Tracking the evolutionary history of Cortinarius species in section Calochroi, with transoceanic...

19
RESEARCH ARTICLE Open Access Tracking the evolutionary history of Cortinarius species in section Calochroi, with transoceanic disjunct distributions Sigisfredo Garnica 1* , Philipp Spahn 2 , Bernhard Oertel 3 , Joseph Ammirati 4 and Franz Oberwinkler 1 Abstract Background: Cortinarius species in section Calochroi display local, clinal and circumboreal patterns of distribution across the Northern Hemisphere where these ectomycorrhizal fungi occur with host trees throughout their geographical range within a continent, or have disjunct intercontinental distributions, the origins of which are not understood. We inferred evolutionary histories of four species, 1) C. arcuatorum, 2) C. aureofulvus, 3) C. elegantior and 4) C. napus, from populations distributed throughout the Old World, and portions of the New World (Central- and North America) based on genetic variation of 154 haplotype internal transcribed spacer (ITS) sequences from 83 population samples. By describing the population structure of these species across their geographical distribution, we attempt to identify their historical migration and patterns of diversification. Results: Models of population structure from nested clade, demographic and coalescent-based analyses revealed genetically differentiated and geographically structured haplotypes in C. arcuatorum and C. elegantior, while C. aureofulvus showed considerably less population structure and C. napus lacked sufficient genetic differentiation to resolve any population structure. Disjunct populations within C. arcuatorum, C. aureofulvus and C. elegantior show little or no morphological differentiation, whereas in C. napus there is a high level of homoplasy and phenotypic plasticity for veil and lamellae colour. The ITS sequences of the type specimens of C. albobrunnoides and C. albobrunnoides var. violaceovelatus were identical to one another and are treated as one species with a wider range of geographic distribution under C. napus. Conclusions: Our results indicate that each of the Calochroi species has undergone a relatively independent evolutionary history, hypothesised as follows: 1) a widely distributed ancestral population of C. arcuatorum diverged into distinctive sympatric populations in the New World; 2) two divergent lineages in C. elegantior gave rise to the New World and Old World haplotypes, respectively; and 3) the low levels of genetic divergence within C. aureofulvus and C. napus may be the result of more recent demographic population expansions. The scenario of migration via the Bering Land Bridge provides the most probable explanation for contemporaneous disjunct geographic distributions of these species, but it does not offer an explanation for the low degree of genetic divergence between populations of C. aureofulvus and C. napus. Our findings are mostly consistent with the designation of New World allopatric populations as separate species from the European counterpart species C. arcuatorum and C. elegantior. We propose the synonymy of C. albobrunnoides, C. albobrunnoides var. violaceovelatus and C. subpurpureophyllus var. sulphureovelatus with C. napus. The results also reinforce previous observations that linked C. arcuatorum and C. aureofulvus displaying distributions in parts of North America and Europe. Interpretations of the population structure of these fungi suggest that host tree history has heavily influenced their modern distributions; however, the complex issues related to co-migration of these fungi with their tree hosts remain unclear at this time. * Correspondence: [email protected] 1 Organismic Botany, Institute of Evolution and Ecology, University of Tübingen, Auf der Morgenstelle 1, D-72076 Tübingen, Germany Full list of author information is available at the end of the article Garnica et al. BMC Evolutionary Biology 2011, 11:213 http://www.biomedcentral.com/1471-2148/11/213 © 2011 Garnica et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Transcript of Tracking the evolutionary history of Cortinarius species in section Calochroi, with transoceanic...

RESEARCH ARTICLE Open Access

Tracking the evolutionary history of Cortinariusspecies in section Calochroi with transoceanicdisjunct distributionsSigisfredo Garnica1 Philipp Spahn2 Bernhard Oertel3 Joseph Ammirati4 and Franz Oberwinkler1

Abstract

Background Cortinarius species in section Calochroi display local clinal and circumboreal patterns of distributionacross the Northern Hemisphere where these ectomycorrhizal fungi occur with host trees throughout theirgeographical range within a continent or have disjunct intercontinental distributions the origins of which are notunderstood We inferred evolutionary histories of four species 1) C arcuatorum 2) C aureofulvus 3) C elegantiorand 4) C napus from populations distributed throughout the Old World and portions of the New World (Central-and North America) based on genetic variation of 154 haplotype internal transcribed spacer (ITS) sequences from83 population samples By describing the population structure of these species across their geographicaldistribution we attempt to identify their historical migration and patterns of diversification

Results Models of population structure from nested clade demographic and coalescent-based analyses revealedgenetically differentiated and geographically structured haplotypes in C arcuatorum and C elegantior while Caureofulvus showed considerably less population structure and C napus lacked sufficient genetic differentiation toresolve any population structure Disjunct populations within C arcuatorum C aureofulvus and C elegantior showlittle or no morphological differentiation whereas in C napus there is a high level of homoplasy and phenotypicplasticity for veil and lamellae colour The ITS sequences of the type specimens of C albobrunnoides and Calbobrunnoides var violaceovelatus were identical to one another and are treated as one species with a wider rangeof geographic distribution under C napus

Conclusions Our results indicate that each of the Calochroi species has undergone a relatively independentevolutionary history hypothesised as follows 1) a widely distributed ancestral population of C arcuatorum divergedinto distinctive sympatric populations in the New World 2) two divergent lineages in C elegantior gave rise to theNew World and Old World haplotypes respectively and 3) the low levels of genetic divergence within Caureofulvus and C napus may be the result of more recent demographic population expansions The scenario ofmigration via the Bering Land Bridge provides the most probable explanation for contemporaneous disjunctgeographic distributions of these species but it does not offer an explanation for the low degree of geneticdivergence between populations of C aureofulvus and C napus Our findings are mostly consistent with thedesignation of New World allopatric populations as separate species from the European counterpart species Carcuatorum and C elegantior We propose the synonymy of C albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus with C napus The results also reinforce previous observations thatlinked C arcuatorum and C aureofulvus displaying distributions in parts of North America and EuropeInterpretations of the population structure of these fungi suggest that host tree history has heavily influenced theirmodern distributions however the complex issues related to co-migration of these fungi with their tree hostsremain unclear at this time

Correspondence sigisfredogarnicauni-tuebingende1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen GermanyFull list of author information is available at the end of the article

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

copy 2011 Garnica et al licensee BioMed Central Ltd This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (httpcreativecommonsorglicensesby20) which permits unrestricted use distribution and reproduction inany medium provided the original work is properly cited

BackgroundSeveral investigators [1-13] have explored intercontinen-tal patterns of ectomycorrhizal fungus species distribu-tions in a phylogenetic context but there is littleinformation on the origin and patterns of speciation inthese fungi Mushrooms in the genus Cortinarius sectionCalochroi (calochroid clade) with over 100 speciesrepresent one of the most conspicuous ectomycorrhizalmembers of contemporary Northern Hemisphere forestecosystems [14] Previous studies of species in Calochroishow various patterns of geographic distributions includ-ing local (C cisticola) clinal (C arcuatorum) or circum-boreal (C aureofulvus C aureopulverulentus Ccupreorufus C elegantior) distributions In mostinstances fungus species distributions correspond to thatof their host trees within continents and often disjunctdistributions across the Northern Hemisphere show asimilar pattern [15-18] Overall the patterns of speciesdistribution in the Calochroi roughly matches the rangeof genera of the host tree families Fagaceae (CastaneaCastanopsis Chrysolepis Fagus Notholithocarpus Quer-cus) and Pinaceae (Abies Larix Picea Pinus Pseudot-suga Tsuga) [17-21] however a few Calochroi speciesform ectotrophic associations with members of the Betu-laceae (Alnus Corylus Carpinus) [2223] Cistaceae (Cis-tus Helianthenum) [2425] and Malvaceae (Tilia) [2425]Host specificity of Calochroi species may be restricted toa single host species with host-switching events beingless common and often restricted to closely related plantspecies [171820] Interestingly allopatric populationsspecies on an intercontinental scale display relativelyhigh host fidelities at the host genus level suggestingpotential co-evolutionaryco-migratory trends [14]Species with a broad distribution provide excellent

models for examining their present patterns of spatialvariation especially where there is an obligate associa-tion as with an ectotrophic fungus An evaluation ofcontemporary geographic patterns of these species canprovide a better understanding of speciation and habitatrequirement as well as determine host preference andhost-switching events that have occurred over time Todate a major difficulty in many mushroom generaincluding species of Cortinarius section Calochroi is thedifficulty of identifying species [1426] which historicallyhas been a major impediment to understanding theirpatterns of geographical distribution ecology diversityand evolutionary relationships There are a limited num-ber of unique morphological features [14] that can beused to construct robust evolutionary hypotheses withinand between closely related species in section CalochroiBased on similarities in basidiomata morphology andcolour some widespread species with largely allopatricdistributions in the Old World (Europe) and the NewWorld (North America) (eg C arcuatorum C

aureofulvus C aureopulverulentus C cupreorufus (as Corichalceus) and C olivellus) were treated as conspecific[171820] On the other hand the discovery of certainphenotypic differences between European populationsand their North American counterparts has led to therecognition of infraspecific taxa for example C elegan-tior var americanus [17] Similarly detailed observationsof morphological variation between European collectionsof C elegantior have resulted in a number of formssubspecies and varieties [19] In our recent phylogeneticstudy [14] we confirmed close phylogenetic affinitiesbetween some disjunct populationsspecies but we alsofound that several European species and their disjunctNorth American counterparts showed significant diver-gence from one anotherRecently the use of DNA sequence data have provided

an independent tool for delineating species boundariesand for revealing patterns of evolution history identify-ing convergent morphological features and assessingestimating species diversity An evaluation of ITS rDNAsequences published by us [14] revealed distinct intra-andor inter-continental phylogeographic structuringwithin some species sometimes accompanied by rela-tively little morphological divergence In contrast wealso found relatively low levels of genetic divergencewithin and between species with allopatric distributionsin North America and Europe that display substantialcolour variation in major components of the basidio-mata While considerable progress has been made inreconstructing interspecific relationships within the sec-tion Calochroi through analyses of morphological fea-tures and DNA sequences [1426] however furtherstudies are needed to examine how sequence divergencerelates to the separation or divergence of species overtime at continental and inter-continental levelsDifferences in genetic structure among extant species

can be used to infer their evolutionary histories As pre-viously stated rapidly evolving ITS regions of rDNAfrom disjunct species of section Calochroi found inregions of Europe and North America show differentlevels of sequence divergence However to date therehas not been sufficient data to draw consistent conclu-sions concerning the genetic variation and divergence ofthese species For example maximum likelihood (ML)analyses [14] suggest a complex evolutionary historywithin C arcuatorum including divergence into fourdistinctive monophyletic groups with similar morpholo-gical features In contrast a close phylogenetic relation-ship was found between species considered to bemorphologically separate from one another European Cnapus and North American C albobrunnoides (varsalbobrunnoides and violaceovelatus and C subpurpureo-phyllus var sulphureovelatus) To gain a more completeunderstanding of the evolutionary histories of Calochroi

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 2 of 19

species (including phylogeographic structure charactervariation and gene flow) we selected four species forstudy 1) C arcuatorum 2) C aureofulvus 3) C elegan-tior and 4) C napus Our species concept is anchored inthe morphology (macroscopic chemical microscopiccharacteristics) of the basidiomata or mushrooms whichis the structure associated with sexual reproduction ofbasidiospores Morphological and molecular evidencesshow that the four species treated here are each clearlydistinct from other known Cortinarius species The fourselected species are somewhat rare to uncommon (espe-cially C aureofulvus and C napus) across their knowngeographical range They are mainly restricted to morecalcareous soils exhibit intercontinental distributionswithin boreo-nemoral areas or in the case of C arcua-torum extend into in the meridional zones [14] Acrosstheir known distributions C arcuatorum is associatedwith oak (Quercus) beech (Fagus) hornbeam (Carpinus)and tan oak (Notholithocarpus) species and with spruce(Picea) and Douglas fir (Pseudotsuga menziesii) inWyoming The tree hosts grow from lowland regions tohigh elevations up to 3000 m in Costa Rica and up to2200 m in Wyoming Three species C aureofulvus Celegantior and C napus are associated with conifers

including spruce (Picea) fir (Abies) pine (Pinus) Dou-glas fir (Pseudotsuga menziesii) and hemlock (Tsuga)species in subalpine regions or less commonly in mid-to low elevation forests (western elegantior clade and Csubpurpureophyllus var sulphureovelatus (= C napus))Although these host tree genera also occur in Asia sofar there is no survey of these Cortinarius species forthis region for inclusion in our studyHere to infer historical events associated with specia-

tion we collected population samples within four spe-cies over geographical distributions spanning EuropeNorth America and Central America (Figure 1) usednested clade demographic and coalescent-basedapproaches to analyse the fast-evolving ITS region ofthe nuclear ribosomal DNA Coalescent-based analysesincorporate random fluctuations in mutation geneticdrift and sampling errors into the calculation of demo-graphic parameters whereas these variations areunderestimated by phylogenetic approaches [27] Weused these analyses to address the following specificquestions What are the mechanisms responsible forthe contemporary patterns of geographic distributionwithin the four species including centres of origin andpotential migration routes And are species with

Figure 1 A map showing the geographic location of the sampling sites of calochroid population samples across Europe NorthAmerica and Central America Population samples of C arcuatorum are represented in blue C aureofulvus in orange C elegantior in red andC napus in green Collection numbers correspond to collections in Additional File 4 Distribution of host plant families green dashed linesindicate distribution of Fagaceae and red dashed lines indicate distribution of Pinaceae respectively

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 3 of 19

allopatric distributions sufficiently distinct from oneanother within and between continents to merit a spe-cific designation To address these questions we com-pared DNA sequence variation of species atcontinental and intercontinental scales We also evalu-ated the phenotypical and ecological features of theCalochroi species Special emphasis was placed ondetecting highly labile traits and their patterns of geo-graphic variation among population samples repre-sented by individual collection containing one toseveral basidiomata from one group or cluster foundgrowing in a relatively small patch (genet)Our findings show that the comparison of population

samples using phenotypical ecological and genetic char-acteristics help to clarify questions concerning the his-torical events involved in the modern distributions ofCalochroi species as well as aspects related to host pre-ferences and host-switching and the use of a combinedphylogenetic morphological and ecological species defi-nition We also discuss the utility and limitations of ITSsequences and phenotypic characteristics as identifica-tion tools for Cortinarius species with wide geographicaldistributions

ResultsDNA amplification phylogenetic placement andrelationshipMore than 100 fungus population samples includingour own collections and those from other herbariawere analysed in this study (see the Methods sectionfor more details) We were unable to amplify the ITSregion for many collections because they were eitherpoorly conserved or too old and the sequence qualitywas too low or we amplified contaminant fungi thatprobably colonised the dried basidiomata Analysis ofITS sequences supported the phylogenetic placementof a total of 83 population samples identified asbelonging to C arcuatorum C aureofulvus C elegan-tior and C napus lineages (data not shown) Each spe-cies was recovered as a monophyletic clade with highbootstrap support (Figure 2) The internal phylogenyof C arcuatorum is composed of four major lineagesa clade including population samples from Costa Rica(I) a clade comprising population samples from Men-docino (Pacific USA) (II) a clade composed of samplesfrom Del Norte Co (Pacific USA) (III) and a largeclade containing mostly European and two samplepopulations from Wyoming (IV) respectively WithinC elegantior several population samples had unre-solved positions (V) a portion of North Americanpopulation samples are distributed into one well-sup-ported subclade (VI) and European population samplesformed a well-supported subclade (VII) The NorthAmerican taxa C albobrunnoides C albobrunnoides

var violaceus and C subpurpureophyllus var sulphur-eovelatus and the European taxon C napus formed amonophyletic group with not internal subclades NewWorld and Old World population samples of C aureo-fulvus clustered together with no significant geneticdifferentiation between them

Patterns of nucleotide variation in the ITS regionSequence divergence of the complete ITS regionbetween intercontinental disjunct populations rangedfrom identical to 463 in C arcuatorum from 016 to035 in C aureofulvus from 068 to 146 in C ele-gantior and from identical to 017 in C napus Afterremoving indels and recombination blocks twenty-fourdistinct ITS haplotypes were identified among the 83population samples from the four species analysed 5haplotypes in C arcuatorum and C aureofulvus 11 inC elegantior and 3 in C napus (Figure 3) Nucleotidediversity (π) estimates for the whole population samplewere 0024 plusmn 19 times 10-3 in C arcuatorum 0002 plusmn 30 times10-4 in C aureofulvus 42 times 10-3 plusmn 42 times 10-4 in C ele-gantior and 50 times 10-4 plusmn 20 times 10-4 in C napus (Table1) When species were partitioned into sampled areaspopulation samples showed similar or lower estimates ofnucleotide diversity The average per-nucleotideexpected heterogeneity θw for the whole sample ofeach species was identical or higher than partitionedsampled areas

Tests of neutrality and population subdivisionThe neutrality tests performed for the whole-speciesdata sampling had non-significant values therefore theequilibrium model of neutral evolution could not berejected Only in C arcuatorum were significant valuesdetected for Fu and Lirsquos D and Fu and Lirsquos F testssuggesting background selection for the whole sample(Table 1) The geographical distributions and frequencyof the haplotypes are given in Additional File 1 Withregard to C arcuatorum haplotype H1 (I) was onlyfound in those samples obtained from Costa Rica (Jardiacutende Dota and Parque Prusia) while haplotypes H2 (II)and H3 (III) are specific to California (Mendocino andDel Norte Counties respectively) The most frequenthaplotype (H4 IV 12 population samples) is distributedthroughout the Old World (Germany France ItalySweden) and also occurs in the New World (Wyoming)see population samples IB19870107 and IB19870239The single haplotype H5 (IV) is restricted to the OldWorld In C aureofulvus haplotypes H1 and H2 arerestricted to the Old World (Austria France GermanySweden) whereas haplotypes H4 and H5 were found inthe New World (Washington State) population samplesHaplotype H3 was found in population samples fromthe Pacific Northwest (Washington State) and Mountain

