673

Mycologia, 94(4), 2002, pp. 673–682.q 2002 by The Mycological Society of America, Lawrence, KS 66044-8897

A phylogenetic study of the genus Cookeina

Richard N. Weinstein1

Donald H. Pfister2

Harvard University Herbaria, 22 Divinity Avenue,Cambridge, Massachusetts 02138

Teresa Iturriaga

Departamento de Biologıa de Organismos, UniversidadSimon Bolıvar, Apartado Postal 89000, Sartenejas,Baruta, Edo. Miranda, Venezuela

Abstract: Cookeina, with seven recognized species, isone of the commonly encountered genera of the Sar-coscyphaceae (Pezizales) in tropical and subtropicalareas around the world. Morphologically the speciesare distinguished by combinations of several featuresincluding ascospore shape and surface relief, pres-ence and origin of apothecial hairs and presence orabsence of gelatinous material within the cortical lay-er of the excipular tissue. Color of the hymenium,attributed to carotenoid pigments, is particularly var-iable in some collections especially those referred toas C. speciosa. In this study phylogenetic analyses werecarried out using rDNA ITS and rDNA LSU sequenc-es. Forty-four collections were studied which includ-ed a broad sampling of color variants of C. speciosafrom a field site in Venezuela. The genus was shownto be monophyletic with several well-supported line-ages. These analyses generally support the estab-lished, morphologically distinguished taxa within amonophyletic genus Cookeina. Collections referred toas C. speciosa segregate within a clade in which hy-menial color differences are associated with groupswithin the clade. Cookeina sinensis is sister to C. tri-choloma but is distinct from it; C. indica fails to re-solve with any of the major clades. The placement ofC. insititia is ambiguous but it falls within Cookeinaand thus is considered in the genus Cookeina ratherthan in a separate genus, Boedijnopeziza.

Key Words: biogeography, ITS sequences, Pezi-zales, Sarcoscyphaceae

Accepted for publication January 9, 2002.1 Current address: Department of Botany, University of Tennessee,326 Hesler Hall, Knoxville, Tennessee 37996.2 Corresponding author, Email: dpfister@oeb.harvard.edu

INTRODUCTION

Of the tropical and subtropical members of the orderPezizales few are as commonly encountered and col-lected as are species belonging to the genus CookeinaKuntze (1891). Their brightly colored and relativelylarge apothecia assure that even the most casual ob-server will see them. The seven recognized morpho-species are defined on a series of distinctive and un-ambiguous macroscopic and microscopic charactersrelated to occurrence and distribution of the hairs onthe surface of the apothecium, spore shape and sur-face ornamentation, and presence or absence of ge-latinous material in the tissue of excipulum. Speciesoccur on fallen angiosperm branches, trunks and, oc-casionally, on durable fruits. Some taxa are wide-spread, such as C. colensoi (Berk.) Seaver, C. tricholoma(Mont.) Kuntze and C. speciosa (Fr. : Fr.) Dennis, butothers are more restricted in distribution, such as, C.sinensis Z. Wang, C. venezuelae (Berk. & M. A. Curtis)Le Gal, C. insititia (Berk. & M. A. Curtis) Kuntze, andC. indica Pfister & R. Kaushal. Hymenial colors rangefrom white to beige, yellow, orange, scarlet, and evento chocolate brown. Color variation is particularly pro-nounced among collections of C. speciosa. Arpin(1969) analyzed the pigments and reported caroten-oids in two species, C. tricholoma and C. sulcipes(Berk.) Kuntze (5C. speciosa).

We initiated this study because we observed bothdistinct distributional patterns of species in this ge-nus and variation in hymenial colors. The study fo-cuses on the phylogenetic relationships among thetaxa and their color variants. We obtained materialof all taxa and we sampled broadly across the geo-graphical range of taxa where we had appropriatematerial. A companion study will treat the genusmonographically.

The family Sarcoscyphaceae, in which Cookeina hasa central position, has a convoluted history, summa-ries of which may be found in Eckblad (1968), Korf(1970), Rifai (1968) and Harrington et al (1999).The Sarcoscyphaceae has been referred to the sub-order Sarcoscyphineae (Rifai 1968). According to ananalysis of the Sarcoscyphineae by Harrington et al(1999) it is a paraphyletic assemblage, but a cladecontaining Sarcoscypha (Fr.) Boud., Phillipsia Berk.,Cookeina and related taxa is well supported and rep-resents the Sarcoscyphaceae of most recent authors.

