Chapter 6

Phylogenetic Relationships of Harpyionycterine Megabats(Chiroptera: Pteropodidae)

NORBERTO P. GIANNINI1,2, FRANCISCA CUNHA ALMEIDA1,3,

AND NANCY B. SIMMONS1

ABSTRACT

After almost 70 years of stability following publication of Andersen’s (1912) monograph onthe group, the systematics of megachiropteran bats (Chiroptera: Pteropodidae) was thrown intoflux with the advent of molecular phylogenetics in the 1980s—a state where it has remained eversince. One particularly problematic group has been the Austromalayan Harpyionycterinae,currently thought to include Dobsonia and Harpyionycteris, and probably also Aproteles. In thiscontribution we revisit the systematics of harpyionycterines. We examine historical hypothesesof relationships including the suggestion by O. Thomas (1896) that the rousettine Boneia bidensmay be related to Harpyionycteris, and report the results of a series of phylogenetic analysesbased on new as well as previously published sequence data from the genes RAG1, RAG2, vWF,c-mos, cytb, 12S, tVal, 16S, and ND2. Despite a striking lack of morphological synapomorphies,results of our combined analyses indicate that Boneia groups with Aproteles, Dobsonia, andHarpyionycteris in a well-supported, expanded Harpyionycterinae. While monophyly of thisgroup is well supported, topological changes within this clade across analyses of different datapartitions indicate conflicting phylogenetic signals in the mitochondrial partition. The positionof the harpyionycterine clade within the megachiropteran tree remains somewhat uncertain.Nevertheless, biogeographic patterns (vicariance-dispersal events) within Harpyionycterinaeappear clear and can be directly linked to major biogeographic boundaries of theAustromalayan region. The new phylogeny of Harpionycterinae also provides a new frameworkfor interpreting aspects of dental evolution in pteropodids (e.g., reduction in the incisordentition) and allows prediction of roosting habits for Harpyionycteris, whose habits areunknown.

INTRODUCTION

During the last decade, the traditionalclassification of Megachiroptera (Mammalia:Chiroptera) has been challenged by a numberof molecular phylogenetic studies that havecalled into question many of the systematicgroupings established by Andersen (1912)and revised recently by Bergmans (1997).One particularly controversial megachirop-teran taxon has been Harpyionycterinae

Miller, 1907, an Austromalayan group cur-rently thought to include HarpyionycterisThomas, 1896, Dobsonia Palmer, 1898, andalmost certainly Aproteles Menzies, 1977(Giannini et al., 2006). There has beendisagreement about the affinities of thenominate genus since its description, largelydue to a suite of unique craniodentalcharacter states seen in Harpyionycteris (seeAndersen, 1912). Thomas (1896) rathervaguely suggested placement of Harpyionyc-

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:01 183 Cust # 582-6

1 Department of Mammalogy, American Museum of Natural History, Central Park West at 79th Street, New York, NY 10024-5192, USA (norberto@amnh.org).

2 Programa de Investigaciones de Biodiversidad Argentina, Facultad de Ciencias Naturales e Instituto Miguel Lillo, Miguel Lillo205, Universidad Nacional de Tucuman, Tucuman, CP 4000, Argentina.

3 Sackler Institute for Comparative Genomics, Molecular Systematics Laboratory, American Museum of Natural History, CentralPark West at 79th Street, New York, NY 10024, USA.

teris with Rousettus Gray, 1921, and BoneiaJentink, 1879. In contrast, Andersen (1912)linked Harpyionycteris with Dobsonia (seebelow), and Boneia with Rousettus. Boneiawas synonymized with Rousettus by Berg-mans and Rozendaal (1988) and this arrange-ment has been followed by many authors(Bergmans, 1994, 1997; Corbet and Hill,1992; Simmons, 2005).

Andersen (1912) formally followed Miller(1907) in recognizing Harpyionycterinae as adistinct group, but wrote in extenso about aclose relationship between Harpyionycterisand Dobsonia. He concluded that ‘‘So evidentis the phylogenetic connection between thesetwo genera that Harpyionycteris may be said,almost with certainty, to be the peculiarlymodified Philippine representative of theAustro-Malayan Dobsonia’’ (Andersen, 1912:803). Miller and Hollister (1921) subsequent-ly described Harpyionycteris celebensis fromSulawesi, providing a further biogeographiclink between Harpyionycteris and Dobsonia.

In spite of Andersen’s (1912) arguments,all subsequent authors (e.g., Corbet and Hill,1992; Koopman, 1993, 1994; Tate, 1951;Slaughter, 1970) followed Miller (1907) inconsidering Harpyionycteris as a peculiarpteropodid best placed in a subfamily of itsown. However, two recent studies providedevidence supporting Andersen’s (1912) hy-pothesis of a close relationship betweenHarypionycteris and Dobsonia. Support forthis grouping was found in Romagnoli andSpringer’s (2000) analysis of morphologicaldata, although caution must be exercised ininterpreting these results because of minimalclade support and the small number ofcharacters used in that study. More compel-lingly, separate and combined parsimony

analyses of two coding genes, one nuclear(exon 28 of the von Willebrand factor gene)and one mitochondrial (cytochrome b),recovered a highly supported clade groupingHarpyionycteris whiteheadi and four speciesof Dobsonia from three distinct speciesgroups (Giannini et al., 2006). As a conse-quence of these results, Giannini et al. (2006)expanded Harpyionycterinae Miller, 1907, toinclude Dobsonia. They also concurred withprevious authors in concluding that Aprotelesprobably also belongs in this subfamily basedboth on general morphological evidence(Flannery, 1995; Menzies, 1977; Gianniniand Simmons, 2005; Springer et al., 1995)and results of all previous phylogeneticanalyses in which Aproteles was included(Colgan and da Costa, 2002; Giannini andSimmons, 2003, 2005; Jones et al., 2002;Kirsch et al., 1995; Romagnoli and Springer,2000; Springer et al., 1995).