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 4 of 19

USA (Colorado and Wyoming) In C elegantior H7 (V)and haplotypes H1 (VII) were predominant in the NewWorld (Wyoming) and the Old World (Austria FranceGermany and Switzerland) respectively Haplotypes H2H3 and H4 are restricted to Old World and haplotypesH5 H6 (VI) H8 H9 H10 and H11 were observed onlyin those population samples obtained from Old WorldWith regard to C napus the predominant haplotype H1was shared by different taxa in the New World (PacificOregon and Washington State and Mountain USAWyoming and Colorado) and the Old World (GermanyNorway Sweden) Haplotype H2 was found in theWyoming populations and haplotype H3 is representedby a single population sample (JFA 10070) from thePacific (Washington State)Hudsonrsquos tests were carried out to estimate population

genetic structures within and among population samples

from each area (Table 2) Analyses of populationsshowed significant differentiation between Costa RicaPacific USA (California) and Mountain USA (Wyom-ing)Europe population samples of C arcuatorum (P lt0001 KST = 06540 KS = 50769 KT = 146769)between EuropePacific USA (Washington State Ore-gon) and Mountain USA (Wyoming Colorado) popula-tion samples of C elegantior (P lt 0001 KST = 00833KS = 31645 KT = 34522) and between EuropePacificUSA (Washington State) and Mountain USA (WyomingColorado) population samples of C aureofulvus (P =0001 KST = 03183 KS = 14166 KT = 20784) Non-significant genetic differentiation between EuropePacific(Washington State Oregon) and Mountain USA(Wyoming Colorado) was found for C napus popula-tion samples (P = 02240 KST = 00518 KS = 03998 KT

= 04216) Alternatively significant P values (P lt 0001)

002

GU363495 C aureofulvus IB19930612

AY174851 C elegantior TUB 011394

EU0567051 C aureofulvus JFA 10065

GU363468 C elegantior IB19890189

DQ663212 C arcuatorum TSJ2000-083

EU057002 C arcuatorum JFA 11893

GU363487 C albobrunnoides IB19970162

GU363457 C arcuatorum TUB 019279

GU363493 C albobrunnoides IB19890479

GU363458 C arcuatorum IB19870239

GU363479 C napus TUB 019281

EU0567061 C aureofulvus JFA 12428

EU056997 C elegantior JFA 11693

GU363497 C aureofulvus IB19890428

AY174850 C elegantior TUB 011388

DQ663228 C aureofulvus AB01-09-31

DQ663292 C elegantior AB00-09-120

UDB0002235 C elegantior AT2005138

AY033120 C arcuatorum IB19980286

GU363459 C arcuatorum IB19870107

GU363494 C aureofulvus IB19850209

GU363477 C elegantior JFA 13226 GU363476 C elegantior TUB 019280

UDB000721 C elegantior AT2004126

GU363482 C albobrunnoides IB19870275

GU363490 C albobrunnoides IB9870186

GU363465 C elegantior IB19890226

EU057001 C arcuatorum JFA 12061

EU056998 C elegantior JFA 11452

EU056999 C elegantior JFA 12438

GU363492 C albobrunnoides IB19890242

GU363453 C arcuatorum IB19950564

DQ663229 C aureofulvus TF2004-065

GU363464 C elegantior IB19980248

GU363485 C albobrunnoides IB19910261

GU363496 C aureofulvus IB19870221

GU363472 C elegantior IB19970107a

GU363480 C napus TUB 019282

EU057003 C arcuatorum JFA 11803

GU363452 C arcuatorum IB19950686

GU363456 C arcuatorum TUB 019278

AY669571 C aureofulvus TUB 011831

GU363460 C elegantior IB19980414

AY174822 C arcuatorum TUB 011421

GU363489 C albobrunnoides IB19970303

EU057069 C napus S F44393

UDB000738 C elegantior AT2004277

GU363486 C albobrunnoides IB19910270

GU363470 C elegantior IB19910141

GU363469 C elegantior IB19970249

EU057005 C arcuatorum JFA 11765

DQ663341 C napus TSJ2001-003

GU363483 C albobrunnoides IB19890298

EU057016 C subpurpureophyllus var sulphureovelatus JFA 11723

EU057067 C napus TUB 012717

GU363466 C elegantior IB19970300

EU057015 C albobrunnoides var violaceus JFA 10070

EF014262 C elegantior TUB 012709

GU363463 C elegantior IB20040204

GU363455 C arcuatorum TUB 019283

GU363481C albobrunnoides JFA 12426

GU363473 C elegantior var americanus IB19890059

AY174824 C arcuatorum TUB 011403

GU363454 C arcuatorum IB19950596

AY174823 C arcuatorum TUB 011447

GU363471 C elegantior JFA 11411

GU363484 C albobrunnoides IB19910237

EU057004 C arcuatorum JFA 11766

AJ236067 C napus KH3

GU363474 C elegantior IB19910140

DQ663213 C arcuatorum TSJ2004-043

UDB002242 C napus AT2005156

GU363467 C elegantior IB19870057

GU363461 C elegantior IB20010192

GU363491 C albobrunnoides IB1997303b

EU057060 C arcuatorum JFA 12037

GU363462 C elegantior IB19790599

EU057000 C arcuatorum JFA 12039

DQ663211 C fulvoincarnatus PML5208

GU363478 C elegantior JFA 13287

GU363475 C elegantior IB1989226

GU363488 C albobrunnoides IB19970194

UDB002241 C napus AT2005152

100

100

100

100

100

100

100

98

91

92

99

74

100

Old World

New World

Europe

Mountain USAPacific USACosta Rica

IV

I

II

III

V

VI

VII

Figure 2 Phylogenetic relationships of New World and Old World population samples of Calochroi species based on ML analysis of ITSrDNA sequences The represented tree was computed from 1000 runs and was mid-point rooted Numbers above branches are bootstrapvalues (values lt 70 not shown) from 1000 replicates Assignations I to VII indicate major groups of C arcuatorum and C elegantior discussed inthe text Type specimens are printed in bold

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 5 of 19

for Hudsonrsquos tests indicated that New World and OldWold population samples of C arcuatorum (KST =04619 KS = 78968 KT = 146769) C aureofulvus (KST

= 03731 KS = 13027 KT = 20784) and C elegantior(KST = 03927 KS = 20963 KT = 34522) were geneti-cally differentiated while New World and Old Worldmacro-populations of C napus (P = 06840 KST =-00251 KS = 04322 KT = 04216) were not significantlydifferentiated from each other

Network analysesTo clarify the phylogenetic relationship among haplo-types within each species networks were constructedusing haplotypes from ITS sequences with non-recom-bining blocks (Figure 4a-d) Nested clade analyses for Carcuatorum resulted in one network with two connectedOld World haplotypes (H4 and H5) and three singleunconnected New World haplotypes (H1 H2 and H3)(Figure 4a) Analyses of C aureofulvus (Figure 4b)yielded a network comprising five haplotypes haplotypesH3 H4 and H5 were scattered throughout the NewWorld whereas haplotypes H1 and H2 were restrictedto the Old World The haplotypes H1 H2 H4 and H5were derived from haplotype H3 Similarly in C elegan-tior haplotypes were contained in a single networkwhere haplotype H7 has an ancestral position to theother haplotypes and the Old World haplotypes H1 H2H3 and H4 grouped together (Figure 4c) The OldWorld haplotypes H1 H2 H3 and H4 appear to bemore related to the Pacific haplotypes H10 (Oregon andWashington State) and H11 (Washington State) In Cnapus haplotype H1 is ancestral to the other haplotypes

and was shared by both New World and Old Worldpopulations whereas the haplotypes H2 and H3 wererestricted to Mountain USA (Wyoming) and Pacific(Washington State) in the New World respectively (Fig-ure 4d)

Migration analysesThe four species showed different histories of recombi-nation two distinct recombination blocks were detectedin C arcuatorum four in C aureofulvus nine in C ele-gantior and one in C napus respectively There was noevidence of recombination within each of these blocksMigration parameter estimates including populationmutation rate time of divergence and direction ofmigration estimates between New World and OldWorld macro-population samples are shown in Addi-tional File 2 These analyses showed that migrationsbetween New World and Old World macro-populationsamples within C arcuatorum and C aureofulvus wereapproximately equal in both directions and that thesehave occurred at low rates While gene flow estimates inC elegantior indicate that migration appeared to beslightly higher (with overlapping confidence intervals)from the New World to the Old World (m1 = 034)rather than vice versa (m2 = 016)

Genealogical analysesCoalescent genealogies for C arcuatorum and C elegan-tior showed that two ancestral lineages gave rise to theextant New and Old World haplotypes (Figures 5 and6) Genealogies inferred for C aureofulvus (Figure 7)and C napus (data not shown) did not have enough

C arcuatorum

C aureofulvus

C elegantior

C napus

Position 11111111222224444444445555555 4555 113334444444 1

345588900146778001230355677881233345 7025 6197780112599 19

5272816645120131129602508634040112762 9774 7115616167823 59

Site Number 1111111111222222222233333333 1234 1111 12

1234567890123456789012345678901234567 1234567890123

Consensus GATATTGGCCACCCAAAACATCTCGATTCCGGCTCTC TCCT CCATTTAGTTGTG GT

Site Type tvvttttttttttttttvtvttttvtvvvttvttvtt tttv vtvtvvvtttttt tt

Character Type

ii---i-----i---imdash-iiiii-i-mdashi-----i-i- i--- i-imdash-i-i-i-i- --

H1 (3) GCGT H1 (4) CA H1 (12) ATA H1 (23)

H2 (2) GATGGGTCT H2 (3) C H2 (2) ATGA H2 (3) A

H3 (8) CCATTTTTGAA H3 (6) H3 (1) ATAC H3 (1) C

H4 (12) ATTGTTCTCTACC H4 (4) T H4 (1) ATTA

H5 (1) ATTGGTTCTCTACC H5 (1) T H5 (1) GCA

H6 (13) GC

H7 (36) C

H8 (8) CC

H9 (4) CC

H10 (4) C

H11 (1) CA

Figure 3 Distribution of polymorphic sites after removing indels and recombination blocks in the ITS haplotypes of calochroidspecies t transitions v transversions i phylogenetically informative sites - uninformative sites

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 6 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

BackgroundSeveral investigators [1-13] have explored intercontinen-tal patterns of ectomycorrhizal fungus species distribu-tions in a phylogenetic context but there is littleinformation on the origin and patterns of speciation inthese fungi Mushrooms in the genus Cortinarius sectionCalochroi (calochroid clade) with over 100 speciesrepresent one of the most conspicuous ectomycorrhizalmembers of contemporary Northern Hemisphere forestecosystems [14] Previous studies of species in Calochroishow various patterns of geographic distributions includ-ing local (C cisticola) clinal (C arcuatorum) or circum-boreal (C aureofulvus C aureopulverulentus Ccupreorufus C elegantior) distributions In mostinstances fungus species distributions correspond to thatof their host trees within continents and often disjunctdistributions across the Northern Hemisphere show asimilar pattern [15-18] Overall the patterns of speciesdistribution in the Calochroi roughly matches the rangeof genera of the host tree families Fagaceae (CastaneaCastanopsis Chrysolepis Fagus Notholithocarpus Quer-cus) and Pinaceae (Abies Larix Picea Pinus Pseudot-suga Tsuga) [17-21] however a few Calochroi speciesform ectotrophic associations with members of the Betu-laceae (Alnus Corylus Carpinus) [2223] Cistaceae (Cis-tus Helianthenum) [2425] and Malvaceae (Tilia) [2425]Host specificity of Calochroi species may be restricted toa single host species with host-switching events beingless common and often restricted to closely related plantspecies [171820] Interestingly allopatric populationsspecies on an intercontinental scale display relativelyhigh host fidelities at the host genus level suggestingpotential co-evolutionaryco-migratory trends [14]Species with a broad distribution provide excellent

models for examining their present patterns of spatialvariation especially where there is an obligate associa-tion as with an ectotrophic fungus An evaluation ofcontemporary geographic patterns of these species canprovide a better understanding of speciation and habitatrequirement as well as determine host preference andhost-switching events that have occurred over time Todate a major difficulty in many mushroom generaincluding species of Cortinarius section Calochroi is thedifficulty of identifying species [1426] which historicallyhas been a major impediment to understanding theirpatterns of geographical distribution ecology diversityand evolutionary relationships There are a limited num-ber of unique morphological features [14] that can beused to construct robust evolutionary hypotheses withinand between closely related species in section CalochroiBased on similarities in basidiomata morphology andcolour some widespread species with largely allopatricdistributions in the Old World (Europe) and the NewWorld (North America) (eg C arcuatorum C

aureofulvus C aureopulverulentus C cupreorufus (as Corichalceus) and C olivellus) were treated as conspecific[171820] On the other hand the discovery of certainphenotypic differences between European populationsand their North American counterparts has led to therecognition of infraspecific taxa for example C elegan-tior var americanus [17] Similarly detailed observationsof morphological variation between European collectionsof C elegantior have resulted in a number of formssubspecies and varieties [19] In our recent phylogeneticstudy [14] we confirmed close phylogenetic affinitiesbetween some disjunct populationsspecies but we alsofound that several European species and their disjunctNorth American counterparts showed significant diver-gence from one anotherRecently the use of DNA sequence data have provided

an independent tool for delineating species boundariesand for revealing patterns of evolution history identify-ing convergent morphological features and assessingestimating species diversity An evaluation of ITS rDNAsequences published by us [14] revealed distinct intra-andor inter-continental phylogeographic structuringwithin some species sometimes accompanied by rela-tively little morphological divergence In contrast wealso found relatively low levels of genetic divergencewithin and between species with allopatric distributionsin North America and Europe that display substantialcolour variation in major components of the basidio-mata While considerable progress has been made inreconstructing interspecific relationships within the sec-tion Calochroi through analyses of morphological fea-tures and DNA sequences [1426] however furtherstudies are needed to examine how sequence divergencerelates to the separation or divergence of species overtime at continental and inter-continental levelsDifferences in genetic structure among extant species

can be used to infer their evolutionary histories As pre-viously stated rapidly evolving ITS regions of rDNAfrom disjunct species of section Calochroi found inregions of Europe and North America show differentlevels of sequence divergence However to date therehas not been sufficient data to draw consistent conclu-sions concerning the genetic variation and divergence ofthese species For example maximum likelihood (ML)analyses [14] suggest a complex evolutionary historywithin C arcuatorum including divergence into fourdistinctive monophyletic groups with similar morpholo-gical features In contrast a close phylogenetic relation-ship was found between species considered to bemorphologically separate from one another European Cnapus and North American C albobrunnoides (varsalbobrunnoides and violaceovelatus and C subpurpureo-phyllus var sulphureovelatus) To gain a more completeunderstanding of the evolutionary histories of Calochroi

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 2 of 19

species (including phylogeographic structure charactervariation and gene flow) we selected four species forstudy 1) C arcuatorum 2) C aureofulvus 3) C elegan-tior and 4) C napus Our species concept is anchored inthe morphology (macroscopic chemical microscopiccharacteristics) of the basidiomata or mushrooms whichis the structure associated with sexual reproduction ofbasidiospores Morphological and molecular evidencesshow that the four species treated here are each clearlydistinct from other known Cortinarius species The fourselected species are somewhat rare to uncommon (espe-cially C aureofulvus and C napus) across their knowngeographical range They are mainly restricted to morecalcareous soils exhibit intercontinental distributionswithin boreo-nemoral areas or in the case of C arcua-torum extend into in the meridional zones [14] Acrosstheir known distributions C arcuatorum is associatedwith oak (Quercus) beech (Fagus) hornbeam (Carpinus)and tan oak (Notholithocarpus) species and with spruce(Picea) and Douglas fir (Pseudotsuga menziesii) inWyoming The tree hosts grow from lowland regions tohigh elevations up to 3000 m in Costa Rica and up to2200 m in Wyoming Three species C aureofulvus Celegantior and C napus are associated with conifers

including spruce (Picea) fir (Abies) pine (Pinus) Dou-glas fir (Pseudotsuga menziesii) and hemlock (Tsuga)species in subalpine regions or less commonly in mid-to low elevation forests (western elegantior clade and Csubpurpureophyllus var sulphureovelatus (= C napus))Although these host tree genera also occur in Asia sofar there is no survey of these Cortinarius species forthis region for inclusion in our studyHere to infer historical events associated with specia-

tion we collected population samples within four spe-cies over geographical distributions spanning EuropeNorth America and Central America (Figure 1) usednested clade demographic and coalescent-basedapproaches to analyse the fast-evolving ITS region ofthe nuclear ribosomal DNA Coalescent-based analysesincorporate random fluctuations in mutation geneticdrift and sampling errors into the calculation of demo-graphic parameters whereas these variations areunderestimated by phylogenetic approaches [27] Weused these analyses to address the following specificquestions What are the mechanisms responsible forthe contemporary patterns of geographic distributionwithin the four species including centres of origin andpotential migration routes And are species with

Figure 1 A map showing the geographic location of the sampling sites of calochroid population samples across Europe NorthAmerica and Central America Population samples of C arcuatorum are represented in blue C aureofulvus in orange C elegantior in red andC napus in green Collection numbers correspond to collections in Additional File 4 Distribution of host plant families green dashed linesindicate distribution of Fagaceae and red dashed lines indicate distribution of Pinaceae respectively

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 3 of 19

allopatric distributions sufficiently distinct from oneanother within and between continents to merit a spe-cific designation To address these questions we com-pared DNA sequence variation of species atcontinental and intercontinental scales We also evalu-ated the phenotypical and ecological features of theCalochroi species Special emphasis was placed ondetecting highly labile traits and their patterns of geo-graphic variation among population samples repre-sented by individual collection containing one toseveral basidiomata from one group or cluster foundgrowing in a relatively small patch (genet)Our findings show that the comparison of population

samples using phenotypical ecological and genetic char-acteristics help to clarify questions concerning the his-torical events involved in the modern distributions ofCalochroi species as well as aspects related to host pre-ferences and host-switching and the use of a combinedphylogenetic morphological and ecological species defi-nition We also discuss the utility and limitations of ITSsequences and phenotypic characteristics as identifica-tion tools for Cortinarius species with wide geographicaldistributions