674 MYCOLOGIA

In that study (Harrington et al 1999) Cookeina wasshown to be sister to Microstoma Bernstein, a positionthat confirms Korf’s (1972, 1973) view that these gen-era are closely related. Korf (1972) proposed a tribe,the Boedijnopezizeae, for Cookeina, Microstoma, andBoedijnopeziza S. Ito & S. Imai. These genera sharefeatures such as simultaneous maturation of asco-spores within apothecia, a particular ascus construc-tion in which the base of the ascus abruptly constrictsto a thin supporting hypha and a dense network ofanastomosing paraphyses. Microstoma is temperate indistribution. Boedijnopeziza, based on B. insititia, isknown from Asia and is recognized as a distinct taxon(Rifai 1968, Korf 1972, 1973) or treated as a synonymof Cookeina (Le Gal 1953, Denison 1967, Pfister 1973,Pfister and Kaushal 1984). Korf (1983) suggested arelationship between the Boedijnopezizeae and Cyt-taria Berk. Subsequent analysis of sequence data in-dicates that Cyttaria is only distantly related to thePezizales (Landvik 1996).

Our goal in this study was to test the monophylyof the genus Cookeina and to evaluate the morpho-species as currently recognized, particularly with re-gard to biogeography and color variation. To accom-plish our goal we used phylogenetic analyses of char-acters derived from nuclear encoded ribosomal DNA(rDNA).

MATERIALS AND METHODS

Material studied. Specimens of Cookeina used for DNAanalysis were collected by the authors or obtained from col-lectors or herbaria (TABLE I). To distinguish specimens andavoid confusion with herbarium and field numbers, vouch-er numbers were assigned to collections. The ingroup con-sisted of 44 specimens of Cookeina, including 16 specimensof C. speciosa displaying distinct color variants. The out-group was Microstoma floccosum.

Molecular techniques. DNA was isolated from approximately50 mg of dried apothecial tissue of each sample, which wasfirst cleaned of extraneous material using a small paintbrushfollowed by short blasts of compressed air. Dried apotheciawere ground to a fine powder in liquid nitrogen. Powderedsamples were extracted in 800 mL SDS lysis buffer (1% SDS;200 mM Tris, pH 7.5; 250 mM NaCl; 25 mM EDTA; 1% po-lyvinylpyrolidone, adjusted to pH 6.8) and incubated, withoccasional mixing, in a water bath for 1 h at 65 C. The su-pernatant was extracted twice with an equal volume of chlo-roform-isoamyl alcohol (24:1) and transferred to a 2.0 mLmicrocentrifuge tube. DNA was purified with a QIAGENBlood and Tissue kit (Valencia, California) by incubating at70 C for 10 min with an equal volume of buffer AL (1:1supernatant: buffer) followed by the addition of 100% eth-anol equal in volume to the original supernatant. DNA waspurified using QIAamp spin columns; bound DNA waswashed twice with applications of 500 mL wash buffer (buffer

AW, Qiagen) and then eluted from the spin column with100–200 mL ddH2O warmed to 70 C. Eluate was passedthrough the column a second time to increase yield.

DNA extract was used for polymerase chain reaction(PCR) of the internal transcribed spacers (ITS 1 and ITS2) and the 5.8S region of the nuclear ribosomal DNA usingthe fungus-specific oligonucleotide primers ITS 4 and ITS5 (White et al 1990). Four mL of PCR product was quanti-fied on an ethidium bromide-stained 1.0% agarose gel, andthe remaining PCR product was purified using Microcon100 microconcentrator columns (Amicon Inc., Beverly, Mas-sachusetts).

For dye-terminator cycle-sequencing involving the ITS re-gion, ITS primers 2, 3, 4 and 5 were used; for the LSUregion primers LR0R, LR3R, LR3 and LR5 were used. Se-quencing reactions were run on an Applied Biosystems 377automated DNA sequencer.

Analytical techniques. Sequences were edited and assem-bled using Sequencher 3.0 (GeneCodes, Ann Arbor, Mich-igan) and aligned manually in the data editor of PAUP4.0d64 (Swofford 1991) with alignment gaps inserted tomaximize aligned sites. Sequences have been deposited inGenBank (TABLE I) and the data matrix is available fromTreeBASE as accession number SN1105.