The harpyionycterines as currently under-stood occupy three major Austromalayansubregions, the Philippine, Wallacean, andPapuan (table 1). The distributions of a fewspecies marginally exceed the boundaries ofthose territories and enter the Australian andSundaic regions (see Byrnes, 2005; Corbet andHill, 1992; Koopman, 1993, 1994; Simmons,2005). The Austromalayan region is crossed bymajor biogeographic boundaries, including theWallace, Lyddeker, and Weber lines. Organ-isms distributed across these various boundar-ies, such as harpyionycterine bats, may con-tribute to the understanding of the complexconnections among the Philippine, Wallacean,and Papuan biogeographic regions and, moregenerally, between Asia and Australia.

In the present study, we explore further themembership and affinities of the expanded

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:02 184 Cust # 582-6

TABLE 1Internal Primers for RAG1 and vWF Genes Designed for This Study

Gene Primer 59–39 sequence

RAG1 RAG1f2 GCCAAGCCCTTCATTGAGACA

RAG1r2 CTATGGAAGGGACTGTCTCAATG

vWF vWFf2 CTTYGTGSTCAGTGGTGTGGA

vWFr2 TCCACACCACTGASCACRAAG

vWFf3 CTATGYCCAGGGCCTGAAGAA

vWFr3 GCTCCTGTTGAAATTGGCCT

184 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

Harpyionycterinae (sensu Giannini et al.,2006) using new molecular data as well aspublished sequences of relevant taxa. We testthe relationship of Boneia and Rousettus toHarpyionycteris as well as the relationship ofAproteles to Dobsonia, and examine theproblem of placing Harpyionycterinae in themegachiropteran tree. Finally, we explore thebiogeographic patterns suggested by phylo-genetic relationships recovered in this study.

METHODS

TAXA

We selected megachiropteran taxa in orderto test monophyly of Harpyionycterinae ascurrently defined (Harpyionycteris, Dobsonia,and Aproteles) and investigate relationshipsof these taxa to other genera of megabats aspreviously proposed in the literature. Specif-ically, we sought to test (1) the association ofBoneia, Rousettus, and Harpyionycteris pro-posed in the description of the latter (Thomas,1896) versus the association of Boneia withRousettus (Andersen, 1912; Bergmans andRozendaal, 1988) and Harpyionycteris withDobsonia (Andersen, 1912; Giannini et al.,2006); (2) the sister relationship of Aprotelesand Dobsonia (Colgan and da Costa, 2002;Giannini and Simmons, 2003, 2005; Menzies,1977; Romagnoli and Springer, 2000); and (3)the inclusion of Dobsonia and their relatives ina pteropodine clade (Romagnoli and Spring-er, 2000) versus a rousettine clade (Andersen,1912; Bergmans, 1997).

Megachiropteran terminals were chosenfrom each main pteropodid clade followingGiannini et al. (2006): Nyctimene albiventerand N. vizcaccia (Nyctimeninae), Cynopterussphinx and Ptenochirus jagori (Cynopteri-nae), Boneia bidens, Eonycteris spelaea, Rou-settus amplexicaudatus, and R. aegyptiacus(Rousettinae), Myonycteris torquata andMegaloglossus woermanni (Epomophorinae,Myonycterini), Epomops franqueti and Epo-mophorus wahlbergi (Epomophorinae, Epo-mophorini), Macroglossus minimus and Syco-nycteris australis (Macroglossinae), Melony-cteris fardoulisi and Melonycteris melanops(Pteropodinae, Notopterini), and Pteropushypomelanus and P. tonganus (Pteropodinae,Pteropodini). Harpyionycterinae megabats

(sensu Giannini et al., 2006) included are:Aproteles bulmerae, Harpyionycteris white-headi, H. celebensis, and a member of eachof the four currently recognized Dobsoniaspecies groups (Simmons, 2005), namelyDobsonia minor (minor species group), inermis(viridis group), D. moluccensis (moluccensisgroup), and D. peronii (peronii group). HereD. magna is considered a junior synonym ofD. moluccensis following Helgen (2007; seealso Byrnes, 2005; Koopman, 1994).

Trees were rooted using Rhinopoma hard-wickii (Rhinopomatidae) and Artibeus jamai-censis (Phyllostomidae). Rhinopomatids arecurrently thought to belong to Rhinolophoi-dea, which is the sister clade of megabats inmultigene phylogenetic analyses, while phyl-lostomids represent the other major clade ofecholocating bats, Yangochiroptera (Eick etal., 2005; Teeling et al., 2005; Miller-Butter-worth et al., 2007).

SEQUENCES

We used sequences of exon 28 of the vonWillebrand factor gene (vWF, 1231 bp) andcytochrome b (cytb, 1140 bp) from Gianniniet al. (2006), and generated new sequences forharpyionycterine species for this study. Wealso incorporated sequences of two othernuclear coding genes, the recombinationactivating gene 1 (RAG1, 1084 bp), and therecombination activating gene 2 (RAG2, 760bp) from Giannini et al. (2008) for the sametaxa, and similarly generated new sequencesfor harpyioncterines. In order to includerelevant species whose samples were notavailable to us, we used published sequencesof the mitochondrial 12S rDNA (12S, ca. 970bp) and Valine tDNA (tVal, ca. 70 bp) andcontributed new sequences for these genes.For additional analyses (see below), weincorporated published sequences of themitochondrial 16S rDNA (16S, ca. 1560 bpfor complete taxa) and NADH dehydroge-nase subunit 2 (ND2, 1044 bp) genes, and theoocyte maturation factor Mos proto-onco-gene (c-mos, ca. 480 bp). Composition of thedata set, accession numbers of all sequencesused, and voucher information for newsequences are given in appendix 1.