ResultsDNA amplification phylogenetic placement andrelationshipMore than 100 fungus population samples includingour own collections and those from other herbariawere analysed in this study (see the Methods sectionfor more details) We were unable to amplify the ITSregion for many collections because they were eitherpoorly conserved or too old and the sequence qualitywas too low or we amplified contaminant fungi thatprobably colonised the dried basidiomata Analysis ofITS sequences supported the phylogenetic placementof a total of 83 population samples identified asbelonging to C arcuatorum C aureofulvus C elegan-tior and C napus lineages (data not shown) Each spe-cies was recovered as a monophyletic clade with highbootstrap support (Figure 2) The internal phylogenyof C arcuatorum is composed of four major lineagesa clade including population samples from Costa Rica(I) a clade comprising population samples from Men-docino (Pacific USA) (II) a clade composed of samplesfrom Del Norte Co (Pacific USA) (III) and a largeclade containing mostly European and two samplepopulations from Wyoming (IV) respectively WithinC elegantior several population samples had unre-solved positions (V) a portion of North Americanpopulation samples are distributed into one well-sup-ported subclade (VI) and European population samplesformed a well-supported subclade (VII) The NorthAmerican taxa C albobrunnoides C albobrunnoides

var violaceus and C subpurpureophyllus var sulphur-eovelatus and the European taxon C napus formed amonophyletic group with not internal subclades NewWorld and Old World population samples of C aureo-fulvus clustered together with no significant geneticdifferentiation between them

Patterns of nucleotide variation in the ITS regionSequence divergence of the complete ITS regionbetween intercontinental disjunct populations rangedfrom identical to 463 in C arcuatorum from 016 to035 in C aureofulvus from 068 to 146 in C ele-gantior and from identical to 017 in C napus Afterremoving indels and recombination blocks twenty-fourdistinct ITS haplotypes were identified among the 83population samples from the four species analysed 5haplotypes in C arcuatorum and C aureofulvus 11 inC elegantior and 3 in C napus (Figure 3) Nucleotidediversity (π) estimates for the whole population samplewere 0024 plusmn 19 times 10-3 in C arcuatorum 0002 plusmn 30 times10-4 in C aureofulvus 42 times 10-3 plusmn 42 times 10-4 in C ele-gantior and 50 times 10-4 plusmn 20 times 10-4 in C napus (Table1) When species were partitioned into sampled areaspopulation samples showed similar or lower estimates ofnucleotide diversity The average per-nucleotideexpected heterogeneity θw for the whole sample ofeach species was identical or higher than partitionedsampled areas

Tests of neutrality and population subdivisionThe neutrality tests performed for the whole-speciesdata sampling had non-significant values therefore theequilibrium model of neutral evolution could not berejected Only in C arcuatorum were significant valuesdetected for Fu and Lirsquos D and Fu and Lirsquos F testssuggesting background selection for the whole sample(Table 1) The geographical distributions and frequencyof the haplotypes are given in Additional File 1 Withregard to C arcuatorum haplotype H1 (I) was onlyfound in those samples obtained from Costa Rica (Jardiacutende Dota and Parque Prusia) while haplotypes H2 (II)and H3 (III) are specific to California (Mendocino andDel Norte Counties respectively) The most frequenthaplotype (H4 IV 12 population samples) is distributedthroughout the Old World (Germany France ItalySweden) and also occurs in the New World (Wyoming)see population samples IB19870107 and IB19870239The single haplotype H5 (IV) is restricted to the OldWorld In C aureofulvus haplotypes H1 and H2 arerestricted to the Old World (Austria France GermanySweden) whereas haplotypes H4 and H5 were found inthe New World (Washington State) population samplesHaplotype H3 was found in population samples fromthe Pacific Northwest (Washington State) and Mountain

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 4 of 19

USA (Colorado and Wyoming) In C elegantior H7 (V)and haplotypes H1 (VII) were predominant in the NewWorld (Wyoming) and the Old World (Austria FranceGermany and Switzerland) respectively Haplotypes H2H3 and H4 are restricted to Old World and haplotypesH5 H6 (VI) H8 H9 H10 and H11 were observed onlyin those population samples obtained from Old WorldWith regard to C napus the predominant haplotype H1was shared by different taxa in the New World (PacificOregon and Washington State and Mountain USAWyoming and Colorado) and the Old World (GermanyNorway Sweden) Haplotype H2 was found in theWyoming populations and haplotype H3 is representedby a single population sample (JFA 10070) from thePacific (Washington State)Hudsonrsquos tests were carried out to estimate population

genetic structures within and among population samples

from each area (Table 2) Analyses of populationsshowed significant differentiation between Costa RicaPacific USA (California) and Mountain USA (Wyom-ing)Europe population samples of C arcuatorum (P lt0001 KST = 06540 KS = 50769 KT = 146769)between EuropePacific USA (Washington State Ore-gon) and Mountain USA (Wyoming Colorado) popula-tion samples of C elegantior (P lt 0001 KST = 00833KS = 31645 KT = 34522) and between EuropePacificUSA (Washington State) and Mountain USA (WyomingColorado) population samples of C aureofulvus (P =0001 KST = 03183 KS = 14166 KT = 20784) Non-significant genetic differentiation between EuropePacific(Washington State Oregon) and Mountain USA(Wyoming Colorado) was found for C napus popula-tion samples (P = 02240 KST = 00518 KS = 03998 KT

= 04216) Alternatively significant P values (P lt 0001)

002

GU363495 C aureofulvus IB19930612

AY174851 C elegantior TUB 011394

EU0567051 C aureofulvus JFA 10065

GU363468 C elegantior IB19890189

DQ663212 C arcuatorum TSJ2000-083

EU057002 C arcuatorum JFA 11893

GU363487 C albobrunnoides IB19970162

GU363457 C arcuatorum TUB 019279

GU363493 C albobrunnoides IB19890479

GU363458 C arcuatorum IB19870239

GU363479 C napus TUB 019281

EU0567061 C aureofulvus JFA 12428

EU056997 C elegantior JFA 11693

GU363497 C aureofulvus IB19890428

AY174850 C elegantior TUB 011388

DQ663228 C aureofulvus AB01-09-31

DQ663292 C elegantior AB00-09-120

UDB0002235 C elegantior AT2005138

AY033120 C arcuatorum IB19980286

GU363459 C arcuatorum IB19870107

GU363494 C aureofulvus IB19850209

GU363477 C elegantior JFA 13226 GU363476 C elegantior TUB 019280

UDB000721 C elegantior AT2004126

GU363482 C albobrunnoides IB19870275

GU363490 C albobrunnoides IB9870186

GU363465 C elegantior IB19890226

EU057001 C arcuatorum JFA 12061

EU056998 C elegantior JFA 11452

EU056999 C elegantior JFA 12438

GU363492 C albobrunnoides IB19890242

GU363453 C arcuatorum IB19950564

DQ663229 C aureofulvus TF2004-065

GU363464 C elegantior IB19980248

GU363485 C albobrunnoides IB19910261

GU363496 C aureofulvus IB19870221

GU363472 C elegantior IB19970107a

GU363480 C napus TUB 019282

EU057003 C arcuatorum JFA 11803

GU363452 C arcuatorum IB19950686

GU363456 C arcuatorum TUB 019278

AY669571 C aureofulvus TUB 011831

GU363460 C elegantior IB19980414

AY174822 C arcuatorum TUB 011421

GU363489 C albobrunnoides IB19970303

EU057069 C napus S F44393

UDB000738 C elegantior AT2004277

GU363486 C albobrunnoides IB19910270

GU363470 C elegantior IB19910141

GU363469 C elegantior IB19970249

EU057005 C arcuatorum JFA 11765

DQ663341 C napus TSJ2001-003

GU363483 C albobrunnoides IB19890298

EU057016 C subpurpureophyllus var sulphureovelatus JFA 11723

EU057067 C napus TUB 012717

GU363466 C elegantior IB19970300

EU057015 C albobrunnoides var violaceus JFA 10070

EF014262 C elegantior TUB 012709

GU363463 C elegantior IB20040204

GU363455 C arcuatorum TUB 019283

GU363481C albobrunnoides JFA 12426

GU363473 C elegantior var americanus IB19890059

AY174824 C arcuatorum TUB 011403

GU363454 C arcuatorum IB19950596

AY174823 C arcuatorum TUB 011447

GU363471 C elegantior JFA 11411

GU363484 C albobrunnoides IB19910237

EU057004 C arcuatorum JFA 11766

AJ236067 C napus KH3

GU363474 C elegantior IB19910140

DQ663213 C arcuatorum TSJ2004-043

UDB002242 C napus AT2005156

GU363467 C elegantior IB19870057

GU363461 C elegantior IB20010192

GU363491 C albobrunnoides IB1997303b

EU057060 C arcuatorum JFA 12037

GU363462 C elegantior IB19790599

EU057000 C arcuatorum JFA 12039

DQ663211 C fulvoincarnatus PML5208

GU363478 C elegantior JFA 13287

GU363475 C elegantior IB1989226

GU363488 C albobrunnoides IB19970194

UDB002241 C napus AT2005152

100

100

100

100

100

100

100

98

91

92

99

74

100

Old World

New World

Europe

Mountain USAPacific USACosta Rica

IV

I

II

III

V

VI

VII

Figure 2 Phylogenetic relationships of New World and Old World population samples of Calochroi species based on ML analysis of ITSrDNA sequences The represented tree was computed from 1000 runs and was mid-point rooted Numbers above branches are bootstrapvalues (values lt 70 not shown) from 1000 replicates Assignations I to VII indicate major groups of C arcuatorum and C elegantior discussed inthe text Type specimens are printed in bold

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 5 of 19

for Hudsonrsquos tests indicated that New World and OldWold population samples of C arcuatorum (KST =04619 KS = 78968 KT = 146769) C aureofulvus (KST

= 03731 KS = 13027 KT = 20784) and C elegantior(KST = 03927 KS = 20963 KT = 34522) were geneti-cally differentiated while New World and Old Worldmacro-populations of C napus (P = 06840 KST =-00251 KS = 04322 KT = 04216) were not significantlydifferentiated from each other

Network analysesTo clarify the phylogenetic relationship among haplo-types within each species networks were constructedusing haplotypes from ITS sequences with non-recom-bining blocks (Figure 4a-d) Nested clade analyses for Carcuatorum resulted in one network with two connectedOld World haplotypes (H4 and H5) and three singleunconnected New World haplotypes (H1 H2 and H3)(Figure 4a) Analyses of C aureofulvus (Figure 4b)yielded a network comprising five haplotypes haplotypesH3 H4 and H5 were scattered throughout the NewWorld whereas haplotypes H1 and H2 were restrictedto the Old World The haplotypes H1 H2 H4 and H5were derived from haplotype H3 Similarly in C elegan-tior haplotypes were contained in a single networkwhere haplotype H7 has an ancestral position to theother haplotypes and the Old World haplotypes H1 H2H3 and H4 grouped together (Figure 4c) The OldWorld haplotypes H1 H2 H3 and H4 appear to bemore related to the Pacific haplotypes H10 (Oregon andWashington State) and H11 (Washington State) In Cnapus haplotype H1 is ancestral to the other haplotypes

and was shared by both New World and Old Worldpopulations whereas the haplotypes H2 and H3 wererestricted to Mountain USA (Wyoming) and Pacific(Washington State) in the New World respectively (Fig-ure 4d)

Migration analysesThe four species showed different histories of recombi-nation two distinct recombination blocks were detectedin C arcuatorum four in C aureofulvus nine in C ele-gantior and one in C napus respectively There was noevidence of recombination within each of these blocksMigration parameter estimates including populationmutation rate time of divergence and direction ofmigration estimates between New World and OldWorld macro-population samples are shown in Addi-tional File 2 These analyses showed that migrationsbetween New World and Old World macro-populationsamples within C arcuatorum and C aureofulvus wereapproximately equal in both directions and that thesehave occurred at low rates While gene flow estimates inC elegantior indicate that migration appeared to beslightly higher (with overlapping confidence intervals)from the New World to the Old World (m1 = 034)rather than vice versa (m2 = 016)

Genealogical analysesCoalescent genealogies for C arcuatorum and C elegan-tior showed that two ancestral lineages gave rise to theextant New and Old World haplotypes (Figures 5 and6) Genealogies inferred for C aureofulvus (Figure 7)and C napus (data not shown) did not have enough

C arcuatorum

C aureofulvus

C elegantior

C napus

Position 11111111222224444444445555555 4555 113334444444 1

345588900146778001230355677881233345 7025 6197780112599 19

5272816645120131129602508634040112762 9774 7115616167823 59

Site Number 1111111111222222222233333333 1234 1111 12

1234567890123456789012345678901234567 1234567890123

Consensus GATATTGGCCACCCAAAACATCTCGATTCCGGCTCTC TCCT CCATTTAGTTGTG GT

Site Type tvvttttttttttttttvtvttttvtvvvttvttvtt tttv vtvtvvvtttttt tt

Character Type

ii---i-----i---imdash-iiiii-i-mdashi-----i-i- i--- i-imdash-i-i-i-i- --

H1 (3) GCGT H1 (4) CA H1 (12) ATA H1 (23)

H2 (2) GATGGGTCT H2 (3) C H2 (2) ATGA H2 (3) A

H3 (8) CCATTTTTGAA H3 (6) H3 (1) ATAC H3 (1) C

H4 (12) ATTGTTCTCTACC H4 (4) T H4 (1) ATTA

H5 (1) ATTGGTTCTCTACC H5 (1) T H5 (1) GCA

H6 (13) GC

H7 (36) C

H8 (8) CC

H9 (4) CC

H10 (4) C

H11 (1) CA

Figure 3 Distribution of polymorphic sites after removing indels and recombination blocks in the ITS haplotypes of calochroidspecies t transitions v transversions i phylogenetically informative sites - uninformative sites

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 6 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

species (including phylogeographic structure charactervariation and gene flow) we selected four species forstudy 1) C arcuatorum 2) C aureofulvus 3) C elegan-tior and 4) C napus Our species concept is anchored inthe morphology (macroscopic chemical microscopiccharacteristics) of the basidiomata or mushrooms whichis the structure associated with sexual reproduction ofbasidiospores Morphological and molecular evidencesshow that the four species treated here are each clearlydistinct from other known Cortinarius species The fourselected species are somewhat rare to uncommon (espe-cially C aureofulvus and C napus) across their knowngeographical range They are mainly restricted to morecalcareous soils exhibit intercontinental distributionswithin boreo-nemoral areas or in the case of C arcua-torum extend into in the meridional zones [14] Acrosstheir known distributions C arcuatorum is associatedwith oak (Quercus) beech (Fagus) hornbeam (Carpinus)and tan oak (Notholithocarpus) species and with spruce(Picea) and Douglas fir (Pseudotsuga menziesii) inWyoming The tree hosts grow from lowland regions tohigh elevations up to 3000 m in Costa Rica and up to2200 m in Wyoming Three species C aureofulvus Celegantior and C napus are associated with conifers

including spruce (Picea) fir (Abies) pine (Pinus) Dou-glas fir (Pseudotsuga menziesii) and hemlock (Tsuga)species in subalpine regions or less commonly in mid-to low elevation forests (western elegantior clade and Csubpurpureophyllus var sulphureovelatus (= C napus))Although these host tree genera also occur in Asia sofar there is no survey of these Cortinarius species forthis region for inclusion in our studyHere to infer historical events associated with specia-

tion we collected population samples within four spe-cies over geographical distributions spanning EuropeNorth America and Central America (Figure 1) usednested clade demographic and coalescent-basedapproaches to analyse the fast-evolving ITS region ofthe nuclear ribosomal DNA Coalescent-based analysesincorporate random fluctuations in mutation geneticdrift and sampling errors into the calculation of demo-graphic parameters whereas these variations areunderestimated by phylogenetic approaches [27] Weused these analyses to address the following specificquestions What are the mechanisms responsible forthe contemporary patterns of geographic distributionwithin the four species including centres of origin andpotential migration routes And are species with

Figure 1 A map showing the geographic location of the sampling sites of calochroid population samples across Europe NorthAmerica and Central America Population samples of C arcuatorum are represented in blue C aureofulvus in orange C elegantior in red andC napus in green Collection numbers correspond to collections in Additional File 4 Distribution of host plant families green dashed linesindicate distribution of Fagaceae and red dashed lines indicate distribution of Pinaceae respectively

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 3 of 19

allopatric distributions sufficiently distinct from oneanother within and between continents to merit a spe-cific designation To address these questions we com-pared DNA sequence variation of species atcontinental and intercontinental scales We also evalu-ated the phenotypical and ecological features of theCalochroi species Special emphasis was placed ondetecting highly labile traits and their patterns of geo-graphic variation among population samples repre-sented by individual collection containing one toseveral basidiomata from one group or cluster foundgrowing in a relatively small patch (genet)Our findings show that the comparison of population

samples using phenotypical ecological and genetic char-acteristics help to clarify questions concerning the his-torical events involved in the modern distributions ofCalochroi species as well as aspects related to host pre-ferences and host-switching and the use of a combinedphylogenetic morphological and ecological species defi-nition We also discuss the utility and limitations of ITSsequences and phenotypic characteristics as identifica-tion tools for Cortinarius species with wide geographicaldistributions

ResultsDNA amplification phylogenetic placement andrelationshipMore than 100 fungus population samples includingour own collections and those from other herbariawere analysed in this study (see the Methods sectionfor more details) We were unable to amplify the ITSregion for many collections because they were eitherpoorly conserved or too old and the sequence qualitywas too low or we amplified contaminant fungi thatprobably colonised the dried basidiomata Analysis ofITS sequences supported the phylogenetic placementof a total of 83 population samples identified asbelonging to C arcuatorum C aureofulvus C elegan-tior and C napus lineages (data not shown) Each spe-cies was recovered as a monophyletic clade with highbootstrap support (Figure 2) The internal phylogenyof C arcuatorum is composed of four major lineagesa clade including population samples from Costa Rica(I) a clade comprising population samples from Men-docino (Pacific USA) (II) a clade composed of samplesfrom Del Norte Co (Pacific USA) (III) and a largeclade containing mostly European and two samplepopulations from Wyoming (IV) respectively WithinC elegantior several population samples had unre-solved positions (V) a portion of North Americanpopulation samples are distributed into one well-sup-ported subclade (VI) and European population samplesformed a well-supported subclade (VII) The NorthAmerican taxa C albobrunnoides C albobrunnoides

var violaceus and C subpurpureophyllus var sulphur-eovelatus and the European taxon C napus formed amonophyletic group with not internal subclades NewWorld and Old World population samples of C aureo-fulvus clustered together with no significant geneticdifferentiation between them