Phylogenetic analyses were performed in PAUP 4.0d64(Swofford 1991). After coding gaps in various fashions andfinding basically similar results, we determined to treat gapsas missing data (gaps 5 missing coding); ambiguous regionsin the alignment (characters 67–76, 79–93, 123–145, 160–182, 238–263, 266–275, 286–312, 530–545, 545–562, 687–716) were recoded as single characters (characters 782–791,respectively) using the methods described by LaGreca(1999), with each newly designated character given a weightequal to the number of characters from the correspondingambiguous region. Using this system, the number of infor-mative characters was increased from 246 to 256. Due tothe size of the data set, we were limited to heuristic search-es, which were performed in two parts. First, 100 heuristicsearches were performed with random taxon addition andTBR branch swapping, with MAXTREES set to 200, keepingup to 2 trees per replicate. Second, all the shortest treesfrom the first part of the analysis were used as starting treesfor complete TBR branch-swapping with MAXTREES set to15 000. Relative robustness of individual clades was assessedby the bootstrap (Felsenstein 1985) using 100 heuristicsearches, simple taxon addition sequences, TBR branchswapping and MAXTREES set to 1000.

Scanning electron microscopy (SEM), light microscopy (LM)and morphological studies. Ascospores of Cookeina collec-tions were gathered by rehydrating a small piece of apothe-cium in several drops of water and then carefully separatingthe hymenial tissue from the excipular tissue. The hymenialtissue was macerated using a scalpel blade and furthersquashed with a pipette tip. An aliquot of this macerate wasexamined under LM. When free floating spores were ob-served, a drop of this spore suspension was pipetted onto acover slip, dried, placed on a stub and sputter-coated witha gold-palladium alloy. Observations were made on an AM-RAY model 1000 SEM.

675WEINSTEIN ET AL: COOKEINA PHYLOGENY

TABLE I. List of collections and sequences of Cookeina and Microstoma species used for phylogenetic analysis

Species Geographic origin Specimena Colorb CodeGenBankaccession

C. colensoiC. colensoiC. colensoiC. colensoiC. colensoiC. colensoiC. colensoiC. colensoiC. colensoiC. indicaC. insititiaC. insititia

IndiaMexicoNew ZealandNew ZealandNew ZealandNew ZealandAustraliaAustraliaGuizhou, ChinaYunnan, ChinaYunnan, ChinaGuizhou, China

FH, PAN 18557CUP 62500PDD 68535PDD 66040PDD 68628PDD 55306DAR 63642DAR 63646HMAS 59537HMASHMAS 70078HMAS 71942

Not recordedNot recordedNot recordedNot recordedNot recordedNot recordedOrangeOrange-saffronNot recordedNot recordedNot recordedNot recorded

11860

112113114115116117121119123124

AF394532AF394040AF394034AF394035AF394036AF394037AF394038AF394039AF394532AF394531AF394030AF394031

C. insititiaC. insititiaC. sinensisC. sinensisC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosaC. speciosa

Yunnan, ChinaYunnan, ChinaYunnan, ChinaYunnan, ChinaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuelaVenezuela

FH Wang sp 1FH Wang sp 2HKAS 14679HMAS 70088FH Iturriaga 10E-D5FH Iturriaga 2610FH Iturriaga 7A-D4FH Iturriaga D2-5AFH Iturriaga D2-4AFH Iturriaga 1D-D6FH Iturriaga 2D-D4FH Iturriaga 4D-D4FH Iturriaga D2-2AFH Iturriaga 1C-D4FH Iturriaga 4A-D4FH Iturriaga 1E-D5

Not recordedNot recordedNot recordedNot recordedLight brown (9C5-9C6)Pale yellow (4A3-4A4)Light coral (10A5)Coral (10A7)Coral (10A7)Mauve (10C4-10C6)Mauve (10C4-10C6)Mauve (10C4-10C6)Orange (7A8)Orange (7A8)Deep coral (11A8)Deep coral (11A8)

12512611148222527282930313233343637

AF394032AF394033AF394028AF394027AF394015AF394005AF394006AF394007AF394012AF394016AF394017AF394004AF394008AF394011AF394014AF394003