Total DNA was obtained from preservedtissue samples with the DNeasy tissue kit

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:03 185 Cust # 582-6

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 185

(QIAGEN). PCR amplification was carriedout using previously published primers (12S-tVal: Springer et al., 1995; cytb: Bastian et al.,2002; RAG1 and RAG2: Teeling et al., 2000;vWF: Porter et al., 1996). To obtain bothforward and reverse sequences for each generegion, additional internal primers weredeveloped for the RAG1 and vWF genes(table 1). All sequences were obtained with anautomated ABI 3730XL sequencer. Sequenceediting and prealignment were done with theSequencher 4.2 software (Gene Codes). Ac-cession numbers for the new sequences areFJ218461-84. Although for some speciesmore than one sample was available, onlyone was included in each analysis. The aligneddataset is provided at ftp://ftp.amnh.org/pub/group/mammalogy/downloads/.

PHYLOGENETIC ANALYSES

Sequences of our coding and ribosomalgenes were submitted to parsimony analysis.In the present study, we used data for thegenes for which we contributed new sequences(i.e., 12S-tVal, cytb, RAG1, RAG2, and vWF)to conduct three principal sets of phylogeneticanalyses: (1) individual gene partitions; (2)nuclear and mitochondrial partitions; and (3)all genes combined. Here we report results ofthe main partitions and the combined set, andcomment on results from the individual geneswhere relevant. In additional analyses, weused genes not sequenced in our study (16S,ND2, and c-mos in individual and combinedsets; see details below).

Analyses involving only coding genes(individual analyses for c-mos, cytb, RAG1,RAG2, vWF, and the nuclear partition) wereperformed under conventional parsimony(static alignment). The tree search strategyconsisted of 500 replicates of random addi-tion sequences of taxa (RAS) each followedby tree bisection reconnection branch swap-ping (TBR). An additional round of TBRwas done on optimal trees so obtained. Cladesupport was assessed using Bremer or decayvalues (Bremer, 1994) and character resam-pling (Goloboff et al., 2003). We calculatedBremer values via incremental sampling ofsuboptimal trees (see Giannini and Bertelli,2004). Briefly, we saved up to 2000 subopti-mal trees one step longer than the previous

optimum in successive stages. That is, we firstsearched for suboptimal trees 1 step longerthan the optimal tree length, next savingsuboptimals at every step from 2 through 10steps longer than the optimal trees. Second,group frequency (based on unbiased sym-metric resampling; Goloboff et al., 2003) wascalculated on the basis of 5000 replications.These analyses were executed in TNT 1.0(Goloboff et al., 2004, 2008).

For all analyses involving ribosomal andtransfer RNA genes (i.e., individual 12S-tValpartition, the mitochondrial partition, thecombined set, and additional analyses using16S), we used direct optimization (DO;Wheeler, 1996). In this approach, alignmentis viewed as dynamic (i.e., varying acrosstopologies in tree space) and is thereforelinked to tree search. The transformation costof whole unaligned sequences on a topology iscalculated via optimization. Costs of eachtransformation type (indel or substitution)were established a priori—as in every align-ment procedure. According to these costs, thelength of the tree is calculated as the sum oftransformations on each internal node in thedownpass optimization, which minimizesindels and substitutions (steps) of wholesequences on the candidate tree. Minimum-length trees are chosen among those visited.In practice, this is done by way of treebuilding and branch swapping. We set equalcosts of indels and substitutions, and oursearch strategy consisted of replicated swap-ping + refinements (e.g., Giannini and Sim-mons, 2003, 2005). Briefly, 100 RAS + TBRbranch swapping were run, collecting alloptimal trees from each replication. Thosetrees were submitted to tree fusing (Goloboff,1999). In analyses combining ribosomal andcoding genes, the latter were treated asprealigned; i.e., no indels were allowed incoding genes (static alignment). All DOanalyses, including searches to calculateBremer support values, were performed usingPOY 4 (Varon et al., 2007).

Additional phylogenetic analyses were con-ducted to take advantage of additional pub-lished data. First, we added published 16S andc-mos sequences to our combined set toprovide a ‘‘total-evidence’’ analysis based onsequences available for the majority of ourtaxonomic samples. Second, we ran a specific

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:03 186 Cust # 582-6

186 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

analysis of ND2 with the few sequences ofmegabats available in GenBank (which un-fortunately had low overlap with our taxo-nomic sampling) to provide a separate test ofthe affinities of Boneia. Data available includ-ed ND2 sequences of Dobsonia minor amongthe harpyionycterines and also Boneia bidens.This analysis was important because our dataon Boneia originated from a single individual(ZMA 23100) from which we extracted DNAand generated all sequences, whereas the ND2sequences originated from DNA that wasamplified from a different individual (Gen-Bank accession number AY504581).

In addition to the parsimony analyses, weperformed a maximum-likelihood (ML) treesearch for the combined nuclear data set usinga GTR + C + I partitioned model (i.e.,parameters were estimated from the data foreach gene separately) using the programRAxML (Stamatakis, 2006; Stamatakis et al,2008). Our search strategy consisted of execut-ing 100 rapid bootstrap inferences and there-after a thorough ML search. Statistical supportwas obtained with 100 bootstrap replications.

BIOGEOGRAPHIC ANALYSIS

Our biogeographic analysis consisted ofmapping onto the optimal tree obtained fromour total-evidence phylogenetic analysis asingle biogeographic character composed ofsix states, each of which consisted of knowndistribution areas of terminals (see below). Weperformed a partial (downpass-only) Fitch(1971) optimization in order to recover thepattern of minimal biogeographic events re-quired by the tree. In this case, we did notintend to provide wide coverage of Megachir-optera; our aim was instead to identify themain dispersal-vicariance events of our in-group, the Harpyionycterinae—those eventsthat led to the origin of currently recognizedgenera. This attempt is necessarily incompleteas we did not include all Dobsonia species.Therefore we restricted our analysis to thesampled harpyionycterines and their areas ofoccurrence, which were coded in a singlemultistate, unordered character whose statesconsisted of broad biogeographic subregionsof the eastern Austromalayan region asrecognized by Corbet and Hill (1992): Sula-wesi (SUL), the Philippines (PHI), Lesser

Sunda Islands (LSI), the Moluccas (MOL),New Guinea (NGU), Australia (AUS), andthe Solomon Islands (SOL). Taxa occurringin more than one area were scored aspolymorphic (for instance, Dobsonia minorwas scored NGU/SUL).