Patterns of nucleotide variation in the ITS regionSequence divergence of the complete ITS regionbetween intercontinental disjunct populations rangedfrom identical to 463 in C arcuatorum from 016 to035 in C aureofulvus from 068 to 146 in C ele-gantior and from identical to 017 in C napus Afterremoving indels and recombination blocks twenty-fourdistinct ITS haplotypes were identified among the 83population samples from the four species analysed 5haplotypes in C arcuatorum and C aureofulvus 11 inC elegantior and 3 in C napus (Figure 3) Nucleotidediversity (π) estimates for the whole population samplewere 0024 plusmn 19 times 10-3 in C arcuatorum 0002 plusmn 30 times10-4 in C aureofulvus 42 times 10-3 plusmn 42 times 10-4 in C ele-gantior and 50 times 10-4 plusmn 20 times 10-4 in C napus (Table1) When species were partitioned into sampled areaspopulation samples showed similar or lower estimates ofnucleotide diversity The average per-nucleotideexpected heterogeneity θw for the whole sample ofeach species was identical or higher than partitionedsampled areas

Tests of neutrality and population subdivisionThe neutrality tests performed for the whole-speciesdata sampling had non-significant values therefore theequilibrium model of neutral evolution could not berejected Only in C arcuatorum were significant valuesdetected for Fu and Lirsquos D and Fu and Lirsquos F testssuggesting background selection for the whole sample(Table 1) The geographical distributions and frequencyof the haplotypes are given in Additional File 1 Withregard to C arcuatorum haplotype H1 (I) was onlyfound in those samples obtained from Costa Rica (Jardiacutende Dota and Parque Prusia) while haplotypes H2 (II)and H3 (III) are specific to California (Mendocino andDel Norte Counties respectively) The most frequenthaplotype (H4 IV 12 population samples) is distributedthroughout the Old World (Germany France ItalySweden) and also occurs in the New World (Wyoming)see population samples IB19870107 and IB19870239The single haplotype H5 (IV) is restricted to the OldWorld In C aureofulvus haplotypes H1 and H2 arerestricted to the Old World (Austria France GermanySweden) whereas haplotypes H4 and H5 were found inthe New World (Washington State) population samplesHaplotype H3 was found in population samples fromthe Pacific Northwest (Washington State) and Mountain

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 4 of 19

USA (Colorado and Wyoming) In C elegantior H7 (V)and haplotypes H1 (VII) were predominant in the NewWorld (Wyoming) and the Old World (Austria FranceGermany and Switzerland) respectively Haplotypes H2H3 and H4 are restricted to Old World and haplotypesH5 H6 (VI) H8 H9 H10 and H11 were observed onlyin those population samples obtained from Old WorldWith regard to C napus the predominant haplotype H1was shared by different taxa in the New World (PacificOregon and Washington State and Mountain USAWyoming and Colorado) and the Old World (GermanyNorway Sweden) Haplotype H2 was found in theWyoming populations and haplotype H3 is representedby a single population sample (JFA 10070) from thePacific (Washington State)Hudsonrsquos tests were carried out to estimate population

genetic structures within and among population samples

from each area (Table 2) Analyses of populationsshowed significant differentiation between Costa RicaPacific USA (California) and Mountain USA (Wyom-ing)Europe population samples of C arcuatorum (P lt0001 KST = 06540 KS = 50769 KT = 146769)between EuropePacific USA (Washington State Ore-gon) and Mountain USA (Wyoming Colorado) popula-tion samples of C elegantior (P lt 0001 KST = 00833KS = 31645 KT = 34522) and between EuropePacificUSA (Washington State) and Mountain USA (WyomingColorado) population samples of C aureofulvus (P =0001 KST = 03183 KS = 14166 KT = 20784) Non-significant genetic differentiation between EuropePacific(Washington State Oregon) and Mountain USA(Wyoming Colorado) was found for C napus popula-tion samples (P = 02240 KST = 00518 KS = 03998 KT

= 04216) Alternatively significant P values (P lt 0001)

002

GU363495 C aureofulvus IB19930612

AY174851 C elegantior TUB 011394

EU0567051 C aureofulvus JFA 10065

GU363468 C elegantior IB19890189

DQ663212 C arcuatorum TSJ2000-083

EU057002 C arcuatorum JFA 11893

GU363487 C albobrunnoides IB19970162

GU363457 C arcuatorum TUB 019279

GU363493 C albobrunnoides IB19890479

GU363458 C arcuatorum IB19870239

GU363479 C napus TUB 019281

EU0567061 C aureofulvus JFA 12428

EU056997 C elegantior JFA 11693

GU363497 C aureofulvus IB19890428

AY174850 C elegantior TUB 011388

DQ663228 C aureofulvus AB01-09-31

DQ663292 C elegantior AB00-09-120

UDB0002235 C elegantior AT2005138

AY033120 C arcuatorum IB19980286

GU363459 C arcuatorum IB19870107

GU363494 C aureofulvus IB19850209

GU363477 C elegantior JFA 13226 GU363476 C elegantior TUB 019280

UDB000721 C elegantior AT2004126

GU363482 C albobrunnoides IB19870275

GU363490 C albobrunnoides IB9870186

GU363465 C elegantior IB19890226

EU057001 C arcuatorum JFA 12061

EU056998 C elegantior JFA 11452

EU056999 C elegantior JFA 12438

GU363492 C albobrunnoides IB19890242

GU363453 C arcuatorum IB19950564

DQ663229 C aureofulvus TF2004-065

GU363464 C elegantior IB19980248

GU363485 C albobrunnoides IB19910261

GU363496 C aureofulvus IB19870221

GU363472 C elegantior IB19970107a

GU363480 C napus TUB 019282

EU057003 C arcuatorum JFA 11803

GU363452 C arcuatorum IB19950686

GU363456 C arcuatorum TUB 019278

AY669571 C aureofulvus TUB 011831

GU363460 C elegantior IB19980414

AY174822 C arcuatorum TUB 011421

GU363489 C albobrunnoides IB19970303

EU057069 C napus S F44393

UDB000738 C elegantior AT2004277

GU363486 C albobrunnoides IB19910270

GU363470 C elegantior IB19910141

GU363469 C elegantior IB19970249

EU057005 C arcuatorum JFA 11765

DQ663341 C napus TSJ2001-003

GU363483 C albobrunnoides IB19890298

EU057016 C subpurpureophyllus var sulphureovelatus JFA 11723

EU057067 C napus TUB 012717

GU363466 C elegantior IB19970300

EU057015 C albobrunnoides var violaceus JFA 10070

EF014262 C elegantior TUB 012709

GU363463 C elegantior IB20040204

GU363455 C arcuatorum TUB 019283

GU363481C albobrunnoides JFA 12426

GU363473 C elegantior var americanus IB19890059

AY174824 C arcuatorum TUB 011403

GU363454 C arcuatorum IB19950596

AY174823 C arcuatorum TUB 011447

GU363471 C elegantior JFA 11411

GU363484 C albobrunnoides IB19910237

EU057004 C arcuatorum JFA 11766

AJ236067 C napus KH3

GU363474 C elegantior IB19910140

DQ663213 C arcuatorum TSJ2004-043

UDB002242 C napus AT2005156

GU363467 C elegantior IB19870057

GU363461 C elegantior IB20010192

GU363491 C albobrunnoides IB1997303b

EU057060 C arcuatorum JFA 12037

GU363462 C elegantior IB19790599

EU057000 C arcuatorum JFA 12039

DQ663211 C fulvoincarnatus PML5208

GU363478 C elegantior JFA 13287

GU363475 C elegantior IB1989226

GU363488 C albobrunnoides IB19970194

UDB002241 C napus AT2005152

100

100

100

100

100

100

100

98

91

92

99

74

100

Old World

New World

Europe

Mountain USAPacific USACosta Rica

IV

I

II

III

V

VI

VII

Figure 2 Phylogenetic relationships of New World and Old World population samples of Calochroi species based on ML analysis of ITSrDNA sequences The represented tree was computed from 1000 runs and was mid-point rooted Numbers above branches are bootstrapvalues (values lt 70 not shown) from 1000 replicates Assignations I to VII indicate major groups of C arcuatorum and C elegantior discussed inthe text Type specimens are printed in bold

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 5 of 19

for Hudsonrsquos tests indicated that New World and OldWold population samples of C arcuatorum (KST =04619 KS = 78968 KT = 146769) C aureofulvus (KST

= 03731 KS = 13027 KT = 20784) and C elegantior(KST = 03927 KS = 20963 KT = 34522) were geneti-cally differentiated while New World and Old Worldmacro-populations of C napus (P = 06840 KST =-00251 KS = 04322 KT = 04216) were not significantlydifferentiated from each other

Network analysesTo clarify the phylogenetic relationship among haplo-types within each species networks were constructedusing haplotypes from ITS sequences with non-recom-bining blocks (Figure 4a-d) Nested clade analyses for Carcuatorum resulted in one network with two connectedOld World haplotypes (H4 and H5) and three singleunconnected New World haplotypes (H1 H2 and H3)(Figure 4a) Analyses of C aureofulvus (Figure 4b)yielded a network comprising five haplotypes haplotypesH3 H4 and H5 were scattered throughout the NewWorld whereas haplotypes H1 and H2 were restrictedto the Old World The haplotypes H1 H2 H4 and H5were derived from haplotype H3 Similarly in C elegan-tior haplotypes were contained in a single networkwhere haplotype H7 has an ancestral position to theother haplotypes and the Old World haplotypes H1 H2H3 and H4 grouped together (Figure 4c) The OldWorld haplotypes H1 H2 H3 and H4 appear to bemore related to the Pacific haplotypes H10 (Oregon andWashington State) and H11 (Washington State) In Cnapus haplotype H1 is ancestral to the other haplotypes

and was shared by both New World and Old Worldpopulations whereas the haplotypes H2 and H3 wererestricted to Mountain USA (Wyoming) and Pacific(Washington State) in the New World respectively (Fig-ure 4d)

Migration analysesThe four species showed different histories of recombi-nation two distinct recombination blocks were detectedin C arcuatorum four in C aureofulvus nine in C ele-gantior and one in C napus respectively There was noevidence of recombination within each of these blocksMigration parameter estimates including populationmutation rate time of divergence and direction ofmigration estimates between New World and OldWorld macro-population samples are shown in Addi-tional File 2 These analyses showed that migrationsbetween New World and Old World macro-populationsamples within C arcuatorum and C aureofulvus wereapproximately equal in both directions and that thesehave occurred at low rates While gene flow estimates inC elegantior indicate that migration appeared to beslightly higher (with overlapping confidence intervals)from the New World to the Old World (m1 = 034)rather than vice versa (m2 = 016)

Genealogical analysesCoalescent genealogies for C arcuatorum and C elegan-tior showed that two ancestral lineages gave rise to theextant New and Old World haplotypes (Figures 5 and6) Genealogies inferred for C aureofulvus (Figure 7)and C napus (data not shown) did not have enough

C arcuatorum

C aureofulvus

C elegantior

C napus

Position 11111111222224444444445555555 4555 113334444444 1

345588900146778001230355677881233345 7025 6197780112599 19

5272816645120131129602508634040112762 9774 7115616167823 59

Site Number 1111111111222222222233333333 1234 1111 12

1234567890123456789012345678901234567 1234567890123

Consensus GATATTGGCCACCCAAAACATCTCGATTCCGGCTCTC TCCT CCATTTAGTTGTG GT

Site Type tvvttttttttttttttvtvttttvtvvvttvttvtt tttv vtvtvvvtttttt tt

Character Type

ii---i-----i---imdash-iiiii-i-mdashi-----i-i- i--- i-imdash-i-i-i-i- --

H1 (3) GCGT H1 (4) CA H1 (12) ATA H1 (23)

H2 (2) GATGGGTCT H2 (3) C H2 (2) ATGA H2 (3) A

H3 (8) CCATTTTTGAA H3 (6) H3 (1) ATAC H3 (1) C

H4 (12) ATTGTTCTCTACC H4 (4) T H4 (1) ATTA

H5 (1) ATTGGTTCTCTACC H5 (1) T H5 (1) GCA

H6 (13) GC

H7 (36) C

H8 (8) CC

H9 (4) CC

H10 (4) C

H11 (1) CA

Figure 3 Distribution of polymorphic sites after removing indels and recombination blocks in the ITS haplotypes of calochroidspecies t transitions v transversions i phylogenetically informative sites - uninformative sites

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 6 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

allopatric distributions sufficiently distinct from oneanother within and between continents to merit a spe-cific designation To address these questions we com-pared DNA sequence variation of species atcontinental and intercontinental scales We also evalu-ated the phenotypical and ecological features of theCalochroi species Special emphasis was placed ondetecting highly labile traits and their patterns of geo-graphic variation among population samples repre-sented by individual collection containing one toseveral basidiomata from one group or cluster foundgrowing in a relatively small patch (genet)Our findings show that the comparison of population

samples using phenotypical ecological and genetic char-acteristics help to clarify questions concerning the his-torical events involved in the modern distributions ofCalochroi species as well as aspects related to host pre-ferences and host-switching and the use of a combinedphylogenetic morphological and ecological species defi-nition We also discuss the utility and limitations of ITSsequences and phenotypic characteristics as identifica-tion tools for Cortinarius species with wide geographicaldistributions

ResultsDNA amplification phylogenetic placement andrelationshipMore than 100 fungus population samples includingour own collections and those from other herbariawere analysed in this study (see the Methods sectionfor more details) We were unable to amplify the ITSregion for many collections because they were eitherpoorly conserved or too old and the sequence qualitywas too low or we amplified contaminant fungi thatprobably colonised the dried basidiomata Analysis ofITS sequences supported the phylogenetic placementof a total of 83 population samples identified asbelonging to C arcuatorum C aureofulvus C elegan-tior and C napus lineages (data not shown) Each spe-cies was recovered as a monophyletic clade with highbootstrap support (Figure 2) The internal phylogenyof C arcuatorum is composed of four major lineagesa clade including population samples from Costa Rica(I) a clade comprising population samples from Men-docino (Pacific USA) (II) a clade composed of samplesfrom Del Norte Co (Pacific USA) (III) and a largeclade containing mostly European and two samplepopulations from Wyoming (IV) respectively WithinC elegantior several population samples had unre-solved positions (V) a portion of North Americanpopulation samples are distributed into one well-sup-ported subclade (VI) and European population samplesformed a well-supported subclade (VII) The NorthAmerican taxa C albobrunnoides C albobrunnoides

var violaceus and C subpurpureophyllus var sulphur-eovelatus and the European taxon C napus formed amonophyletic group with not internal subclades NewWorld and Old World population samples of C aureo-fulvus clustered together with no significant geneticdifferentiation between them

Patterns of nucleotide variation in the ITS regionSequence divergence of the complete ITS regionbetween intercontinental disjunct populations rangedfrom identical to 463 in C arcuatorum from 016 to035 in C aureofulvus from 068 to 146 in C ele-gantior and from identical to 017 in C napus Afterremoving indels and recombination blocks twenty-fourdistinct ITS haplotypes were identified among the 83population samples from the four species analysed 5haplotypes in C arcuatorum and C aureofulvus 11 inC elegantior and 3 in C napus (Figure 3) Nucleotidediversity (π) estimates for the whole population samplewere 0024 plusmn 19 times 10-3 in C arcuatorum 0002 plusmn 30 times10-4 in C aureofulvus 42 times 10-3 plusmn 42 times 10-4 in C ele-gantior and 50 times 10-4 plusmn 20 times 10-4 in C napus (Table1) When species were partitioned into sampled areaspopulation samples showed similar or lower estimates ofnucleotide diversity The average per-nucleotideexpected heterogeneity θw for the whole sample ofeach species was identical or higher than partitionedsampled areas

Tests of neutrality and population subdivisionThe neutrality tests performed for the whole-speciesdata sampling had non-significant values therefore theequilibrium model of neutral evolution could not berejected Only in C arcuatorum were significant valuesdetected for Fu and Lirsquos D and Fu and Lirsquos F testssuggesting background selection for the whole sample(Table 1) The geographical distributions and frequencyof the haplotypes are given in Additional File 1 Withregard to C arcuatorum haplotype H1 (I) was onlyfound in those samples obtained from Costa Rica (Jardiacutende Dota and Parque Prusia) while haplotypes H2 (II)and H3 (III) are specific to California (Mendocino andDel Norte Counties respectively) The most frequenthaplotype (H4 IV 12 population samples) is distributedthroughout the Old World (Germany France ItalySweden) and also occurs in the New World (Wyoming)see population samples IB19870107 and IB19870239The single haplotype H5 (IV) is restricted to the OldWorld In C aureofulvus haplotypes H1 and H2 arerestricted to the Old World (Austria France GermanySweden) whereas haplotypes H4 and H5 were found inthe New World (Washington State) population samplesHaplotype H3 was found in population samples fromthe Pacific Northwest (Washington State) and Mountain

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 4 of 19

USA (Colorado and Wyoming) In C elegantior H7 (V)and haplotypes H1 (VII) were predominant in the NewWorld (Wyoming) and the Old World (Austria FranceGermany and Switzerland) respectively Haplotypes H2H3 and H4 are restricted to Old World and haplotypesH5 H6 (VI) H8 H9 H10 and H11 were observed onlyin those population samples obtained from Old WorldWith regard to C napus the predominant haplotype H1was shared by different taxa in the New World (PacificOregon and Washington State and Mountain USAWyoming and Colorado) and the Old World (GermanyNorway Sweden) Haplotype H2 was found in theWyoming populations and haplotype H3 is representedby a single population sample (JFA 10070) from thePacific (Washington State)Hudsonrsquos tests were carried out to estimate population

genetic structures within and among population samples

from each area (Table 2) Analyses of populationsshowed significant differentiation between Costa RicaPacific USA (California) and Mountain USA (Wyom-ing)Europe population samples of C arcuatorum (P lt0001 KST = 06540 KS = 50769 KT = 146769)between EuropePacific USA (Washington State Ore-gon) and Mountain USA (Wyoming Colorado) popula-tion samples of C elegantior (P lt 0001 KST = 00833KS = 31645 KT = 34522) and between EuropePacificUSA (Washington State) and Mountain USA (WyomingColorado) population samples of C aureofulvus (P =0001 KST = 03183 KS = 14166 KT = 20784) Non-significant genetic differentiation between EuropePacific(Washington State Oregon) and Mountain USA(Wyoming Colorado) was found for C napus popula-tion samples (P = 02240 KST = 00518 KS = 03998 KT