C. speciosaC. speciosaC. speciosaC. speciosaC. tricholomaC. tricholomaC. tricholomaC. tricholomaC. tricholomaC. tricholomaC. tricholomaC. tricholomaC. venezuelaeC. venezuelaeC. venezuelaeC. venezuelae

ThailandThailandBorneo, MalaysiaColombiaYunnan, ChinaYunnan, ChinaVenezuelaVenezuelaVenezuelaPuerto RicoThailandBorneo, MalaysiaGuadeloupePuerto RicoVenezuelaVenezuela

FH Pfister 7131FH Pfister 7143C TL 6035FH Muneton 296HMAS 23238HMASFH Iturriaga 1D-D5FH Iturriaga 1B-D5FH Iturriaga 2705FH Pfister 38FH Pfister 7170C TL 6101FH Pfister 1161FH Cantrell 3381FH Iturriaga 6066FH Iturriaga 6065

Not recordedNot recordedNot recordedWhiteNot recordedNot recordedOrange (7A8)Orange (7A8)Not recordedNot recordedNot recordedNot recordedNot recordedNot recordedSalmon (10A5)Salmon (10A5)

71317143603511012012238394015

71706101

19204243

AF394009AF394010AF394018AF394013AF394025AF394026AF394021AF394022AF394023AF394024AF394020AF394019AF394042AF394041AF394043AF394044

M. floccosumM. floccosum

MexicoMexico

FH K. GriffithFH K. Griffith

Not recordedNot recorded

4546

AF394045AF394046

a Voucher specimens are deposited in the herbaria indicated in boldface type as follows: CUP, Plant Pathology Herbarium,Cornell University; DAR, Plant Pathology Branch Herbarium, Biological and Chemical Research Institute, New South Wales;HMAS, Mycological herbarium, Systematic Mycology and Lichenology Laboratory, Beijing; FH, Farlow Herbarium, HarvardUniversity; PAN, Botany Department Punjab University; PDD, New Zealand Fungal Herbarium, Landcare Research, Auckland,New Zealand. Code refers to the numbers used in this study. Color from Methuen is noted in parentheses where available.

b Numbers and letters in parentheses refer to Methuen (Kornerup and Wanscher 1978) notations.

676 MYCOLOGIA

Field studies. A detailed field study was undertaken in theAmazonian rainforest in Yutaje, Amazonas State, Venezuela.Cookeina species were collected in numbered plots on pre-cisely designated and coded substrates. Thus, single collec-tions could be traced precisely to a particular substrate andits neighbors on that substrate could be identified.

Color nomenclature. Colors were assigned in the field andcorrelated with Methuen (Kornerup and Wanscher 1978)color names and numbers. Methuen color notations are giv-en in TABLE I.

RESULTS

Alignment. ITS1 and ITS2 were aligned along withthe intervening 5.8S region and flanking partial se-quences of 18S and 25S rDNA. All positions weregenerally able to be aligned within collections of thesame species; alignment across species often requiredthe addition of gaps. Because of the large data setthis resulted in ten main areas of ambiguous align-ment (as described in the Methods). The 5.8S regionwas almost identical across all taxa, and the alignedlength of all sequences (including inserted gaps) was781 bp.

Parsimony analyses. Under gap 5 missing data therewere 256 informative characters yielding 15 000equally parsimonious trees of 3083 steps (consistencyindex, CI 5 0.960; retention index, RI 5 0.980). Phy-logenetic analysis identified four well-supported lin-eages of rDNA as measured by bootstrapping (FIG.1). The monophyly of the C. speciosa and C. tricho-loma lineages was strongly supported (for each, boot-strap 5 100%) and the lineage consisting of C. col-ensoi and C. venezuelae was supported by a bootstrapvalue of 99%. Further bootstrap support of 100% and93%, respectively, was obtained for both the C. col-ensoi and C. venezuelae clades, while the C. insititiaclade was supported by a bootstrap value of 100%.The only specimen of C. indica, although falling with-in Cookeina, failed to clearly resolve with any of thegreater clades.