RESULTS

NUCLEAR PARTITION

Parsimony analysis of nuclear data (RAG1,RAG2, and vWF sequences) resulted in 10most-parsimonious trees of 1922 steps (strictconsensus tree in fig. 1). In this tree, Mega-chiroptera is highly supported (Bremer sup-port value, BS 5 38). Mutually monophyleticNyctimeninae and Cynopterinae form a cladethat corresponds with Andersen’s (1912)Cynopterus section, and is sister to a poorlysupported multichotomy of four clades con-taining all other megachiropterans. The fourgroups are: a reduced macroglossine clade(Macroglossus + Syconycteris); a reducedpteropodine clade including Melonycterisand Pteropus (see Bergmans, 1997; Gianniniet al., 2008); a clade composed of rousettine(Rousettus and Eonycteris) and epomophor-ine megabats, including Myonycterini (Mega-loglossus and Myonycteris) and Epomophor-ini (Epomophorus and Epomops); and aharpyionycterine clade (sensu Giannini etal., 2006) that includes Boneia as sister toHarpyionycteris, and both sister to Dobsonia.Our ML analysis (fig. 2) recovered the samemajor groupings as the parsimony analysis(fig. 1), resolving the backbone polytomy ofthe parsimony result (cf. fig. 1) with varyingdegrees of support (bootstrap values 32–82).This topology is identical to one of theoptimal topologies obtained under parsimony(i.e., in this analysis, parsimony is conserva-tive in reflecting conflicts in the data).Regarding harpyionycterine megabats, thebranching pattern is the same in both MLand MP analyses (cf. figs. 1 and 2).

MITOCHONDRIAL PARTITION

The analysis of the 12S-tVal + cytbsequences under direct optimization pro-duced a single tree at 3220 steps (fig. 3).Megachiroptera is highly supported (BS 5

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:03 187 Cust # 582-6

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 187

36), but much of the backbone of themegachiropteran subtree is not very wellsupported (2 # BS # 6). Melonycterisappears as sister taxon to Nyctimene, Syco-nycteris, and a large group composed of theremainder of megachiropterans. The latter isfurther subdivided into a clade containing(Pteropodini + Cynopterinae), the previouslyreported rousettine-epomophorine clade, andanother version of the harpyionycterine cladeincluding Aproteles, Boneia, Dobsonia, Har-pyionycteris, and, surprisingly, Macroglossus.In this clade, all intergeneric relationships areweakly supported (1 # BS # 2).

COMBINED AND ADDITIONAL ANALYSES

A combined parsimony analysis usingboth the nuclear genes (RAG1, RAG2, andvWF) and our mitochondrial data (cytb and12S-tVal) resulted in three trees of 4451 steps(strict consensus in fig. 4). In this analysis,Megachiroptera is highly supported (BS 5

97). Melonycteris is sister to a trichotomyincluding Macroglossus, Syconycteris, and apoorly supported (BS 5 3) group containingthe remainder of terminals, which is furthersubdivided into a harpyionycterine clade((Harpyionycteris + Boneia) + (Aproteles +

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:03 188 Cust # 582-6

Fig. 1. Strict consensus tree resulting from a parsimony analysis of three nuclear coding genescombined (RAG1, RAG2, and vWF). Bremer support values are given above branches and resampling(jackknife) frequencies are given below branches.

188 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

Dobsonia)), and another group containingthe rousettine-epomophorine clade and acomposite clade (Pteropodini (Nyctimeninae+ Cynopterinae)).

Adding to this data set published sequenc-es of the mitochondrial 16S and the nuclearoncogene c-mos genes produces a single treeof 6119 steps (fig. 5). In this tree, Nyctimeneis sister to all other megachiropterans, andthis tree partition appeared highly supported(BS 5 19). Successive clades recovered were(Pteropodini + Cynopterinae), the rousettine-epomophorine clade, and a highly supportedharpyionycterine clade (BS 5 23) in which

Dobsonia was sister to Aproteles, Boneia, andHarpyionycteris. Several important dispersal-vicariance events are suggested by mappingdistribution of terminals onto the optimaltree of our combined phylogenetic analysis.These patterns are discussed below.

Finally, we obtained four trees of 1717steps in a separate analysis based on ND2sequences from pteropodid species availableon GenBank. The strict consensus tree(fig. 6) was generally poorly supported butplaced Boneia bidens and Dobsonia minor assister taxa. This result is compatible withthose of previous analyses that separately

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:06 189 Cust # 582-6

Fig. 2. Maximum likelihood tree resulting from the analysis of the nuclear dataset (RAG1, RAG2, andvWF; gene partitions modeled using GTR + C + I; ML score -11,136.227759). Values near nodes representbootstrap estimates of clade support based on 100 replications.

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 189

support an association of Boneia with har-pyionycterine megabats—D. minor amongthe latter.

DISCUSSION

PHYLOGENETIC RELATIONSHIPS

AND CLASSIFICATION

The phylogenetic analyses of nuclear datarecovered a harpyionycterine clade composed

of Harpyionycteris, Aproteles, Dobsonia, and,rather unexpectedly, Boneia. Although Mac-roglossus was added to this group in theanalysis of the mitochondrial partition (withnegligible support, 1 # BS # 2; fig. 3), thegroup membership of the nuclear analysis wasrecovered in the total-evidence analysis withhigh support (BS 5 23; fig. 5). This representsstrong evidence that Boneia is not a synonymof Rousettus but a distinct genus that belongsin Harypionycterinae. We also confirmed the

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:11 190 Cust # 582-6

Fig. 3. Single optimal tree resulting from a direct optimization, parsimony analysis of threemitochondrial genes combined (12S-tVal + cytb). Bremer support values are given below branches.