= 04216) Alternatively significant P values (P lt 0001)

002

GU363495 C aureofulvus IB19930612

AY174851 C elegantior TUB 011394

EU0567051 C aureofulvus JFA 10065

GU363468 C elegantior IB19890189

DQ663212 C arcuatorum TSJ2000-083

EU057002 C arcuatorum JFA 11893

GU363487 C albobrunnoides IB19970162

GU363457 C arcuatorum TUB 019279

GU363493 C albobrunnoides IB19890479

GU363458 C arcuatorum IB19870239

GU363479 C napus TUB 019281

EU0567061 C aureofulvus JFA 12428

EU056997 C elegantior JFA 11693

GU363497 C aureofulvus IB19890428

AY174850 C elegantior TUB 011388

DQ663228 C aureofulvus AB01-09-31

DQ663292 C elegantior AB00-09-120

UDB0002235 C elegantior AT2005138

AY033120 C arcuatorum IB19980286

GU363459 C arcuatorum IB19870107

GU363494 C aureofulvus IB19850209

GU363477 C elegantior JFA 13226 GU363476 C elegantior TUB 019280

UDB000721 C elegantior AT2004126

GU363482 C albobrunnoides IB19870275

GU363490 C albobrunnoides IB9870186

GU363465 C elegantior IB19890226

EU057001 C arcuatorum JFA 12061

EU056998 C elegantior JFA 11452

EU056999 C elegantior JFA 12438

GU363492 C albobrunnoides IB19890242

GU363453 C arcuatorum IB19950564

DQ663229 C aureofulvus TF2004-065

GU363464 C elegantior IB19980248

GU363485 C albobrunnoides IB19910261

GU363496 C aureofulvus IB19870221

GU363472 C elegantior IB19970107a

GU363480 C napus TUB 019282

EU057003 C arcuatorum JFA 11803

GU363452 C arcuatorum IB19950686

GU363456 C arcuatorum TUB 019278

AY669571 C aureofulvus TUB 011831

GU363460 C elegantior IB19980414

AY174822 C arcuatorum TUB 011421

GU363489 C albobrunnoides IB19970303

EU057069 C napus S F44393

UDB000738 C elegantior AT2004277

GU363486 C albobrunnoides IB19910270

GU363470 C elegantior IB19910141

GU363469 C elegantior IB19970249

EU057005 C arcuatorum JFA 11765

DQ663341 C napus TSJ2001-003

GU363483 C albobrunnoides IB19890298

EU057016 C subpurpureophyllus var sulphureovelatus JFA 11723

EU057067 C napus TUB 012717

GU363466 C elegantior IB19970300

EU057015 C albobrunnoides var violaceus JFA 10070

EF014262 C elegantior TUB 012709

GU363463 C elegantior IB20040204

GU363455 C arcuatorum TUB 019283

GU363481C albobrunnoides JFA 12426

GU363473 C elegantior var americanus IB19890059

AY174824 C arcuatorum TUB 011403

GU363454 C arcuatorum IB19950596

AY174823 C arcuatorum TUB 011447

GU363471 C elegantior JFA 11411

GU363484 C albobrunnoides IB19910237

EU057004 C arcuatorum JFA 11766

AJ236067 C napus KH3

GU363474 C elegantior IB19910140

DQ663213 C arcuatorum TSJ2004-043

UDB002242 C napus AT2005156

GU363467 C elegantior IB19870057

GU363461 C elegantior IB20010192

GU363491 C albobrunnoides IB1997303b

EU057060 C arcuatorum JFA 12037

GU363462 C elegantior IB19790599

EU057000 C arcuatorum JFA 12039

DQ663211 C fulvoincarnatus PML5208

GU363478 C elegantior JFA 13287

GU363475 C elegantior IB1989226

GU363488 C albobrunnoides IB19970194

UDB002241 C napus AT2005152

100

100

100

100

100

100

100

98

91

92

99

74

100

Old World

New World

Europe

Mountain USAPacific USACosta Rica

IV

I

II

III

V

VI

VII

Figure 2 Phylogenetic relationships of New World and Old World population samples of Calochroi species based on ML analysis of ITSrDNA sequences The represented tree was computed from 1000 runs and was mid-point rooted Numbers above branches are bootstrapvalues (values lt 70 not shown) from 1000 replicates Assignations I to VII indicate major groups of C arcuatorum and C elegantior discussed inthe text Type specimens are printed in bold

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 5 of 19

for Hudsonrsquos tests indicated that New World and OldWold population samples of C arcuatorum (KST =04619 KS = 78968 KT = 146769) C aureofulvus (KST

= 03731 KS = 13027 KT = 20784) and C elegantior(KST = 03927 KS = 20963 KT = 34522) were geneti-cally differentiated while New World and Old Worldmacro-populations of C napus (P = 06840 KST =-00251 KS = 04322 KT = 04216) were not significantlydifferentiated from each other

Network analysesTo clarify the phylogenetic relationship among haplo-types within each species networks were constructedusing haplotypes from ITS sequences with non-recom-bining blocks (Figure 4a-d) Nested clade analyses for Carcuatorum resulted in one network with two connectedOld World haplotypes (H4 and H5) and three singleunconnected New World haplotypes (H1 H2 and H3)(Figure 4a) Analyses of C aureofulvus (Figure 4b)yielded a network comprising five haplotypes haplotypesH3 H4 and H5 were scattered throughout the NewWorld whereas haplotypes H1 and H2 were restrictedto the Old World The haplotypes H1 H2 H4 and H5were derived from haplotype H3 Similarly in C elegan-tior haplotypes were contained in a single networkwhere haplotype H7 has an ancestral position to theother haplotypes and the Old World haplotypes H1 H2H3 and H4 grouped together (Figure 4c) The OldWorld haplotypes H1 H2 H3 and H4 appear to bemore related to the Pacific haplotypes H10 (Oregon andWashington State) and H11 (Washington State) In Cnapus haplotype H1 is ancestral to the other haplotypes

and was shared by both New World and Old Worldpopulations whereas the haplotypes H2 and H3 wererestricted to Mountain USA (Wyoming) and Pacific(Washington State) in the New World respectively (Fig-ure 4d)

Migration analysesThe four species showed different histories of recombi-nation two distinct recombination blocks were detectedin C arcuatorum four in C aureofulvus nine in C ele-gantior and one in C napus respectively There was noevidence of recombination within each of these blocksMigration parameter estimates including populationmutation rate time of divergence and direction ofmigration estimates between New World and OldWorld macro-population samples are shown in Addi-tional File 2 These analyses showed that migrationsbetween New World and Old World macro-populationsamples within C arcuatorum and C aureofulvus wereapproximately equal in both directions and that thesehave occurred at low rates While gene flow estimates inC elegantior indicate that migration appeared to beslightly higher (with overlapping confidence intervals)from the New World to the Old World (m1 = 034)rather than vice versa (m2 = 016)

Genealogical analysesCoalescent genealogies for C arcuatorum and C elegan-tior showed that two ancestral lineages gave rise to theextant New and Old World haplotypes (Figures 5 and6) Genealogies inferred for C aureofulvus (Figure 7)and C napus (data not shown) did not have enough

C arcuatorum

C aureofulvus

C elegantior

C napus

Position 11111111222224444444445555555 4555 113334444444 1

345588900146778001230355677881233345 7025 6197780112599 19

5272816645120131129602508634040112762 9774 7115616167823 59

Site Number 1111111111222222222233333333 1234 1111 12

1234567890123456789012345678901234567 1234567890123

Consensus GATATTGGCCACCCAAAACATCTCGATTCCGGCTCTC TCCT CCATTTAGTTGTG GT

Site Type tvvttttttttttttttvtvttttvtvvvttvttvtt tttv vtvtvvvtttttt tt

Character Type

ii---i-----i---imdash-iiiii-i-mdashi-----i-i- i--- i-imdash-i-i-i-i- --

H1 (3) GCGT H1 (4) CA H1 (12) ATA H1 (23)

H2 (2) GATGGGTCT H2 (3) C H2 (2) ATGA H2 (3) A

H3 (8) CCATTTTTGAA H3 (6) H3 (1) ATAC H3 (1) C

H4 (12) ATTGTTCTCTACC H4 (4) T H4 (1) ATTA

H5 (1) ATTGGTTCTCTACC H5 (1) T H5 (1) GCA

H6 (13) GC

H7 (36) C

H8 (8) CC

H9 (4) CC

H10 (4) C

H11 (1) CA

Figure 3 Distribution of polymorphic sites after removing indels and recombination blocks in the ITS haplotypes of calochroidspecies t transitions v transversions i phylogenetically informative sites - uninformative sites

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 6 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

USA (Colorado and Wyoming) In C elegantior H7 (V)and haplotypes H1 (VII) were predominant in the NewWorld (Wyoming) and the Old World (Austria FranceGermany and Switzerland) respectively Haplotypes H2H3 and H4 are restricted to Old World and haplotypesH5 H6 (VI) H8 H9 H10 and H11 were observed onlyin those population samples obtained from Old WorldWith regard to C napus the predominant haplotype H1was shared by different taxa in the New World (PacificOregon and Washington State and Mountain USAWyoming and Colorado) and the Old World (GermanyNorway Sweden) Haplotype H2 was found in theWyoming populations and haplotype H3 is representedby a single population sample (JFA 10070) from thePacific (Washington State)Hudsonrsquos tests were carried out to estimate population

genetic structures within and among population samples

from each area (Table 2) Analyses of populationsshowed significant differentiation between Costa RicaPacific USA (California) and Mountain USA (Wyom-ing)Europe population samples of C arcuatorum (P lt0001 KST = 06540 KS = 50769 KT = 146769)between EuropePacific USA (Washington State Ore-gon) and Mountain USA (Wyoming Colorado) popula-tion samples of C elegantior (P lt 0001 KST = 00833KS = 31645 KT = 34522) and between EuropePacificUSA (Washington State) and Mountain USA (WyomingColorado) population samples of C aureofulvus (P =0001 KST = 03183 KS = 14166 KT = 20784) Non-significant genetic differentiation between EuropePacific(Washington State Oregon) and Mountain USA(Wyoming Colorado) was found for C napus popula-tion samples (P = 02240 KST = 00518 KS = 03998 KT

= 04216) Alternatively significant P values (P lt 0001)

002

GU363495 C aureofulvus IB19930612

AY174851 C elegantior TUB 011394

EU0567051 C aureofulvus JFA 10065

GU363468 C elegantior IB19890189

DQ663212 C arcuatorum TSJ2000-083

EU057002 C arcuatorum JFA 11893

GU363487 C albobrunnoides IB19970162

GU363457 C arcuatorum TUB 019279

GU363493 C albobrunnoides IB19890479

GU363458 C arcuatorum IB19870239

GU363479 C napus TUB 019281

EU0567061 C aureofulvus JFA 12428

EU056997 C elegantior JFA 11693

GU363497 C aureofulvus IB19890428

AY174850 C elegantior TUB 011388

DQ663228 C aureofulvus AB01-09-31

DQ663292 C elegantior AB00-09-120

UDB0002235 C elegantior AT2005138

AY033120 C arcuatorum IB19980286

GU363459 C arcuatorum IB19870107

GU363494 C aureofulvus IB19850209

GU363477 C elegantior JFA 13226 GU363476 C elegantior TUB 019280

UDB000721 C elegantior AT2004126

GU363482 C albobrunnoides IB19870275

GU363490 C albobrunnoides IB9870186

GU363465 C elegantior IB19890226

EU057001 C arcuatorum JFA 12061

EU056998 C elegantior JFA 11452

EU056999 C elegantior JFA 12438

GU363492 C albobrunnoides IB19890242

GU363453 C arcuatorum IB19950564

DQ663229 C aureofulvus TF2004-065

GU363464 C elegantior IB19980248

GU363485 C albobrunnoides IB19910261

GU363496 C aureofulvus IB19870221

GU363472 C elegantior IB19970107a

GU363480 C napus TUB 019282

EU057003 C arcuatorum JFA 11803

GU363452 C arcuatorum IB19950686

GU363456 C arcuatorum TUB 019278

AY669571 C aureofulvus TUB 011831

GU363460 C elegantior IB19980414

AY174822 C arcuatorum TUB 011421

GU363489 C albobrunnoides IB19970303

EU057069 C napus S F44393

UDB000738 C elegantior AT2004277

GU363486 C albobrunnoides IB19910270

GU363470 C elegantior IB19910141

GU363469 C elegantior IB19970249

EU057005 C arcuatorum JFA 11765

DQ663341 C napus TSJ2001-003

GU363483 C albobrunnoides IB19890298

EU057016 C subpurpureophyllus var sulphureovelatus JFA 11723

EU057067 C napus TUB 012717

GU363466 C elegantior IB19970300

EU057015 C albobrunnoides var violaceus JFA 10070

EF014262 C elegantior TUB 012709

GU363463 C elegantior IB20040204

GU363455 C arcuatorum TUB 019283

GU363481C albobrunnoides JFA 12426

GU363473 C elegantior var americanus IB19890059

AY174824 C arcuatorum TUB 011403

GU363454 C arcuatorum IB19950596

AY174823 C arcuatorum TUB 011447

GU363471 C elegantior JFA 11411

GU363484 C albobrunnoides IB19910237

EU057004 C arcuatorum JFA 11766

AJ236067 C napus KH3

GU363474 C elegantior IB19910140

DQ663213 C arcuatorum TSJ2004-043

UDB002242 C napus AT2005156

GU363467 C elegantior IB19870057

GU363461 C elegantior IB20010192

GU363491 C albobrunnoides IB1997303b

EU057060 C arcuatorum JFA 12037

GU363462 C elegantior IB19790599

EU057000 C arcuatorum JFA 12039

DQ663211 C fulvoincarnatus PML5208

GU363478 C elegantior JFA 13287

GU363475 C elegantior IB1989226

GU363488 C albobrunnoides IB19970194

UDB002241 C napus AT2005152

100

100

100

100

100

100

100

98

91

92

99

74

100

Old World

New World

Europe

Mountain USAPacific USACosta Rica

IV

I

II

III

V

VI

VII

Figure 2 Phylogenetic relationships of New World and Old World population samples of Calochroi species based on ML analysis of ITSrDNA sequences The represented tree was computed from 1000 runs and was mid-point rooted Numbers above branches are bootstrapvalues (values lt 70 not shown) from 1000 replicates Assignations I to VII indicate major groups of C arcuatorum and C elegantior discussed inthe text Type specimens are printed in bold

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 5 of 19

for Hudsonrsquos tests indicated that New World and OldWold population samples of C arcuatorum (KST =04619 KS = 78968 KT = 146769) C aureofulvus (KST

= 03731 KS = 13027 KT = 20784) and C elegantior(KST = 03927 KS = 20963 KT = 34522) were geneti-cally differentiated while New World and Old Worldmacro-populations of C napus (P = 06840 KST =-00251 KS = 04322 KT = 04216) were not significantlydifferentiated from each other

Network analysesTo clarify the phylogenetic relationship among haplo-types within each species networks were constructedusing haplotypes from ITS sequences with non-recom-bining blocks (Figure 4a-d) Nested clade analyses for Carcuatorum resulted in one network with two connectedOld World haplotypes (H4 and H5) and three singleunconnected New World haplotypes (H1 H2 and H3)(Figure 4a) Analyses of C aureofulvus (Figure 4b)yielded a network comprising five haplotypes haplotypesH3 H4 and H5 were scattered throughout the NewWorld whereas haplotypes H1 and H2 were restrictedto the Old World The haplotypes H1 H2 H4 and H5were derived from haplotype H3 Similarly in C elegan-tior haplotypes were contained in a single networkwhere haplotype H7 has an ancestral position to theother haplotypes and the Old World haplotypes H1 H2H3 and H4 grouped together (Figure 4c) The OldWorld haplotypes H1 H2 H3 and H4 appear to bemore related to the Pacific haplotypes H10 (Oregon andWashington State) and H11 (Washington State) In Cnapus haplotype H1 is ancestral to the other haplotypes

and was shared by both New World and Old Worldpopulations whereas the haplotypes H2 and H3 wererestricted to Mountain USA (Wyoming) and Pacific(Washington State) in the New World respectively (Fig-ure 4d)

Migration analysesThe four species showed different histories of recombi-nation two distinct recombination blocks were detectedin C arcuatorum four in C aureofulvus nine in C ele-gantior and one in C napus respectively There was noevidence of recombination within each of these blocksMigration parameter estimates including populationmutation rate time of divergence and direction ofmigration estimates between New World and OldWorld macro-population samples are shown in Addi-tional File 2 These analyses showed that migrationsbetween New World and Old World macro-populationsamples within C arcuatorum and C aureofulvus wereapproximately equal in both directions and that thesehave occurred at low rates While gene flow estimates inC elegantior indicate that migration appeared to beslightly higher (with overlapping confidence intervals)from the New World to the Old World (m1 = 034)rather than vice versa (m2 = 016)

Genealogical analysesCoalescent genealogies for C arcuatorum and C elegan-tior showed that two ancestral lineages gave rise to theextant New and Old World haplotypes (Figures 5 and6) Genealogies inferred for C aureofulvus (Figure 7)and C napus (data not shown) did not have enough

C arcuatorum

C aureofulvus

C elegantior

C napus

Position 11111111222224444444445555555 4555 113334444444 1

345588900146778001230355677881233345 7025 6197780112599 19

5272816645120131129602508634040112762 9774 7115616167823 59

Site Number 1111111111222222222233333333 1234 1111 12

1234567890123456789012345678901234567 1234567890123

Consensus GATATTGGCCACCCAAAACATCTCGATTCCGGCTCTC TCCT CCATTTAGTTGTG GT

Site Type tvvttttttttttttttvtvttttvtvvvttvttvtt tttv vtvtvvvtttttt tt

Character Type

ii---i-----i---imdash-iiiii-i-mdashi-----i-i- i--- i-imdash-i-i-i-i- --

H1 (3) GCGT H1 (4) CA H1 (12) ATA H1 (23)

H2 (2) GATGGGTCT H2 (3) C H2 (2) ATGA H2 (3) A

H3 (8) CCATTTTTGAA H3 (6) H3 (1) ATAC H3 (1) C

H4 (12) ATTGTTCTCTACC H4 (4) T H4 (1) ATTA

H5 (1) ATTGGTTCTCTACC H5 (1) T H5 (1) GCA

H6 (13) GC

H7 (36) C

H8 (8) CC

H9 (4) CC

H10 (4) C

H11 (1) CA

Figure 3 Distribution of polymorphic sites after removing indels and recombination blocks in the ITS haplotypes of calochroidspecies t transitions v transversions i phylogenetically informative sites - uninformative sites