The C. speciosa clade includes color variants fromVenezuela, a white specimen from Colombia, two iso-lates from Thailand and one from Sabah, Malaysia,which together form two sub-clades with good sup-port. One major sub-clade consisting of mostly Ve-nezuelan specimens (bootstrap 5 95%) contains twogroups: one that includes ‘‘deep coral’’ variants(specimen nos. 36 and 37; bootstrap 5 100%) andanother with ‘‘mauve/light brown’’ variants (speci-men nos. 22, 30, 31 and 32; bootstrap 5 97%). Thespecimen from Sabah, Malaysia is basal to this sub-clade. The other major well-supported sub-clade(bootstrap 5 99%) also consists of two smallergroups. One of these groups (specimen nos. 25, 27,

28, 29, 33, 34, and 110; bootstrap 5 100%) consistsmostly of Venezuelan specimens ranging in colorfrom orange, coral, light coral, to yellow. In addition,the white specimen from Colombia falls into thisclade. The other group consists of collections fromThailand (bootstrap 5 100%).

The C. tricholoma clade includes specimens fromSabah, China, Puerto Rico, Thailand and Venezuela.There is very little difference in ITS sequences withinthis clade despite the geographical spread of the col-lections (FIG. 2). Cookeina sinensis is basal in the C.tricholoma clade with 100% bootstrap support.

The clade comprising C. venezuelae and C. colensoiis strongly supported at the highest order (bootstrap5 99%) and resolution for each sub-clade alone ishigh (bootstrap of 93% and 100% respectively). Re-gional variation is apparent in each of these clades:all southern hemisphere (Australia and New Zea-land) specimens of C. colensoi cluster together withvirtually no ITS sequence differences among speci-mens while the specimens from the northern hemi-sphere (China, India, and Mexico) resolve basally tothe southern hemisphere collections. Cookeina vene-zuelae shows considerable variation within the rela-tively local Caribbean region, with Venezuelan spec-imens resolving together (bootstrap 5 100%), andspecimens from Guadeloupe and Puerto Rico clus-tering together (bootstrap 5 100%).

Cookeina insititia also demonstrates considerablevariation within the clade, especially considering thateach of the four specimens analyzed were collectedin China, three from Yunnan Province (FIG. 2).

SEM spore images. SEM images of spores from Cook-eina (FIG. 3) aid and clarify descriptions derived fromLM studies. Spores of several species are described asstriate in LM studies but SEM reveals that the natureand arrangement of the striation varies. Spore orna-mentation in C. speciosa, described as striate, consistof irregular anastomozing longitudinal ridges thatappear somewhat wrinkled. Spores of C. tricholoma,also described as striate under LM, exhibit a moreregular pattern of relatively straight and parallel lon-gitudinal ridges. Cookeina indica, described as havingfine longitudinal markings, has regular parallel lon-gitudinal ridges that are wider than those in C. tri-choloma. Cookeina venezuelae, typically described ashaving both longitudinal and transverse striations,can be more accurately described as having longitu-dinal ribs with fine transverse interconnecting ridges.Spores of both C. colensoi and C. insititia are uni-formly smooth. Spores of C. sinensis were difficult toobtain because the material was scanty and question-ably mature, but those we found were smooth.

677WEINSTEIN ET AL: COOKEINA PHYLOGENY

FIG. 1. Strict consensus of 15 000 equally parsimonious trees (3083 steps) generated under gap 5 missing data. Numbersabove the branches represent bootstrap support for 1000 replicates. Location codes are as follows: A—Australia, B—Borneo(Sabah), C—China, Co—Colombia, G—Guadeloupe, I—India, M—Mexico, NZ—New Zealand, PR—Puerto Rico, T—Thai-land, V—Venezuela.

Field studies. The field study focused on C. speciosabut also yielded collections of C. tricholoma. Twelvecollections of C. speciosa from the field plots wereincluded in this study. Two collections, nos. 27 and36, recorded as different in color, were collected onthe same substrate in the same plot on the same day.

However, these collections occur in different cladesof the C. speciosa group. This strongly supports thehypothesis that the color variants are distinct. Simi-larly, two other collection, nos. 22 and 37, were madein the same plot, on the same substrate, on the sameday. Both of these collections fall in the same clade.

678 MYCOLOGIA

FIG. 2. Phylogram representing one of 1000 equally parsimonious trees for ITS sequences. Terminal taxa are individualcollections (see TABLE I); branch lengths are according to scale. Geographical codes are as for FIG. 1.