190 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

inclusion of Aproteles in this group (seediscussion in Giannini et al., 2006). Thus,we formally expand Harpyionycterinae Mil-ler, 1907, to include Boneia Jentink, 1879,Harpyionycteris Thomas, 1896, DobsoniaPalmer, 1898, and Aproteles Menzies, 1977.

Resolving the composition of Harpyionyc-terinae has proven difficult throughout thehistory of megachiropteran systematics.Thomas (1896) recognized the difficulties offinding allies of Harpyionycteris among the

megachiropterans known at the time of itsdescription. Thomas (1896: 243) attributedthe similarities in canine morphology withHarpyia (5 Nyctimene) to either homoplasy(‘‘accident’’ in his words) or plesiomorphy(i.e., ‘‘common descent from the (presum-ably) cuspidate-toothed ancestor of the Pter-opodidae’’). Thomas (1896: 243) suggestedthat ‘‘On the whole … [Harpyionycteris] maybe most conveniently placed near Xanthar-pyia [5 Rousettus] and Boneia, with which it

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:16 191 Cust # 582-6

Fig. 4. Strict consensus tree of a parsimony analysis of RAG1, RAG2, vWF, cytb and 12S-tVal genescombined. Bremer support values are given below branches.

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 191

shares certain external characters, an indicalclaw, and the cheek-tooth formula of P. 3/3,M. 2/3.’’ The index finger claw and the cheek-tooth formula are likely plesiomorphic forMegachiroptera (see Giannini and Simmons,2005, 2007), and the external characters werenot specified (but see the external resem-blance of Harpyionycteris celebensis andBoneia bidens in fig. 7). Nevertheless, it seemsclear that Thomas (1896) did correctlyperceive a link between these forms.

Despite Thomas’s (1896) intuition and themolecular support discovered in the currentstudy, to our knowledge, no author subse-quent to Thomas considered a subfamily ortribal group inclusive of both Harpyionycterisand Boneia. This is reflected in a completelack of potential morphological synapomor-phies for such grouping in the literature. Inrecent systematic studies involving Boneia,discussion has centered on the nature of itsrelationship with Rousettus (e.g., Bergmans,

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:20 192 Cust # 582-6

Fig. 5. Single optimal tree of a parsimony analysis of RAG1, RAG2, vWF, cytb, 12S-tVal, 16S, and c-mos genes combined. Bremer support values are given below branches.

192 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

1994; Bergmans and Rozendaal, 1988). Thisis not surprising given the prominent rolethat Rousettus has played in megachiropteransystematics as a result of long being consid-ered close to the ancestral form of megabatsfollowing Andersen (1912). Although thisassessment has been refuted by all cladisticstudies since the early 1990s (e.g., Colgan andda Costa, 2002; Giannini and Simmons,2003, 2005; Kirsch et al., 1995; Romagnoliand Springer, 2000; Springer et al., 1995),most megachiropteran genera have beencompared with Rousettus on these grounds,and Boneia was no exception. Given the lackof unusual external features, unremarkableskull shape, dental formula close to thegeneralized condition of megabats, and thehabit of roosting gregariously in caves,Boneia was considered so close an ally ofRousettus as to be included in Rousettus as asubgenus (see Bergmans and Rozendaal,1988). The morphological evidence discussedby the authors in favor of this associationincludes both morphometric and qualitativedata and is rather robust, which makes ourfinding of a Boneia-Harpyionycteris clademore surprising. As shown in fig. 8, theskulls of Boneia and Harpyionycteris bearlittle mutual resemblance. In Boneia, the

rostrum is thin and strongly deflected relativeto the basicranial axis, the dentition issimplified but otherwise typically pteropodid,the zygomatic arch is weak, and in overallappearance it resembles Rousettus. By con-trast, Harpyionycteris shows a strong ‘‘dob-sonine’’ skull with the most peculiar rostrumand dentition seen in megabats (see com-ments in Giannini et al., 2006).

However difficult it may seem to reconcilethe current composition of the harpyionyc-terine clade with the morphological variationtraditionally considered in megabat system-atics, our phylogeny does suggest someinteresting scenarios of morphological andecological evolution. One particularly signif-icant example involves the dentition. Har-pyionycterines exhibit dental formulae thatvary across taxa, but are similar in the sensethat one formula can be parsimoniouslytransformed into any other in one-step lossesor gains. A most relevant case is the firstupper incisor, which is the sole tooth missingfrom the generalized pteropodid formula inBoneia (a variably deciduous tooth in thistaxon; Bergmans and Rozendaal, 1988). Thisis also the incisor demonstrably lacking inDobsonia and hypothesized to be absent inHarpyionycteris (see Giannini and Simmons,2007: table 2). Aproteles lacks both upperincisors. Our phylogeny, once it is acceptedas a supported hypothesis, allows us tosuggest that the loss or reduction of I1 iseffectively homologous across harpyionycter-ine taxa, and that the loss of incisors inAproteles likely happened sequentially (I1first lost first, followed by I2).

BIOGEOGRAPHIC AND ECOLOGICAL PATTERNS

Harpyionycterinae sensu Giannini et al.(2006; i.e., composed of Harpyionycteris,Dobsonia, and Aproteles), expanded here toinclude Boneia, comprises 18 species whosejoint distribution encompasses the Philippine,Wallacean, and Papuan subregions of theAustromalayan area, plus two species thatnarrowly escape the biogeographic boundar-ies of those subregions (fig. 9; table 2). Thelatter two taxa are D. peronii, which alsooccurs in Bali (Sundaic subregion) and D.moluccensis, which also occurs in Queens-land, Australia (Corbet and Hill, 1992;

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:24 193 Cust # 582-6

Fig. 6. Strict consensus tree of a parsimonyanalysis of the ND2 gene. Bremer and jackknifevalues (above and below branches, respectively)are reported for the harpyionycterine clade recov-ered in this analysis. Numbers in parenthesisindicate number of terminals included in thecynopterine and pteropodine clades recovered.