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 6 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

for Hudsonrsquos tests indicated that New World and OldWold population samples of C arcuatorum (KST =04619 KS = 78968 KT = 146769) C aureofulvus (KST

= 03731 KS = 13027 KT = 20784) and C elegantior(KST = 03927 KS = 20963 KT = 34522) were geneti-cally differentiated while New World and Old Worldmacro-populations of C napus (P = 06840 KST =-00251 KS = 04322 KT = 04216) were not significantlydifferentiated from each other

Network analysesTo clarify the phylogenetic relationship among haplo-types within each species networks were constructedusing haplotypes from ITS sequences with non-recom-bining blocks (Figure 4a-d) Nested clade analyses for Carcuatorum resulted in one network with two connectedOld World haplotypes (H4 and H5) and three singleunconnected New World haplotypes (H1 H2 and H3)(Figure 4a) Analyses of C aureofulvus (Figure 4b)yielded a network comprising five haplotypes haplotypesH3 H4 and H5 were scattered throughout the NewWorld whereas haplotypes H1 and H2 were restrictedto the Old World The haplotypes H1 H2 H4 and H5were derived from haplotype H3 Similarly in C elegan-tior haplotypes were contained in a single networkwhere haplotype H7 has an ancestral position to theother haplotypes and the Old World haplotypes H1 H2H3 and H4 grouped together (Figure 4c) The OldWorld haplotypes H1 H2 H3 and H4 appear to bemore related to the Pacific haplotypes H10 (Oregon andWashington State) and H11 (Washington State) In Cnapus haplotype H1 is ancestral to the other haplotypes

and was shared by both New World and Old Worldpopulations whereas the haplotypes H2 and H3 wererestricted to Mountain USA (Wyoming) and Pacific(Washington State) in the New World respectively (Fig-ure 4d)

Migration analysesThe four species showed different histories of recombi-nation two distinct recombination blocks were detectedin C arcuatorum four in C aureofulvus nine in C ele-gantior and one in C napus respectively There was noevidence of recombination within each of these blocksMigration parameter estimates including populationmutation rate time of divergence and direction ofmigration estimates between New World and OldWorld macro-population samples are shown in Addi-tional File 2 These analyses showed that migrationsbetween New World and Old World macro-populationsamples within C arcuatorum and C aureofulvus wereapproximately equal in both directions and that thesehave occurred at low rates While gene flow estimates inC elegantior indicate that migration appeared to beslightly higher (with overlapping confidence intervals)from the New World to the Old World (m1 = 034)rather than vice versa (m2 = 016)

Genealogical analysesCoalescent genealogies for C arcuatorum and C elegan-tior showed that two ancestral lineages gave rise to theextant New and Old World haplotypes (Figures 5 and6) Genealogies inferred for C aureofulvus (Figure 7)and C napus (data not shown) did not have enough

C arcuatorum

C aureofulvus

C elegantior

C napus

Position 11111111222224444444445555555 4555 113334444444 1

345588900146778001230355677881233345 7025 6197780112599 19

5272816645120131129602508634040112762 9774 7115616167823 59

Site Number 1111111111222222222233333333 1234 1111 12

1234567890123456789012345678901234567 1234567890123

Consensus GATATTGGCCACCCAAAACATCTCGATTCCGGCTCTC TCCT CCATTTAGTTGTG GT

Site Type tvvttttttttttttttvtvttttvtvvvttvttvtt tttv vtvtvvvtttttt tt

Character Type

ii---i-----i---imdash-iiiii-i-mdashi-----i-i- i--- i-imdash-i-i-i-i- --

H1 (3) GCGT H1 (4) CA H1 (12) ATA H1 (23)

H2 (2) GATGGGTCT H2 (3) C H2 (2) ATGA H2 (3) A

H3 (8) CCATTTTTGAA H3 (6) H3 (1) ATAC H3 (1) C

H4 (12) ATTGTTCTCTACC H4 (4) T H4 (1) ATTA

H5 (1) ATTGGTTCTCTACC H5 (1) T H5 (1) GCA

H6 (13) GC

H7 (36) C

H8 (8) CC

H9 (4) CC

H10 (4) C

H11 (1) CA

Figure 3 Distribution of polymorphic sites after removing indels and recombination blocks in the ITS haplotypes of calochroidspecies t transitions v transversions i phylogenetically informative sites - uninformative sites

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 6 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

genetic polymorphisms to resolve transoceanic disjunc-tion events In C arcuatorum a widely distributed (NewWorld (Wyoming) - Old World (Europe)) ancestralmacro-population could be the origin of extant haplo-types In C elegantior an ancestral macro-populationdiverged at 07 into two distinct lineages in the NewWorld whereas Old World haplotypes diverged at 04

respectively The oldest mutations of both New Worldand Old World macro-populations of C arcuatorumhave a mean age of 09 while the C elegantior macro-populations have a mean age of approximately 07 In Carcuatorum extant Old World haplotypes are youngerthan New World haplotypes T lt 02 and T gt 05respectively In C elegantior New World haplotype

Table 1 Population statistics diversity estimates and neutrality tests based on ITS region of calochroid taxa studied

Population statistics Tests of neutrality

Speciespopulation

n s h hd k π(SD)

θw TajimarsquosD

Fu andLirsquos DStatistic

Fu andLirsquos FStatistic

Fursquos FsStatistic

C arcuatorum

Old World 11 1 2 0182 0181 33 times 10-4

(26 times 10-4)61 times 10-4 -1128

(NS)-1289(NS)

-1399(NS)

-0410(NS)

New World 15 36 4 0685 12229 0021 0019 1635(NS)

1502 1502(NS)

9684(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 10 20 2 0356 7111 0013 (56 times 10-3) 0013 0027(NS)

1567 1326(NS)

9432(NS)

Costa Rica 3 0 1 0000 0000 0000 0000 ND ND ND ND

All(l = 559)

26 37 5 0698 13741 0024(19 times 10-3)

0017 1569(NS)

1568 1843 13357(NS)

C aureofulvus

Old World 7 1 2 0571 0571 0001(21 times 10-4)

73 times 10-4 1341(NS)

0953(NS)

1101(NS)

0856(NS)

New World 11 2 3 0618 0690 123 times 10-3

(29 times 10-4)122 times 10-3 0036

(NS)-0033(NS)

-0269(NS)

-0113(NS)

Mountain USA 2 0 1 0000 0000 0000 0000 ND ND ND ND

Pacific USA 9 2 3 0667 0777 0001(32 times 10-4)

0001 0195(NS)

-0221(NS)

-0135(NS)

-0108(NS)

All(l = 560)

18 4 5 0804 1346 0002(30 times 10-4)

0002 0473(NS)

0211(NS)

0324(NS)

-0505(NS)

C elegantior

Old World 16 3 4 0442 04833 9 times 10-4

(36 times 10-4)0002 -1316

(NS)-1122(NS)

-1351(NS)

-1867(NS)

New World 67 7 7 0662 1003 0002(27 times 10-4)

0003 -0791(NS)

-0420(NS)

-0638(NS)

-1614(NS)

Mountain USA 62 4 5 0608 0739 0001(18 times 10-4)

0002 -0288(NS)

-0153(NS)

-0228(NS)

-0753(NS)

Pacific 5 1 2 0400 0400 78 times 10-4

(46 times 10-4)93 times 10-4 -0816

(NS)-0816(NS)

-0771(NS)

0090(NS)

All(l = 516)

83 13 11 0760 2139 42 times 10-3(42 times 10-4)

0005 -0495(NS)

-0782(NS)

-0808(NS)

-1468(NS)

C napus

Old World 8 0 1 0000 0000 0000 0000 ND ND ND ND

New World 19 2 3 0368 0385 68 times 10-4

(25 times 10-4)10 times 10-3 -0777 -0573 -0720 -0725

(NS)

Mountain USA 16 1 2 0325 0325 58 times 10-4

(22 times 10-4)53 times 10-4 0155

(NS)0688(NS)

0627(NS)

0551(NS)

Pacific USA 3 1 2 0667 0666 11 times 10-3

(56 times 10-4)0002 ND ND ND 0201

(NS)

All(l = 564)

27 2 3 0271 0279 50 times 10-4

(20 times 10-4)92 times 10-4 -0977

(NS)-0702(NS)

-0898(NS)

-1088(NS)

l nucleotide sequence length n sample size s number of segregating sites h haplotypes hd haplotype diversity k average number of nucleotide pair wisedifferences π number of nucleotide differences per site (SD = standard deviation of π) θw Wattersonrsquos estimate of θ per site ND not determined because nopolymorphism was found NS non-significant P lt 002

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 7 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

deep divergence was older (T gt 06) than New Worldhaplotypes (T = 04) but extant New World haplotypesare descendents from more recent lineages

Basidioma phenotypical featuresDetailed descriptions of the macro- and microscopicalstructures for each species are given in Additional File3 The most relevant phenotypical features are sum-marised below(i) Basidioma colourThere was little or no detectable variation of basidio-mata colour within C arcuatorum C aureofulvus andC elegantior (see Figures 8 9 and 10) However popula-tion samples of C napus showed considerable variationin colour of both lamellae and veil (Figure 11)(ii) Basidioma sizeSome variation in basidioma size was observed within Carcuatorum where populations from Costa Rica and somefrom Europe presented somewhat slender basidiomata incomparison with other population samples (Figure 8)(iii) Colour reaction of dried specimensA pink to pinkish orange colour reaction with 40KOH on both pileus surface and mycelia at the stipebase characterised the basidiomata of C arcuatorumpopulations Population samples of C napus yielded agreyish to dark red brown colour reaction on thepileus surface whereas the mycelia showed a pale pinkto intense pink colour reaction All population samplesin C aureofulvus had a black reaction on both pileus

surface and mycelia on the stipe bulb while in C ele-gantior both the pileus surface and mycelia turnedwine red (Figure 12)

DiscussionThe markedly high species diversity within the genusCortinarius together with their obligate ectomycorrhizalstatus makes the factors that have driven their specia-tion of particular interest to a broad spectrum of biolo-gists In addition the occurrence of widespread taxawithin Cortinarius raises intriguing questions about theorigins of transoceanic disjunct species and the influ-ence of the historical events that have created contem-porary geographical distributions

Phylogeographical structure and historical demographyof Calochroi speciesWhile it is difficult to determine the origin and migra-tion routes of ectomycorrhizal fungi between NorthAmerica Europe and Asia one can gain some insightinto the question by looking at host migration patternsand assuming co-migration of plant and fungus Geneticdata reveal considerable divergence between populationsacross the core range of distribution of C arcuatorumand C elegantior in the New World (Figure 4a c Table2) Since the divergent ancestral lineages were found inthe New World we postulate this area as the centre oforigin for C arcuatorum and C elegantior and that OldWorld populations are the result of relatively recentdemographic expansion (Figures 4a c 5 and 6) Migra-tion with host plants [28] through Beringia is one of themost plausible hypotheses to explain the disjunct popu-lation distributions between the New World and OldWorld but migrations through the North Atlantic landbridge [29] or via long-distance dispersal [30] though aless likely explanation cannot be completely ruled outGenetic breaks between C arcuatorum populationswithin the New World have probably been triggered bypaleogeographic and paleoclimatic events eg the rise ofthe Sierra Nevada and other mountains followed bydesertification of the Great Basin or other events suchas Pliestocene glaciation [3031] The European popula-tions of C arcuatorum and most of the North Americanpopulations occur mainly with trees of the family Faga-ceae with the exception of the population samples inWyoming where this species occurs with conifersThese known population locations are currently geogra-phically isolated from members of the Fagaceae how-ever it is likely that Picea andor Tsuga were sympatricwith Quercus in the past [32] allowing for the possibilityof a host switch from Quercus to a conifer hostThe coalescent-based analysis suggests that extant

populations diverged from a quite recent commonancestor early in the evolution of C elegantior with little

Table 2 Population subdivision in calochroid taxainferred from Hudsonrsquos test statistics Ks (upper rightmatrix) and Kst (lower left matrix)

Old World Mountain USA Pacific USA Costa Rica

C arcuatorum

Old World 01818 (NS) 37041 01636

Mountain USA -01818 (NS) 76666 00000 (NS)

Pacific USA 07511 04025 68148

Costa Rica 09754 10000 (NS) 03185

C aureofulvus

Old World 05714 13399

Mountain USA 03949 18888 (NS)

Pacific USA 04044 -01052 (NS)

C elegantior

Old World 17739 06607

Mountain USA 04581 19434

Pacific USA 07411 01880

C napus

Old World 03908 (NS) 00952 (NS)

Mountain USA -00081 (NS) 05655 (NS)

Pacific USA 04761 (NS) 00329 (NS)

Significance including incompatible sites and recombination blocks wasevaluated by performing 1000 permutations using SNAP Map NS non-significant 001 le P lt 005 0001 le P lt 001 P lt 0001

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 8 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

gene flow among disjunt populations (Figure 6 Addi-tional File 2) Despite this very little morphologicalchange has evolved between New World and Old Worldlineages Hence the close relatedness of Pacific and Eur-opean haplotypes of C elegantior could indicate anexpansion northward along the western side of themountains through the Bering Land Bridge whichresulted in the extant populations in New World (Figure4c) This evolutionary scenario is compatible with themigration of Tricholoma (matsutake) between westernNorth America and eastern Asia as postulated by Cha-pela amp Garbelotto [8] The phylogeographical patternsobserved in C elegantior support the hypothesis thatthe current distributions of populations are closely asso-ciated with the historical events of their host plantsSimilarly conifers are the primary hosts for populations

of C aureofulvus and for C napus In most parts oftheir ranges these Cortinarius species are sympatric orallopatric although their frequency of occurrence varieswith C aureofulvus being the least commonly encoun-tered C napus being moderately frequent and C ele-gantior populations being the most frequentlyencountered especially at higher elevations with PiceaPopulations of C elegantior and C napus are most fre-quently encountered in subalpine conifer forests how-ever they also occur in more mesic mid-elevations andcoastal sites Picea is currently a main host plant genusfor C aureofulvus C elegantior and C napus in westernNorth America this has probably been the case for anextended period of time with a shift in abundance ofPicea from eastern to western North America severaltimes over the last 21000 years [32] however Abies

197

475

523

550

503

H1 (P F B)

453

377

411

488

487

191

406

371

111

422

401

67

372

198 15

ba

c dH3 (P)

H2 (P)

H2 (P) H4 (P)

H11 (P)

H10 (P)

H8 (P)H6 (P)H9 (P)

H5 (P)

H7 (P M2)

H1 (P)

H2 (P)H3 (P)

H5 (P)

H3 (P)

H4 (P)

H1 (P)

H2 (F P)

H1 (F M1)

H3 (F P)H4 (F P)

H5 (F P)

Figure 4 TCS haplotype networks based on non-recombining ITS sequences a C arcuatorum b C aureofulvus c C elegantior d C napusParsimony probabilities were set at 95 Sizes of circular and rectangular areas are proportional to the number of individuals with thathaplotype Haplotypes with a blue background are from the Old World and those with red backgrounds are from the New World Haplotypescomprising New and Old World populations are indicated by lilac backgrounds Distributions of ecological and morphological features areabbreviated as follows F Fagaceae P Pinaceae (pure stand in Wyoming) P Pinaceae B Betulaceae M1 populations of C arcuatorum withlarger spores M2 some populations of C elegantior with sulphur yellow veil and lacking of veil patches on the pileus surface

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 9 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

Pinus and Tsuga are also potential hosts for these spe-cies Douglas fir another possible host for these speciesappeared in western North America in the mid-Pleisto-cene [33] Recent phylogenetic studies of Picea indicatethat it originated in North America (P breweriana andP sitchensis as basal taxa) and that the present

distribution of Picea could be a result of two dispersalevents one from North America to Asia by the BeringLand Bridge [34] and a second from Asia to EuropeMost of the northeastern Asian species and the Eur-opean P abies could have arisen from a recent radiation

Figure 5 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (CR = Costa Rica Pacific CA =California Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C arcuatorum performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =116 times 10-18) and standard deviation (SD = 953 times 10-16) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Figure 6 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington State OR= Oregon Mountain USA WY = Wyoming) and Old World (EU= Europe) populations of C elegantior performed usingGenetree The inferred genealogy is a rooted genealogy based onfive million simulations of the coalescent Simulations were carriedout assuming a constant population size Maximum likelihoodestimates of the tree with the highest root probability (likelihood =255 times 10-15) and standard deviation (SD = 468 times 10-13) showingthe distribution of mutations The direction of divergence is fromthe top of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 10 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

[35] The literature also notes that the earliest (Paleo-cene) fossil of Picea is from Montana and that it is wellrepresented in the Eocene in western North Americabut was not common in Asia until the Oligocene andnot present in Europe until the Pliocene Therefore it isfully possible that populations within C aureofulvus Celegantior and C napus have long been associated withPicea and could have migrated with this and possibleother host plants during glacial and interglacial periodsfor millions of years The findings presented here inconjunction with those of Dentinger et al [4] and Wuet al [7] strongly support the key role of Beringia in thephylogeographic processes leading to speciation andintraspecific population structures in the Northern

Hemisphere Subsequently the comigration with theirassociated phanerogams [2835] through eg short-dis-tance spore dispersal may be an important meansexplaining the expansion of these fungi from Asia toEuropeWe cannot estimate times for the divergence of dis-

junct populations of Calochroi with certainty owing to alack of fossil records However it is plausible tohypothesise that they co-migrated with their host plantsand that the lineages may have subsequently diverged asa result of fragmentation and geographic isolation ofancestral populations due to geologic and subsequentclimatic changes However each species has a somewhatdifferent biogeographical history that likely has beeninfluenced by different biotic and abiotic factors such asdispersal potential and host (phanerogam) history Theoccurrence of these taxa in Asia has not been documen-ted to date Consequently further sampling across thisregion is necessary in order to better understand theirhistorical distributions across the Northern HemisphereThe recent work on porcini mushrooms [4] supportsthe idea of a potential link between European Asianand North American species in the section Calochroi