679WEINSTEIN ET AL: COOKEINA PHYLOGENY

FIG. 3. Ascospore ornamentation as seen by SEM. A. Cookeina speciosa. Face view showing anastomosing ridges (FH,Aldava 296). B. Cookeina indica. Face view showing 13–15 longitudinal ribbon-like ridges (HMAS, Z. Wang). C. Cookeinavenezuelae. Profile view showing 6–7 longitudinal ribs with regular, fine transverse connecting ridges (FH 1107). D. Cookeinatricholoma. Face view showing 22–24 longitudinal ridges (OSC 67751). E. Cookeina colensoi. Face view (PDD 68628). F. Cookeinasinensis. Profile view (HMAS 14679). G. Cookeina insititia. Profile view, surface smooth (HMAS, Z. Wang). H. Microstomafloccosum. (FH, K. Griffith). Scale bar 5 10 mm.

680 MYCOLOGIA

Five collections (nos. 27, 31, 32, 34, 36) were madein plot D4. Of these nos. 31, 32, and 36 group to-gether in the dark-color clade; nos. 27 and 34 are inthe light-color sister clade. Three collections (nos. 28,29, 33) were made in plot D2. Collection 28 and 29were from the same substrate, and together with 33resolve in the light-color clade.

DISCUSSION

General morphology. Phylogenetic analyses reveal sev-eral distinct rDNA lineages with resolution intoclades that correlate with the morphological groupsthat have been recognized as species. The monophy-letic group consisting of C. speciosa, C. tricholoma,and C. sinensis are hirsute, stipitate and lack a de-fined gelatinous layer in the excipulum at maturity.Spores are longitudinally ridged in the two formerspecies and smooth in the latter. Cookeina venezuelaeand C. colensoi, which also constitute a monophyleticgroup, have a well-defined excipular gel layer, are ses-sile or short stipitate and glabrous or with low scalesor pustules. Spores are ornamented in C. venezuelaeand smooth in C. colensoi. Cookeina indica, with or-namented spores, and C. insititia, with smoothspores, are intermediate regarding excipular featuresand did not clearly resolve in relation to the largergroupings.

Spore ornamentation has no clear correlation toany of the major lineages. Cookeina sinensis, for ex-ample, has smooth spores or slightly wrinkled spores(Wang 2001) yet resolves with taxa having striatespores, while C. venezuelae, with a unique spore or-namentation, as dissimilar to striate spores as they areto smooth spores, resolves with the smooth-spored C.colensoi. This lack of correlation contrasts with obser-vations by Hansen et al (1999), who found that ITSlineages within Phillipsia, also a member of the Sar-coscyphaceae, were all supported by spore morphol-ogy.

Cookeina insititia is the only species with stipitate,hirsute apothecia and a gelatinous layer, but its rela-tionship to one or the other of the main groups isuncertain. Likewise, C. indica, the only stipitate, gla-brous species without a gelatinous layer, also fails toconsistently resolve with one or the other of thegreater monophyletic clades.

Color and regional variation in C. speciosa. Cookeinaspeciosa is noteworthy within the Pezizales for havinga broad spectrum of apothecial colors. Our studiesto date show no consistent anatomical differencesamong the color forms. On the other hand, rDNAanalysis indicates collections in the same color rangesgroup together into two well-supported clades. Thus,

one clade with a 65% bootstrap includes collectionswith darker colored apothecia, in the brown-mauve-dark coral (purplish) range; the second clade, with a99% bootstrap, has lighter colored apothecia, in theyellow, orange, light coral, coral and white range. InArpin’s (1969) study he found no evidence for dif-ferences in the carotenoid content in two color var-iants of C. speciosa (as C. sulcipes) he studied.

In this study the northern region of South Americarepresents the reference point from which variationcan be assessed since a range of color forms weredocumented from there and incorporated in thedata set. It is interesting therefore that the collectionfrom Sabah groups with the more darkly pigmentedspecimens from Venezuela, while the collectionsfrom Thailand always group with the clade compris-ing the more lightly pigmented specimens from Ve-nezuela. The sequence difference between the Sabahand Thailand specimens is considerable, given thegeographical proximity of the two locations. This dis-parity cannot be accounted for by biogeographicalfeatures such as the Wallace line, which passes to theeast of Borneo. The two regions are part of the samebiogeographical zone, although associated with dif-ferent forest types. Furthermore, field studies in Ve-nezuela demonstrate a range of ITS and apothecialcolor differences present even within small areas andon the same pieces of downed wood.