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 193

Simmons, 2005). Keeping in mind that ourbiogeographic interpretation rests upon (a)our limited taxonomic sampling (cf. table 1),and (b) the optimal tree from our total-evidence analysis, our biogeographic analysisreconstructed the Papuan subregion as theunambiguous area of origin of harpyionyc-terines (fig. 9). From this region, the ancestorof Harpyionycteris and Boneia crossed theLyddekker’s and Weber’s lines westward to

establish populations and speciate in Sula-wesi, where differentiation of the two generaoccurred. Later, an ancestral Harpyionycterisspecies split into a lineage that colonized thePhilippines (crossing the Wallace line north-ward) giving rise to H. whiteheadi, and intoanother lineage that remained in Sulawesi tobecome H. celebensis.

The dispersal-vicariance events (sensuRonquist, 1997) that lead to the present-day

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:57:25 194 Cust # 582-6

Fig. 7. External appearance of live specimens of Harpyionycteris celebensis (top) and Boneia bidens(bottom). Photos: courtesy Jan Haft (with permission).

194 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

diversity within Dobsonia can only be matterof speculation given our limited sample (butsee Byrnes, 2005, for a detailed account). Inour analysis, the genus also originated inNew Guinea. In a more conventional scenar-io in which Aproteles is sister to Dobsonia asin the tree in fig. 4, the area of the ancestralharpyionycterine bat is reconstructed asSulawesi + New Guinea. Accepting a Papuan(or Papuan-Wallacean) origin as probable(fig. 9), and taking together the distributionof all species, several Dobsonia lineagesapparently diverged in various directions:northeastward to colonize the Bismarck

Archipelago and the Solomon Islands; south-ward to reach Cape York Peninsula (Aus-tralia), and westward to reach the Moluccas,Lesser Sunda Islands, Sulawesi, and beyondthe Wallace line in Bali (populations of D.peronii) and the Philippines (D. chapmani,restricted to Negros and Cebu Is.; Alcala etal. 2004; Heaney et al., 1998; Paguntalan etal., 2004). Other speciation events took placein land-bridge islands of New Guinea (By-rnes, 2005).

The optimal tree from our total-evidenceanalysis also suggests some interesting eco-logical patterns including insights into the

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:06 195 Cust # 582-6

Fig. 8. Lateral (upper row) and ventral (lower row) views of the skulls of Rousettus amplexicaudatus(A, AMNH 226391), Boneia bidens (B, AMNH 254542), and Harpyionycteris whiteheadi (C, AMNH142766). Scale 5 5 mm.

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 195

evolution of roosting habits (see also discus-sion in Byrnes, 2005). Roost availability is oneof the limiting factors of bat populations(Kunz and Pierson, 1994). Megachiropteransare either cavity dwellers (occupying struc-tures from tree holes to large caves) or foliagedwellers (a habit that varies from takingsolitary refuge on leaves or vines to living inlarger groups in actively defoliated trees;Kunz and Pierson, 1994). The roosting habitsof Harpyionycteris are unknown, but thisgenus is nested within successive sister cladesthat include many typical cave-dwelling bats:Boneia (Bergmans and Rozendaal, 1988),Aproteles (Flannery, 1995; Menzies, 1977),and several Dobsonia species including D.moluccensis (Bonaccorso, 1998; Flannery,1995; Helgen, 2007), and probably also D.peronii (Chiroptera Specialist Group, 1996)and D. inermis. By contrast, D. minor hasbeen reported to roost both in foliage(Flannery, 1995) and in caves (Boeadi andBergmans, 1987). The prediction we drawfrom reconstructing roosting habits (foliage

vs. cavities) in the optimal tree of our mostcomplete analysis is that Harpyionycteriscould be also a cavity dweller, as are manymembers of the harpyionycterine clade(fig. 10). It is interesting that the harpyionyc-terine clade appeared in our analysis as sisterto the rousettine-epomophorine clade (withhigh support in our analysis, BS 5 21). Takentogether, these two major megabat cladesshare a common ancestor that is also recon-structed as a cavity dweller (fig. 10). Onceagain, limited taxonomic sampling preventsfurther speculation, but we risk the hypothesisthat cavity dwelling may have had a majorimpact in the diversification of megabats.

CONCLUSIONS

Our results strongly confirm monophyly ofa harpyionycterine clade inclusive of Har-pyionycteris, Dobsonia, Aproteles, and Boneia(the latter previously included in Rousettus asa subgenus). Accordingly, we recognizeBoneia as a distinct genus and recommend

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:11 196 Cust # 582-6

TABLE 2Distribution of Each of the 18 Species of Harpyionycterinae Miller, 1907

Biogeographic regions follow Corbet and Hill (1992).

Species Distribution Biogeographic regions

Aproteles bulmerae Papua New Guinea (highlands) Papuan

Boneia bidens Northern Sulawesi Wallacean (Sulawesi Division)

Dobsonia anderseni Bismarck Arch., Admiralty Isls. Papuan

Dobsonia beauforti Waigeo, Batanta, Salawati, Gebe,

Gag, and Biak Isls.

Papuan, Wallacean (Moluccan Division)

Dobsonia chapmani Cebu and Negros Isls. Philippines

Dobsonia crenulata N Moluccas, Togian, Sangihe,

Talaud, and Peleng Isls., Sulawesi

Wallacean (Sulawesi and Moluccan Divisions)

Dobsonia emersa Biak, Numfoor, and Owii Isls. Papuan

Dobsonia exoleta Sulawesi, Muna, Togian, Sula Isls. Wallacean (Sulawesi Division)

Dobsonia inermis Solomon Islands Solomons

Dobsonia minor Lowlands of New Guinea and

adjacent Isls., Sulawesi

Papuan, Wallacean (Sulawesi Division)

Dobsonia moluccensis Moluccas, New Guinea and

land-bridge islands, to Queensland

Papuan, Wallacean (Moluccan Division),

Australian

Dobsonia pannietensis Louisiade Arch., D’Entrecasteaux

Isls., Trobriand Isls.