ITS sequences and phenotype as identification tools forCortinarius species with wider ranges of geographicaldistribution utility and limitationsNested clade demographic and coalescent-based ana-lyses based on the ITS region of C arcuatorum C aur-eofulvus C elegantior and C napus allowed us to infertheir evolutionary histories Overall C arcuatorum andC elegantior exhibited geographically structured haplo-types with evidence for a more ancient populationexpansion These findings agree with those of Geml etal [911] who observed a similar population structurein the widespread fly agaric Amanita muscaria Interest-ingly geographically structured populations within Carcuatorum and C elegantior were often accompaniedby some divergence in basidioma colouration and size inthe absence of any described species (see Additional File3) By contrast some populations within C arcuatorumand C aureofulvus with disjunct distributions in NorthAmerica and Europe exhibited little morphological dif-ferentiation and had identical to relatively low levels ofITS sequence divergence (identical to 03 (2 bp) diver-gence) Two hypotheses might explain this low degreeof nucleotide variation First these taxa might be rela-tively young having recently spread across their rangesand accordingly having had insufficient time to accu-mulate ITS sequence divergence An alternative hypoth-esis is that these results could be a consequence of lowmutation rates at the locus studied Although some dif-ferences in colouration of the basidiomata were foundbetween western populations of C albobrunnoides var

Figure 7 Coalescent-based genealogy of the internaltranscribed spacers (ITS1 and ITS2) showing the distributionmutations for New World (Pacific WA = Washington StateMountain USA WY = Wyoming CO = Colorado) and Old World(EU = Europe) populations of C aureofulvus performed usingGenetree The inferred genealogy is based on five millionsimulations of the coalescent Simulations were carried outassuming a constant population size Maximum likelihood estimatesof the tree with the highest root probability (likelihood = 175 times 10-8) and standard deviation (SD = 161 times 10-6) showing thedistribution of mutations The direction of divergence is from thetop of the genealogy (oldest = past) to bottom (youngest =present) coalescence is from the bottom (present) to the top (past)The numbers below each branch indicate the correspondinghaplotype and its frequency in each macro-population andsampling area The time scale is in coalescent units of effectivepopulation size

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 11 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

albobrunnoides C albobrunnoides var violaceovelatusand C subpurpureophyllus var sulphureovelatus andEuropean populations of C napus these differenceswere not sufficient to consider them as separate taxaThis suggests that variation in colouration alone mightnot necessarily reflect different species or subspeciesFor example violet pigments in the basidiomata arehighly labile because they are sensitive to solar radiationand the age of the basidiomata Therefore it is evidentthat it is important to do careful fieldwork and to con-sider the variation of macroscopical and ecological fea-tures before they are used for taxonomic purposesespecially characters related to colouration Accordingto our analyses colour reactions with KOH are relativelystable among populations whereas ITS sequences incertain instances provide several genetic markers to dis-tinguish species as well as population structure withinthe section Calochroi

ConclusionsBased on variations in ITS DNA polymorphism thisstudy revealed different evolutionary histories for NewWorld and Old World populations within Cortinariusarcuatorum C aureofulvus C elegantior and Cnapus Nested clade demographic and coalescent-based analyses provide powerful tools for assessing theeffect of historical and contemporary events of thegeographic distributions of these species proposed asfollows (i) an ancestral population of C arcuatorumputatively with a Wyoming-Europe distribution evolvedinto at least three distinct lineages in association with

a variety of fagaceous and coniferous trees in the NewWorld (ii) two divergent lineages in C elegantior gaverise to New World and Old World haplotypes thatoccur almost exclusively in association with Picea(apart from one Washington State population inTsugaAbies forest) (iii) low genetic diversity amongNew World and Old World populations of C aureoful-vus iv) low genetic diversity among New World andOld World populations of C napus The results for Caureofulvus and C napus could be due to recentdemographic expansion but the origin of such transo-ceanic disjunct distributions remains unclear The sce-nario of a spatial expansion through the Bering LandBridge appears to be the most probable explanation formodern transoceanic disjunct distributions however itdoes not represent a complete explanation for theremarkably low degree of genetic differentiationbetween populations of these two species Our studyreveals patterns of species diversification restricted tothe New World within C acuatorum and C elegantiorhowever the relatively few morphological innovationsfound among allopatric populations suggest that diver-sification events may have been driven by ecologicalopportunities afforded by the shift to new host speciesedaphic and climate conditions The morphologicaland genetic data described here support a single spe-cies C napus (to include C albobrunnoides var albo-brunnoides C albobrunnoides var violaceovelatus andC subpurpureophyllus var sulphureovelatus) withtransoceanic distribution Similarly C aureofulvusshows disjunct distributions in New World and Old

Figure 8 Basidioma morphology and colouration within C arcuatorum a Costa Rican population JFA 12039 b Costa Rican population JFA12061 c Californian (Mendocino) population IB19950686 d Californian (Patricks Creek Campground) population JFA 11765 e Californian(Patricks Creek Campground) population IB19950564 f European population TUB 011403 Photos c and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 12 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

World Colour reactions with KOH represent a power-ful tool for recognising these Calochroi specieswhereas features related mainly to veil and lamellaecolouration displayed high levels of homoplasy andphenotypic plasticity Interpretations of the populationstructures suggest that host tree history as influencedby past events (glaciation mountain building andassociated factors) was one of the driving forces thatshaped the modern distributions of Calochroi speciesFinally our work shows the need for further studiesthat focus on careful fieldwork in the context of thevariation of both morphology and ecology as well as

the use of more genetic information to draw morecomplete models of population histories in these andother species of Cortinarius

Taxonomic implicationsSee Additional File 3 for information of the followingtaxonomic noveltiesCortinarius elegantio-montanus Garnica amp Ammir-

ati (= Cortinarius elegantior (Fr) Fr var americanus MM Moser amp McKnight 1995) new synonym and newstatusCortinarius elegantio-occidentalis Garnica amp

Ammirati new species - allied to Cortinarius elegantior(Fr) Fr 1838Cortinarius fulvo-arcuatorum Garnica amp Ammirati

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius jardinensis Garnica Ammirati amp Halling

new species - allied to Cortinarius arcuatorum RobHenry 1939Cortinarius lilaciotinctus Garnica amp Ammirati new

species - allied to Cortinarius arcuatorum Rob Henry1939Cortinarius napus Fr 1838 (= Cortinarius albobrun-

noides var albobrunnoides MM Moser amp McKnight1995 Cortinarius albobrunnoides MM Moser amp KMcKnight var violaceovelatus MM Moser amp J Ammir-ati 1996 and Cortinarius subpurpureophyllus AHSmith var sulphureovelatus MM Moser 2000) newsynonymThe authors declare that they have no competing

interests

MethodsPopulation samplingWe collected population samples within C arcuatorumRob Henry C aureofulvus MM Moser C elegantior(Fr) Fr var elegantior (including C elegantior (Fr) Frvar americanus MM Moser amp McKnight) and Cnapus Fr (including C albobrunnoides MM Moser ampMcKnight C albobrunnoides var violaceovelatus MMMoser amp Ammirati and C subpurpureophyllus AHSm var sulphureovelatus MM Moser) groups follow-ing their distribution ranges in Europe Central Americaand western North America The following specimenswere initially used for DNA sequencing 1 collection ofC albobrunnoides from the University of WashingtonHerbarium (WTU) and 13 from Innsbruck Herbarium(IB) and 8 collections of C arcuatorum (including twonamed C cf sodagnitus IB19870239 and IB19870107)36 collections of C elegantior (including the type mate-rial of C elegantior var americanus IB19890059) fromEurope and North America deposited at (IB) and 4 fromWTU 14 collections of C aureofulvus and 7 collections

Figure 9 Basidioma morphology and colouration of Caureofulvus a Washington State population IB19890428 bColorado population JFA 12428 c Swedish population IB19850209Photos a and c courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 13 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

Figure 10 Basidioma morphology and colouration within C elegantior a Washington State (Table Mountain) population JFA 13287 bWashington State (Table Mountain) population JFA 13287 a basidioma showing KOH reaction c Washington (Snohomish County) populationTUB 019280 d Wyoming (Teton National Forest) population of C elegantior var americanus IB19890059 (holotype) e Wyoming (Teton NationalForest) population IB19890189 NH3 is showed on stipe context f Austrian population IB19790599 (one basidioma shows a red wine KOHreaction) Photos d e and f courtesy of M Moser

Figure 11 Basidioma morphology and colouration within C napus a Washington State (Rainy Pass) population IB19890479 b Washington(Chelan County) population sample of C albobrunnoides var violaceovelatus IB19890186 (typus) c Oregon (Lincoln County) population sample ofC subpurpureophyllus var sulphureovelatus JFA 11723 d Wyoming population (Shoshone National Forest) IB19970303 e Wyoming (ShoshoneNational Forest) population IB19910237 f European population TUB 019282 Photos a b d and e courtesy of M Moser

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 14 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

of C napus from IB An attempt was made to samplefrom throughout the distribution range of these speciesalthough sampling was dictated by the availability ofrecently collected and accurately documented herbariumspecimens These four taxa were selected because theyrepresent a diverse set of speciespopulations each witha monophyletic lineage displaying allopatric distributionsin Europe and North America as previously revealed inresearch by the authors [14] In order to incorporate asmuch genetic information as possible from a wide rangeof geographical locations for these species sequenceswere obtained from GenBank httpwwwncbinlmnihgov and UNITE httpuniteutee databases Details ofthe population samples used for this study and the geo-graphic location of the sampling sites are provided inAdditional File 4

DNA extraction amplification sequencing and sequenceeditingTotal genomic DNA was extracted from dried lamellafragments using the DNAeasy Plant Mini Kit (QiagenHilden Germany) following the standard protocol Theinternal transcribed spacers (ITS1 ITS2) the 58 S ribo-somal subunit and the D1D2 regions of the nucLSU

were amplified with the primer combination ITS1F (5rsquo-CTTGGTCATTTAGAGGAAGTAA-3rsquo) [36]NL4 (5rsquo-GGTCCGTGTTTCAAGACGG-3rsquo) [37] Subsequentlythose specimens that did not give any amplification pro-ducts were amplified with the primer combinationsITS1FITS4 (5rsquo-TCCTCCGCTTATTGATATGC-3rsquo) [38]or ITS1 (5rsquo-TCCGTAGGTGAACCTGCGG -3rsquo) [38]ITS4 and 58SR (5rsquo-TCGATGAAGAACGCAGCG-3rsquo)LR3 (5rsquo-CCGTGTTTCAAGACGGG-3rsquo) [39] or 58SRNL4 Amplifications of ITS1 ITS2 the 58S ribosomalsubunit and the D1D2 regions of the nucLSU were car-ried out using a touch-down program with the followingconditions initial denaturation at 94degC for 3 min 10cycles with temperatures ranging from 60degC in the firstcycle to 51degC each cycle decreasing by 1degC 25 cycleswith an annealing temperature of 50degC each cycle con-sisted of an annealing step of 05 min an elongationstep of 72degC for 1 min and a denaturation step of 94degCfor 05 min and a final elongation phase at 72degC for 7min Alternatively amplifications of the ITS region forthose specimens with negative results using Taq DNApolymerase (Invitrogen Corporation Carlsbad CAUSA) were carried out using 25 μL Phusiontrade High-Fidelity DNA polymerase-mediated reactions (Finnzymes

Figure 12 Colour change reactions with KOH 40 mycelia at the stipe bulb and on pileus surface of dried specimens a - c Carcuatorum a Blood colour on mycelia in IB19870239 b Blood colour on pileus surface in IB19870239 c Pink on pileus surface in TUB 019283d - f C aureofulvus d Black on mycelia in IB19890428 e Purple on pileus surface in IB19930612 f Reaction in (e) turning black after someminutes g - i C elegantior var americanus IB198959 holotype g Wine red on mycelia h Intense wine red on pileus surface i Reaction in (g)after a while becoming dark wine red j Pink with orange tinge on mycelia in C napus TUB 019282 k - l C albobrunnoides k Pink with orangetinge on mycelia in C albobrunnoides IB19890298 l Dark red-brown on pileus surface in C albobrunnoides IB19970303 Note photographs weretaken in natural daylight

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 15 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

Oy Keilaranta Finland) with an initial heating of 95degCfor 60 seconds followed by 35 cycles of denaturation at94degC 30 seconds annealing at 50degC-55degC 30 secondsand extension at 72degC and 45-90 seconds followed by afinal extension of 72degC for 10 minutes Amplificationproducts were electrophoresed in a 07 agarose gel andstained with ethidium bromide for visualization of thebands PCR products were cleaned using ExoSAP-ITreg

(USB Corporation Cleveland OH USA) reagent diluted1 20 according to the manufacturerrsquos instructions BothDNA strands were cycle-sequenced with the amplifica-tion primers and in some cases internal primers asindicated in Garnica et al [40] Clean PCR productswere sequenced in both directions with a 1 6 diluteddye terminator sequencing kit (Big Dye 31 Applied Bio-systems Foster City CA USA) on an ABI Prism 3130 timesl Genetic Analyzer (Applied Biosystems) Unpublishedmolecular data from our lab indicate non-sequencedivergence among individuals within the same collec-tion therefore we sequenced only one individual percollection Forward and reverse sequence fragmentswere assembled and edited using Sequencher version 41(Gene Codes Corporation Ann Arbor MI USA)

Sequence identity alignments phylogenetic placementand variabilityTo detect potential contaminant sequences we firstscreened our sequences against those available in theGenBank database httpwwwncbinlmnihgovhttp[41]) using the Basic Local Alignment Search Tool(BLAST) To establish the phylogenetic placement ofthe newly generated ITS sequences and those from Gar-nica et al [14] we automatically aligned them inMAFFT v57 [42] with the E-INS-i option Our datamatrix which included 637 sequences and 739 nucleo-tide positions was analysed by ML inference as imple-mented in RAxML version 703 [43] The best-knownlikelihood tree under the GTRMIX model of nucleotidesubstitution was computed from 100 runs starting fromdistinct randomised maximum parsimony starting treesA total of 1000 non-parametric bootstrap replicates [44]were run on the original alignment Graphical proces-sing of the trees with best likelihood and bootstrappingwere generated using TreeViewPPC version 166 [45]and PAUP 40b10 [46] Subsequently sequences foreach taxon were aligned separately in MAFFT Withineach taxon set all polymorphic sites were recheckedfrom the chromatograms Isolates from fungal speci-mens are n + n and therefore provide ambiguous haplo-type data for heterozygotes Heterozygous sites appearas two coincident peaks at the same site in the forwardand reverse sequence chromatograms To resolve thehaplotype structure within the heterozygote we usedClarkrsquos haplotype subtraction algorithm [47] The DNA

sequences newly generated in this study were submittedto GenBank (accession numbers GU363452-GU363497)All sequences within one species were collapsed into

haplotypes removing indels and excluding infinite-siteviolations using Map as implemented in SNAP Work-bench [4849] A site-incompatibility matrix was calcu-lated to assess compatibility among all variablecharacters with incompatible characters being subse-quently removed Sites are called compatible if there is aphylogeny which both can evolve on without homoplasyTwo or more variable sites showing an identical patternof compatibility are considered to form a recombinationblock [50] When more than one recombination blockswere found we used RecMin [51] as implemented inSNAP Workbench to estimate the minimum number ofrecombination events Conversely all sites being compa-tible with each other were considered as evidence for norecombination Sequence polymorphism (haplotype hand nucleotide π diversities) was estimated using DnaSPversion 5 [52] Estimates were calculated for the entiredataset comprising all sequences of one species as wellas for each subpopulation sampled

Neutrality tests and population subdivisionIdentical ITS sequences were collapsed into haplotypesusing SNAP Map [49] after excluding indels and infi-nite-site violations Site compatibility matrices were gen-erated from each haplotype datasets using SNAP Cladeand Matrix [53] Sequence variations were tested fordeviations from neutrality by using Tajimarsquos D [54] Fuand Lirsquos D and F [55] and Fursquos Fs [56] statistics withDNASP v 50007 [52] Tests of neutrality assume a con-stant population size no recombination and no migra-tion The D test by Fu and Li is based on assessingselection by counting external (recent) and internal(ancient) mutations in the genealogy and comparing thevalues to their expectations under the assumption ofneutral evolution Significant Fu and Lirsquos tests suggestbackground selection while a significant Fursquos Fs valueindicates population growth and genetic hitchhiking Assmall sample sizes and population subdivisions areknown to limit the power of the neutrality tests [5758]we subsequently tested our data for population subdivi-sion To test for population genetic differentiation weused SNAP Map [49] to generate the appropriatesequence file Septomatrix to convert the sequence fileto a distance matrix and Permtest [57] to test for geo-graphic intraspecific subdivision among the differentareas sampled as implemented in SNAP Workbench[48] Population genetic differentiation analyses werecalculated according to the sampled areas in CentralAmerica (CR = Costa Rica) Pacific USA (WA =Washington State OR = Oregon CA = California)Mountain USA (WY = Wyoming CO = Colorado) and

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 16 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

Europe (EU) In order to increase sample sizes popula-tions were grouped into macro-populations with regardto the directions of migration hypothesized as followsfor C arcuatorum these were Costa Rica (CR) + Pacific(CA) + Mountain USA (WY) for C aureofulvus thesewere Pacific USA (WA) + Mountain USA (WY CO)for C elegantior Pacific (WA OR) + Mountain USA(WY) for C napus Pacific USA (WA OR) + MountainUSA (WY CO) Each macro-population was comparedagainst the EU macro-population The main criteriaused to define the macro-populations were geographicallocation of the sampling sites geological events and hosttree histories Alternatively population samples wereanalysed as follows for C arcuatorum in Central Amer-ica (CR) + Pacific (CA) macro-populations was com-pared against Mountain USA (WY CO) + Europe (EU)macro-populations whereas in C aureofulvus and Celegantior Pacific USA (WA OR) macro-populationswere compared against Mountain USA (WY) + Europe(EU) macro-populations Hudsonrsquos test statistics wereevaluated under the null hypothesis of no genetic differ-entiation applying 1000 permutations per species sam-ple For this purpose only we used the non-reduceddatasets containing the recombination blocks (seeabove) as recombination has been shown to improvethe power of Hudsonrsquos test [57]

Haplotype network analysesWe constructed haplotype networks by using the pro-gram TCS version 121 [59] Parsimony probability wasset at 95 therefore haplotypes with a probability ofparsimony higher than 95 would be connected andthose with a probability lower than 95 would beunlinked