Color as a reliable taxonomic character. Conventionsin taxonomic mycology have had an important influ-ence on species concepts within certain groups offungi. Ascomycete systematists have largely discount-ed color variations in favor of reliance on microscop-ic characters. Although color variation was noted andused by some earlier mycologists to distinguish spe-cies of Cookeina, more recent authors have groupedthe various color forms into more or less anatomi-cally uniform morpho-species. Our data suggests thata closer look at these species complexes is requiredto understand the characters used in classification.We believe there are at least two taxa within the C.speciosa complex.

Resolution of the taxonomic status of C. sinensis.Cookeina sinensis was recognized on morphologicalgrounds. It has smooth or slightly wrinkled spores butotherwise is morphologically similar to C. tricholoma.We examined both the sterile holotype and a fertileparatype of C. sinensis. Our morphological observa-tions are still tentative because of the inadequacy ofthe available material. Even the paratype material isbarely mature and it is possible that the smoothspores of C. sinensis could represent immature sporesof C. tricholoma. Our studies indicate that the stria-tions on ascospores of C. tricholoma form relatively

681WEINSTEIN ET AL: COOKEINA PHYLOGENY

late in the development of the ascospores. On mo-lecular grounds C. sinensis falls outside and sister tothe C. tricholoma group. The C. tricholoma collectionsfrom such disparate locations as Sabah, China,Puerto Rico, Thailand, and Venezuela, show short in-ternal branch lengths (FIG. 2). While this lack of var-iation may somehow be unique to the biology of C.tricholoma, it also highlights the distinct taxonomicprovenance of C. sinensis, which always resolves sisterto the C. tricholoma collections (bootstrap 5 100%).Furthermore, both specimens of C. sinensis and twospecimens of C. tricholoma were collected in YunnanProvince of China, further weakening the argumentthat C. sinensis may represent either an immaturecollection of C. tricholoma or a regional variant of it.These phylogenetic results, along with the morpho-logical difference that characterizes C. sinensis, pro-vide evidence for its recognition as a distinct species.Wang (2001) reported C. sinensis from Taiwan andprovided SEM photomicrographs of spores from atype collection and one from Taiwan.

Resolution of the taxonomic status of C. insititia.There has been debate over the taxonomic status ofC. insititia. The spore shape, presence of a gel layer(known also in C. colensoi and C. venezuelae) and theform and origin of the hairs seem distinct in C. in-sititia. These features led Ito and Imai (1937) to erecta new genus, Boedijnopeziza. The genus was support-ed by some (Rifai 1968, Korf 1972, 1973) and dis-puted by others (Denison 1967, Eckblad 1968, Le Gal1953, Pfister 1973, Pfister and Kaushal 1984). Molec-ular data supports the position of this species withinthe genus Cookeina, but because of lack of resolutionin the strict consensus tree, its position in relation-ship to the other taxa remains problematic. Korf(1971) included C. colensoi in Boedijnopeziza based onthe presence of gel in the excipular tissue of thatspecies. Pfister (1973) recognized that C. colensoi andC. venezuelae shared this character. One could rec-ognize Boedijnopeziza for C. insititia, C. colensoi, andC. venezuelae since the molecular data do not ruleout such a position, but there remains little otherthan the somewhat ambiguous character of gel tounite them morphologically. In our opinion thechoice to recognize Boedijnopeziza obscures the con-tinuity of characters in this monophyletic group.

Distribution of C. colensoi. Denison (1967) report-ed that C. colensoi occurred chiefly south of the equa-tor, with one collection from Venezuela and nonefrom Central America. In our monographic studieswe have determined that the Venezuelan collectionrepresents a heretofore undescribed species. How-ever, collections from Mexico, India, and China havecome to light and have been included in this analysis.

The species is best considered sub-tropical, with prin-ciple distribution in the southern hemisphere.

ACKNOWLEDGMENTS

We thank the following people who provided specimens forour use in this study: Sharon Cantrell, Peter Johnston, Rich-ard P. Korf, Thomas Læssøe, Zheng Wang, Wen-yingZhuang and the curators of C, CUP, DAR HMAS, KMAS,PDD. We appreciate the comments given us by MeredithBlackwell and Scott LaGreca who read the manuscript invarious versions. Scott LaGreca also assisted us with varioustechnical aspects. We gratefully acknowledge the supportprovided by NSF DEB-9521944.