Papuan

Dobsonia peronii Lesser Sunda Islands (Bali East to

Babar Isls.)

Sundaic (Javan Division) and Wallacean

(Lesser Sunda and Moluccan Divisions)

Dobsonia praedatrix Bismarck Arch. Papuan

Dobsonia viridis C and S Moluccas, Banda, and

Kai Isls.

Wallacean (Moluccan Division)

Harpyionycteris celebensis Sulawesi Wallacean (Sulawesi Division)

Harpyionycteris whiteheadi Philippines Philippine

196 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:12 197 Cust # 582-6

Fig. 9. Joint distribution of species in harpyionycterine megabat genera in the Austromalayan Region(A), and biogeographic patterns in harpyionycterine megabats (B). References for distributions: Aproteles(dots), Boneia (hatched area), Dobsonia (perimeter lines), and Harpyionycteris (shaded areas). Indicated arethe Wallace, Weber, and Lydekker lines. Arrows indicate the distribution of Dobsonia species that escapethe boundaries of the Wallacean and Papuan subregions: (A) Dobsonia chapmani in Negros and CebuIslands (Philippine subregion); (B) Dobsonia peronii in Bali (Sundaic subregion, Javan Division; D. peroniiis widespread in the Lesser Sunda Islands); (C) Dobsonia moluccensis in Queensland, Australia (D.moluccensis, inclusive of D. magna, is widespread in the Moluccas and New Guinea). Biogeographicpatterns: Each species is assigned one or more of the following areas, each of which is a state in a singleunordered biogeographic character: AUS Australia, LSI Lesser Sunda Islands, NGU New Guinea, PHIPhilippines, SOL Solomon Islands, SUL Sulawesi. Biogeographic events are inferred by partial (downpassonly) optimization of distribution areas of species in the harpyionycterine clade in the optimal tree fromthe most complete, combined analysis (B; see fig. 4).

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 197

expanding Harpyionycterinae Miller, 1907,to include these four genera. Although nomorphological synapomorphies are apparentfor this grouping, some morphological andecological patterns are interesting, for in-stance, dental-formula patterns and theprediction that Harpyionycteris may be acavity-dwelling bat. Also inferred from our

phylogenetic hypothesis, we showed thatmajor biogeographic boundaries within theAustromalayan region play a key role inexplaining patterns of dispersal-vicarianceevents in harpyionycterine bats. These resultsand preliminary interpretations encouragethe undertaking of a full-scale phylogeny ofMegachiroptera.

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:15 198 Cust # 582-6

Fig. 10. Reconstruction of roosting habits (cavity versus foliage) in the most parsimonious tree fromthe total dataset of this study (fig. 4). Hatched lines indicate ambiguity (in this case, terminals for whichboth roosting habits were documented). Cavity dwelling is indicated in thick branches. Question marksindicate lack of data for Harpyionycteris. Note that this genus is nested in a clade of cavity dwellers(see text).

198 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

ACKNOWLEDGMENTS

We are indebted with Rob DeSalle (Sack-ler Institute for Comparative Genomics,American Museum of Natural History) forproviding laboratory space and reagents forthe molecular part of this work. We thankLawrence Heaney (Field Museum of NaturalHistory, Chicago), Jim Patton and CarlaCicero (Museum of Vertebrate Zoology,Berkeley), John Wible and Suzanne McLaren(Carnegie Museum, Pittsburgh), DenisO’Meally (Australian Museum), Jeremy Ja-cobs, Louise Emmons, and James Mead(Smithsonian Institution, National Museumof Natural History, Washington DC), andWim Bergmans (ZMA Zoologisch Museum,Amsterdam) for access to tissue samples thatmade possible this contribution. We alsothank Jan Haft for photographs an addition-al information. Finally, our appreciation toDeanna P. Byrnes for her generosity inproviding two unpublished sequences foruse in this study. Funding for this reportwas provided by the National Science Foun-dation (research grant DEB-9873663 toN.B.S.), Coleman and Vernay postdoctoralfellowships at the AMNH to N.P.G., aHenry MacCracken doctoral fellowship atNew York University to F.C.A., and aVernay postdoctoral fellowship at theAMNH to F.C.A.

REFERENCES

Alcala, E.L., R.B. Paalan, L.T. Averia, and A.C.

Alcala. 2004. Rediscovery of the Philippine

bare-backed fruit bat (Dobsonia chapmani Ra-

bor) on southwestern Negros Island, the Phil-

ippines. Silliman Journal 45(2): 126–136.

Andersen, K. 1912. Catalogue of the Chiroptera in

the collection of the British Museum. London:

Megachiroptera British Museum (Natural His-

tory).

Bastian, S.T., Jr., K. Tanaka, R.V. Anunciado,

N.G. Natural, A.C. Sumalde, and T. Nami-

kawa. 2002. Evolutionary relationships of flying

foxes (genus Pteropus) in the Philippines in-

ferred from DNA sequences of cytochrome b

gene. Biochemical Genetics 40: 101–116.

Bergmans, W. 1994. Taxonomy and biogeography

of African fruit bats (Mammalia, Megachirop-

tera). 4. The genus Rousettus Gray, 1821.

Beaufortia 44(4): 79–126.

Bergmans, W. 1997. Taxonomy and biogeography

of African fruit bats (Mammalia, Megachirop-

tera). 5. The genera Lissonycteris Andersen,

1912, Myonycteris Matschie, 1899, and Mega-

loglossus Pagenstecher, 1885. General remarks

and conclusions. Annex: key to all species.

Beaufortia 47: 11–90.

Bergmans, W., and F.G. Rozendaal. 1988. Notes

on collections of fruit bats from Sulawesi and

some off-lying islands (Mammalia, Megachir-

optera). Zoologische Verhandelingen 248: 1–74.