Migration analysisFor those species showing significant population subdi-vision according to Hudsonrsquos test we conducted migra-tion analysis using the IM program [60] As theprogram model assumes neutrality and no recombina-tion the reduced dataset that contained non-recombi-nant characters only was submitted to analysis Thisprogram simultaneously infers multiple genealogicalparameters like effective population sizes for ancestral(θA) and modern (θ1 and θ2) populations time sincedivergence between populations (t) and migration rates(m1 and m2) between two subpopulations using a Mar-kov Chain Monte Carlo (MCMC) approach We usedchain lengths of 10000000 steps for C elegantior data-set and 20000000 steps for C arcuatorum and C aur-eofulvus to guarantee a sufficiently large effective samplesize (ESS) as well as a flat or presumably completeappearance of the posterior distribution curve For eachdataset we executed a total of 15 independent

simulations of the chain (starting from different randomnumber seeds for each) and evaluated the estimatedmeans of migration rates m1 and m2 Subsequently aone-sided t-test was used to statistically assess thehypothesis of directional migration between the twosubpopulations

Genealogical analysesAncestral intraspecific genealogies were inferred usingGenetree Version 9 [61] in SNAP Workbench [48] Thegenealogies with the highest root probability ages ofmutations the times to the most recent common ances-tor (TMRCA) and geographic distribution of the muta-tions were assessed from coalescent simulationsCoalescent simulations were carried out assuming aninfinite-site model constant size and population subdivi-sion As for the previous analyses a reduced dataset wasused that lacked incompatible and recombinant sites tomeet the model assumptions The backward migrationmatrix was set up by averaging the backward migrationrate estimates obtained from the preceding IM simula-tions Genealogical analyses for each species were esti-mated using 4000000 steps per coalescent simulationFor the most likely tree topology two independent runsusing different starting seeds were executed both result-ing in the same topology

Distribution of phenotypic features among populationsTo evaluate the distribution of morphological andmacrochemical features of the basidiomata of the popu-lations used in this study we analysed the macro- andmicroscopical morphological structures in detail as wellas analysing the macrochemical reactions with KOH fol-lowing standard procedures Shape measurement andcolour of the microscopic features were obtained fromdried specimens by mounting them in 3 KOH Mea-surements of basidiospores (n = 51) are given as lengthand breadth where the values in brackets represent theuncommon extremes Basidiospore lengthwidth quoti-ent (Q) mean and standard deviation values werecalculatedIn accordance with article 29 and 30A2 of the Inter-

national Code of Botanical Nomenclature copies of thisarticle are deposited at the following ten botanical andor publicly accessible libraries 1) Departamento deBotaacutenica Instituto de Biologiacutea UNAM Circuito exter-ior sn Cuidad Universitaria Copilco Coyoacaacuten Dis-trito Federal CP 04510 AP 70-233 Mexico 2) MarianKoshland Bioscience and Natural Resources Library2101 Valley Life Sciences Bldg 6500 University ofCalifornia Berkeley CA 94720-6500 USA 3) NaturalSciences Library Box 352900 University of WashingtonSeattle WA 98195 USA 4) Serial and ElectronicResources Washington State University Libraries PC

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 17 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

Box 645610 100 Dairy Rd Pullman WA 99164-5610USA 5) Acquisitions The LuEsther T Mertz LibraryThe New York Botanical Garden 2900 Southern BlvdBronx NY 10458 USA 6) Librarian Canadian ForestService Natural Resources Canada Pacific Forestry Cen-tre Victoria 506 West Burnside Road BC Canada V8Z1M5 7) UBC Herbarium Dept of Botany 3529-6270University Blvd Vancouver B C Canada V6T 1Z4 8)Plant Research Library Wm Saunders Bldg 49 CEFOtawa ON Canada 9) Main Library Herbarium RoyalBotanic Gardens Kew Surrey TW9 3AB UK 10) Her-barium MSB Ludwig-Maximilians-Universitaumlt MuumlnchenMenzinger Straszlige 67 80638 Muumlnchen Germany and11) Herbarium Tubingense (TUB) University of Tuumlbin-gen Auf der Morgenstelle 1 D-74076 TuumlbingenGermany

Additional material

Additional File 1 Identity of haplotypes of calochroid taxa inferredfrom ITS rDNA sequences Haplotype frequencies are shown inparentheses In some cases a single collection can carry two or moredifferent haplotypes

Additional File 2 Population mutation rate effective sample size(ESS) time of divergence and direction of migration estimateswithin C arcuatorum C aureofulvus and C elegantior populationsamples The parameters are as follows θ1 θ2 and θA are the meanpopulation rates for the Old World the New World and the ancestralpopulation respectively t is the mean time of divergence of populationsfrom a common ancestor m1 represents the mean number ofmigrations into the Old World and m2 is the mean number of migrationsinto the New World Numbers in parenthesis are standard deviationsReliable estimation of the highest probability density (HPD) intervalscould not be achieved as the posterior distribution turned out to beincomplete in several cases All values represent arithmetic means takenfrom 15 independent runs per species

Additional File 3 Taxonomy Macroscopical descriptions are based onfresh material whereas microscopical structures were analysed fromdried specimens

Additional File 4 Population samples used in this study and theirrespective voucher numbers GenBank numbers collection datecollection site host tree(s) collector and determinator Holotypespecimens are marked in bold Herbarium abbreviations IB = HerbariumInnsbruck Austria JFA = J F Ammirati material deposited at the BurkeMuseum Herbarium University of Washington Herbarium Seattle USA(WTU) Costa Rican National Biodiversity Institute (INB) and University ofCosta Rica San Joseacute (USJ) TUB = Herbarium Tuumlbingense University ofTuumlbingen S = Herbarium Stockholm Sweden

AcknowledgementsThe authors are grateful to Sabine Silberhorn for technical assistance withlab work and Gunther Saar for providing dried specimens Special thanks toRegine Kuhnert-Finkernagel and Ursula Peintner from IB herbarium(Innsbruck Herbarium Austria) for kindly facilitating use of herbariumspecimens colour slides and Meinhard Moserrsquos original macroscopicaldescriptions We thank Estella Leopold and Caroline Stromberg (University ofWashington) and Cathy Whitlock (Montana State University) for discussionsof plant geography Editor and two anonymous reviewers contributedimportant suggestions and guidance This work was supported by grantsfrom the Deutsche Forschungsgemeinschaft (OB 2427-1 29) and fundingfrom the Universidad de Costa Rica Instituto Nacional de Biodiversidad

(INBio) Costa Rica the New York Botanical Garden (Roy Halling) and the DE Stuntz Memorial Foundation Seattle Washington

Author details1Organismic Botany Institute of Evolution and Ecology University ofTuumlbingen Auf der Morgenstelle 1 D-72076 Tuumlbingen Germany 2Division ofAnimal Genetics University of Tuumlbingen Auf der Morgenstelle 28 D-72076Tuumlbingen Germany 3INRES Gartenbauwissenschaft University of Bonn Aufdem Huumlgel 6 D-53121 Bonn Germany 4Department of Biology Box 351330University of Washington Seattle Washington 98195 USA

Authorsrsquo contributionsSG JA and FO conceived the study design SG generated the DNAsequences and wrote the manuscript PS and SG performed populationgenetics analyses macro- and microscopical descriptions of North Americanpopulations were made by JA macroscopical descriptions of Europeanpopulations were carried out by BO SG studied the European collectionsmicroscopically All authors read and approved the final manuscript

Received 20 October 2010 Accepted 19 July 2011Published 19 July 2011

References1 Kretzer A Li Y Szaro T Bruns TD Internal transcribed spacer sequences

from 38 recognized species of Suillus sensu lato Phylogenetic andtaxonomic implications Mycologia 1996 88776-785

2 Martin F Diacuteez J Dell B Delaruelle C Phylogeography of theectomycorrhizal Pisolithus species as inferred from nuclear ribosomalDNA ITS sequences New Phytol 2002 153345-357

3 den Bakker HC Zuccarello GC Kuyper TW Noordellos ME Phylogeographicpatterns in Leccinum sect Scabra and the status of the arctic-alpinespecies L rotundifoliae Mycol Res 2007 111663-72

4 Dentinger BTM Ammirati JF Both EE Desjardin DE Halling RE Henkel TWMoreau P-A Nagasaw E Soytong K Taylor AF Watling R Moncalvo JMMcLaughlin DJ Molecular phylogenetics of porcini mushrooms (Boletussection Boletus) Mol Phylogenet Evol 2010 571276-92

5 Rochet J Moreau P-A Manzi S Gardes M Comparative phylogenies andhost specialization in the alder ectomycorrhizal fungi Alnicola Alpovaand Lactarius BMC Evol Biol 2011 1140-53

6 Halling R Ectomycorrhizae Co-evolution significance and biogeographyAnn Missouri Bot Gard 2001 885-13

7 Wu QX Mueller GM Lutzoni FM Huang YQ Guo SY Phylogenetic andbiogeographic relationships of Eastern Asian and Eastern NorthAmerican disjunct Suillus species (Fungi) as inferred from nuclearribosomal RNA ITS sequences Mol Phylogenet Evol 2000 1727-37

8 Chapela IH Garbelotto M Phylogeography and evolution in matsutakeand close allies inferred by analyses of ITS sequences and AFLPsMycologia 2004 96730-741

9 Geml J Laursen GA OrsquoNeill K Nusbaum HC Taylor DL Beringian originsand cryptic speciation events in the fly agaric (Amanita muscaria) MolEcol 2006 15225-239

10 Hosaka K Castellano MA Spatafora JW Biogeography of Hysterangiales(Phallomycetidae Basidiomycota) Mycol Res 2008 112449-462

11 Geml J Tulloss RE Laursen GA Sazanova NA Taylor DL Evidence forstrong inter- and intracontinental phylogeographic structure in Amanitamuscaria a wind-dispersed ectomycorrhizal basidiomycete MolPhylogenet Evol 2008 48694-701

12 Halling RE Osmundson TW Neves M-A Pacific boletes Implications forbiogeographic relationships Mycol Res 2008 112437-447

13 Matheny PB Aime MC Bougher NL Buyck B Desjardin DE Horak EKropp BR Lodge DJ Trappe JM Hibbett DS Out of the palaeotropicsHistorical biogeography and diversification of the cosmopolitanmushroom family Inocybaceae J Biogeogr 2009 36577592

14 Garnica S Weiszlig M Oertel B Ammirati J Oberwinkler F Phylogeneticrelationships in Cortinarius section Calochroi inferred from nuclear DNAsequences BMC Evol Biol 2009 91

15 Kauffman CH Cortinarius Fries North American Flora 1932 10282-34816 Smith AH Studies in the genus Cortinarius I Contributions from the

University of Michigan Herbarium 1939 21-42

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 18 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19

17 Moser M McKnight KH Ammirati JF Studies on North American cortinariiI New and interesting taxa from the Greater Yellowstone areaMycotaxon 1995 55301-346

18 Moser M Ammirati JF Studies on North American cortinarii IV New andinteresting Cortinarius species (subgenus Phlegmacium) from oak forestsin Northern California Sydowia 1997 4925-48

19 Moser M Die Gattung Phlegmacium (Schleimkoumlpfe) Bad Heilbrunn GermanyJulius Klinkhardt 1960

20 Moser M McKnight KH Sigl M The genus Cortinarius (Agaricales) in theGreater Yellowstone area Mycorrhizal host associations and taxonomicconsiderations In Plant and their Environments Proceedings of the FirstBiennial Scientific Conference on the Greater Yellowstone Ecosystem Edited byDespain DG 1994 239-246

21 Moser M Ammirati JF Studies in North American cortinarii VI New andinteresting taxa in subgenus Phlegmacium from the pacific states ofNorth America Mycotaxon 2000 741-36

22 Moser MM Some aspects of Cortinarius associated with Alnus Journal desJEC 2001 447-76

23 Soop K Knuntsson T Un nuveau cortinaire de la section CalochroiJournal des JEC 2005 751-55

24 Froslashslev TG Jeppesen TS Laessoslashe T Seven new calochroid and fulvoidspecies of Cortinarius Mycol Res 2006 1101046-1058

25 Froslashslev TG Systematics of Cortinarius with emphasis on sectionCalochroi PhD thesis University of Copenhagen Faculty of Science 2007

26 Froslashslev TG Jeppesen TS Laessoslashe T Kjoslashller R Molecular phylogenetics anddelimitation of species in Cortinarius section Calochroi (BasidiomycotaAgaricales) in Europe Mol Phylogenet Evol 2007 44217-227

27 Rosenberg NA Nordborg M Genealogical trees coalescent theory andthe analysis of genetic polymorphisms Nat Rev Genet 2002 3380-390

28 Manos PS Stanford AM The historical biogeography of Fagaceaetracking the tertiary history of temperate and subtropical forests of theNorthern Hemisphere Intl J Plant Sci 2001 162S77-S93

29 Denk T Grimsson F Zetter R Episodic migration of oaks to IcelandEvidence for a North Atlantic ldquoland bridgerdquo in the latest Miocene Am JBot 2010 97276-287

30 Milne RI Northern hemisphere plant disjunctions a window on tertiaryland bridges and climatic change Ann Bot - London 2006 98465-472

31 Gitzendanner MA Strenge DD Soltis PS Chloroplast DNA intraspecificphylogeography of plants from the Pacific Northwest of America PlantSyst Evol 1997 206353-373

32 Williams JW Shuman BN Webb T III Bartlein PJ Leduc PL Late-quaternaryvegetation dynamics in North America scaling from taxa to biomes EcolMonogr 2004 74309-334

33 Hermann RK The genus Pseudotsuga Ancestral History and PastDistribution In Forest Research Laboratory Special Publication 2b Edited byCorvallis OR Oregon State University College of Forestry 1985

34 Brubaker LB Anderson PM Edwards ME Lozhkin AV Beringia as a glacialrefugium for boreal trees and shrubs new perspectives from mappedpollen data J Biogeogr 2005 32833-848

35 Ran J-H Wie X-X Wang X-Q Molecular phylogeny and biogeography ofPicea (Pinaceae) Implications for phylogeographical studies usingcytoplasmic haplotypes Mol Phylogenet Evol 2006 41405-419

36 Gardes M Bruns D ITS primers with enhanced specificity forbasidiomycetes application to the identification of mycorrhizae andrusts Mol Ecol 1993 2113-118

37 OrsquoDonnell K Fusarium and its near relatives In The Fungal HolomorphMitotic Meiotic and Pleomorphic Speciation in Fungal Systematics Edited byReynolds DR Taylor JW Wallingford UK CAB International 1993225-233

38 White TJ Bruns T Lee S Taylor J Amplification and direct sequencing offungal ribosomal RNA genes for phylogenetics In PCR Protocols A Guideto Methods and Amplifications Edited by Innis MA Gelfand H Sninsky JSWhite TJ New York Academic Press 1990315-322

39 Vilgalys R Hester M Rapid genetic identification and mapping ofenzymatically amplified ribosomal DNA from several Cryptococcusspecies J Bacteriol 1990 1724238-4246

40 Garnica S Weiszlig M Oertel B Oberwinkler F Phylogenetic relationships ofEuropean Phlegmacium species (Cortinarius Agaricales) Mycologia 2003951155-1170

41 Altschul SF Madden TL Schaumlffer AA Zhang J Zhang Z Miller WLipman DJ Gapped BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 1997 253389-3402

42 Katoh K Kuma K Toh H Miyata T MAFFT version 5 improvement inaccuracy of multiple sequence alignment Nucleic Acids Res 200533511-518

43 Stamatakis A RAxML-VI-HPC maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models Bioinformatics 2006222688-2690

44 Felsenstein J Confidence limits on phylogenies an approach using thebootstrap Evolution 1985 39783-791

45 Page RDM TreeView an application to display phylogenetic trees onpersonal computers Comput Appl Biosci 1996 12357-358

46 Swofford DL PAUP Phylogenetic Analysis Using Parsimony (and OtherMethods) Version 40b10 Sunderland Massachusetts Sinauer Associates2002

47 Clark AG Inference of haplotypes from PCR-amplified samples of diploidpopulations Mol Biol Evol 1990 7111-122

48 Prince EW Carbone I SNAP workbench management tool forevolutionary population genetic analysis Bioinformatics 2005 21402-404

49 Aylor DL Price EW Carbone I SNAP combine and map modules formultilocus population genetic analysis Bioinformatics 2006 221399-1401

50 Carbone I Liu Y-C Hillman BI Milgroom MG Recombination andmigration of Cryphonectria hypovirus 1 as inferred from genegenealogies and the coalescent Genetics 2004 1661611-1629

51 Myers SR Griffiths RC Bounds on the minimum number ofrecombination events in a sample history Genetics 2003 163375-394

52 Librado P Rozas J DnaSP v5 A software for comprehensive analysis ofDNA polymorphism data Bioinformatics 2009 251451-1452

53 SNAP Clade and Matrix Version 2 [httpwwwcalsncsueduplantpathfacultycarbonehomehtml]

54 Tajima F Statistical method for testing the neutral mutation hypothesisby DNA polymorphism Genetics 1989 123585-595

55 Fu YX Li W-H Statistical tests of neutrality of mutations Genetics 1993133693-709

56 Fu YX Statistical test of neutrality of mutations against populationgrowth hitchhiking and background selection Genetics 1997147915-925

57 Hudson RR Boos DD Kaplan NL A statistical test for detectinggeographic subdivision Mol Biol Evol 1992 9138-151

58 Simonsen KL Churchill GA Aquadro CF Properties of statistical tests ofneutrality for DNA polymorphism data Genetics 1995 141413-429

59 Clement M Posada D Crandall KA TCS a computer program to estimategene genealogies Mol Ecol 2000 91657-1660

60 Hey J Nielsen R Multilocus methods for estimating population sizesmigration rates and divergence time with applications to thedivergence of Drosophila pseudoobscura and D persimilis Genetics 2004167747-760

61 Griffiths RC Tavareacute S Ancestral inference in population genetics Stat Sci1994 9307-319

doi1011861471-2148-11-213Cite this article as Garnica et al Tracking the evolutionary history ofCortinarius species in section Calochroi with transoceanic disjunctdistributions BMC Evolutionary Biology 2011 11213

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Garnica et al BMC Evolutionary Biology 2011 11213httpwwwbiomedcentralcom1471-214811213

Page 19 of 19