LITERATURE CITED

Arpin N. 1969. Les carotenoıdes des discomycetes: Essai chi-miotaxinomique. Bull Mens Soc Linn Soc Bot Lyon38(suppl):1–169.

Denison WC. 1967. Central American Pezizales. II. The ge-nus Cookeina. Mycologia 59:306–317.

Eckblad F-E. 1968. The genera of operculate discomycetes.A re-evaluation of their taxonomy, phylogeny and no-menclature. Nytt Mag Bot 15:1–191.

Felsenstein J. 1985. Confidence limits on phylogenetics: anapproach using the bootstrap. Evolution 39:783–791.

Harrington FA, Pfister DH, Potter D, Donoghue MJ. 1999.Phylogenetic studies within the Pezizales. I. 18S rDNAsequence data and classification. Mycologia 91:41–50.

Hansen K, Pfister DH, Hibbett DS. 1999. Phylogenetic re-lationship among species of Phillipsia inferred frommolecular and morphological data. Mycologia 91:299–314.

Ito S, Imai S. 1937. Fungi of the Bonin Islands II. TransSapporo Nat Hist Soc 15:52–59.

Korf RP. 1970. Nomenclatural notes. VII. Family and tribenames in the Sarcoscyphineae (Discomycetes) and anew taxonomic disposition of the genera. Taxon 19:782–786.

———. 1971. Some new discomycete names. Phytologia 21:201–207.

———. 1972. A synoptic key to the genera of the Pezizales.Mycologia 64:937–994.

———. 1973. Discomycetes and Tuberales. In: AinsworthGC, Sparrow FK, Sussman AS, eds. The fungi: an ad-vanced treatise. Vol. 4A. New York: Academic Press. p249–319.

———. 1983. Cyttaria (Cyttariales): coevolution with Noth-ofagus, and evolutionary relationship to the Boedijno-pezizeae (Pezizales, Sarcoscyphaceae). Austral J Bot,Suppl Ser 10:77–87.

Kornerup A, Wanscher JH. 1978. Methuen handbook ofcolour. Ed. 3. London: Eyre Methuen. 252 p.

Kuntze O. 1891. Revisio generum plantarum vasculariumomnium atque cellularium multarum secundum legesnomeclaturae internationales cum enumeratione plan-tarum exoticarum in itinere mundi collectarum. Vol.2. Leipzig: A. Felix. 1011 p.

682 MYCOLOGIA

LaGreca S. 1999. A phylogenetic evaluation of the Ramalinaamericana chemotype complex (lichenized Ascomyco-ta, Ramalinaceae) based on rDNA ITS sequence data.Bryologist 102:602–608.

Landvik S. 1996. Phylogenetic rDNA studies of discomycetes(Ascomycota). Umea, Sweden: Umea Univ. 42 p.

Le Gal M. 1953. Le discomycetes de Madagascar. ProdrFlore Mycol Madagascar 4:1–465.

Pfister DH. 1973. Notes on Caribbean discomycetes IV.Cookeina venezuelae, C. colensoi and the genus Boedijno-peziza. Phytologia 27:55–62.

———, Kaushal R. 1984. Cookeina indica, a new speciesfrom India with a key to the species of Cookeina. My-cotaxon 20:117–121.

Rifai MA. 1968. The Australasian Pezizales in the herbariumof the Royal Botanic Gardens, Kew. Verh Kon NedAkad Wetensch, Afd Natuurk, Tweede Sect 57:1–295.

Swofford DL. 1991. PAUP-phylogenetic analysis using par-simony, ver. 4.064d. Champaign, Illinois: Illinois Natu-ral History Survey.

Wang YZ. 2001. Discomycetes of the Sarcoscyphaceae in Tai-wan. Mycotaxon 79:329–336.

White TJ, Bruns T, Lee S, Taylor JW. 1990. Amplificationand direct sequencing of fungal ribosomal RNA genesfor phylogenetics. In: Innis MA, Gelfand DH, SninskyJJ, White T, eds. PCR protocols: a guide to methodsand applications. San Diego: Academic Press. p 315–322.