Boeadi, B., and W. Bergmans. 1987. First record

of Dobsonia minor (Dobson, 1879) from Sula-

wesi, Indonesia (Mammalia, Megachiroptera).

Bulletin Zoologisch Museum Universiteit van

Amsterdam 11: 69–74.

Bonaccorso, F.J. 1998. Bats of Papua New

Guinea. Washington, DC: Conservation Inter-

national.

Bremer, K. 1994. Branch support and tree

stability. Cladistics 10: 295–304.

Byrnes, D.G.P. 2005. Evolution among the bare-

backed fruit bats, Dobsonia (Chiroptera, Pter-

opodidae). Unpublished Ph.D. dissertation.

University of Wisconsin, Madison.

Chiroptera Specialist Group, Dobsonia peroni. In

International Union for Conservation of Nature

and Natural Resources 2007. 2007 IUCN red

list of threatened species (downloaded 21

August 2008 from http://www.iucnredlist.org/).

Colgan, D.J., and P. da Costa. 2002. Megachir-

opteran evolution studied with 12S rDNA and

c-mos DNA sequences. Journal of Mammalian

Evolution 9: 3–22.

Corbet, G.B., and J.E. Hill. 1992. Mammals of the

Indomalayan region: a systematic review. Lon-

don: Oxford University Press. 2 vols, 488 pp.

Eick, G.N., D.S. Jacobs, and C.A. Matthee. 2005.

A nuclear DNA phylogenetic perspective on the

evolution of echolocation and historical bioge-

ography of extant bats (Chiroptera). Molecular

Biology and Evolution 22: 1869–1886.

Flannery, T.F. 1995. Mammals of Papua New

Guinea. Sydney: Robert Brown and Associa-

tion, 439 pp.

Fitch, W.M. 1971. Toward defining the course of

evolution: minimum change for a specific tree

topology. Systematic Zoology 20: 406–416.

Giannini, N.P., F.C. Almeida, N.B. Simmons, and

R. DeSalle. 2006. Phylogenetic relationships of

the enigmatic Harpy Fruit Bat, Harpyionycteris

(Mammalia: Chiroptera: Pteropodidae). Amer-

ican Museum Novitates 3533: 1–12.

Giannini, N.P., F.C. Almeida, N.B. Simmons, and

K.M. Helgen. 2008. The systematic position of

Pteropus leucopterus and its bearing on the

monophyly and relationships of Pteropus (Chi-

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:20 199 Cust # 582-6

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 199

roptera: Pteropodidae). Acta Chiropterologica10: 11–20.

Giannini, N.P., and S. Bertelli. 2004. A phylogenyof extant penguins based on integumentarycharacters. Auk 121: 422–434.

Giannini, N.P., and N.B. Simmons. 2003. Aphylogeny of megachiropteran bats (Mammalia:Chiroptera: Pteropodidae) based on directoptimization analysis of one nuclear and fourmitochondrial genes. Cladistics 19: 496–511.

Giannini, N.P., and N.B. Simmons. 2005. Conflictand congruence in a combined DNA-morphol-ogy analysis of megachiropteran bat relation-ships (Mammalia: Chiroptera: Pteropodidae).Cladistics 21: 411–437.

Giannini, N.P., and N.B. Simmons. 2007. ElementHomology and the Evolution of Dental For-mulae in Megachiropteran Bats (Mammalia:Chiroptera: Pteropodidae). American MuseumNovitates 3559: 1–27.

Goloboff, P.A. 1999. Analyzing large data sets inreasonable times: solutions for composite opti-ma. Cladistics 15: 415–428.

Goloboff, P.A., J.S. Farris, M. Kallersjo, B.Oxelman, M. Ramırez, and C. Szumik. 2003.Improvements to resampling measures of groupsupport. Cladistics 19: 324–332.

Goloboff, P.A., J.S. Farris, and K. Nixon. 2004.T.N.T.: tree analysis using new technologies.Program and documentation (available at www.cladistics.org).

Goloboff, P.A., J.S. Farris, and K.C. Nixon. 2008.TNT, a free program for phylogenetic analysis.Cladistics 24(5): 774–786.

Heaney, L.R., D.S. Balete, M.L. Dolar, A.C.Alcala, A.T.L. Dans, P.C. Gonzales, N.R. Ingle,M.V. Lepiten, W.L.R. Oliver, P.S. Ong, E.A.Rickart, B.R. Tabaranza, Jr., and R.C.B.Utzurrum. 1998. A synopsis of the mammalianfauna of the Philippine Islands. FieldianaZoology New Series 88: 1–61.

Helgen, K.M. 2007. A taxonomic and geographicoverview of the mammals of Papua. In A.J.Marshall and B.M. Beehler (editors), Ecology ofIndonesia series. Vol. 6. Part 1, The ecology ofPapua: 689–749. Singapore: Periplus Editions.

Jones, K.E., A. Purvis, A. MacLarnon, O.R.P.Bininda-Emonds, and N.B. Simmons. 2002. Aphylogenetic supertree of the bats (Mammalia:Chiroptera). Biological Review 77: 223–259.

Kirsch, J.A., T.F. Flannery, M.S. Springer, andF.-J. Lapointe. 1995. Phylogeny of the Pter-opodidae (Mammalia: Chiroptera) based onDNA hybridization with evidence of bat mono-phyly. Australian Journal of Zoology 43:395–428.

Koopman, K.F. 1993. Order Chiroptera. In D.E.Wilson and D.M. Reeder (editors), Mammal

species of the world: a taxonomic and geo-graphic reference. 2nd ed.: 137–241. Washing-ton, DC: Smithsonian Institution Press.

Koopman, K.F. 1994. Chiroptera: systematics. InHandbook of zoology, vol. 8, part 60, Mam-malia: 1–217. Berlin: Walter de Gruyter.

Kunz, T.H., and E.D. Pierson. 1994. Bats of theworld: an introduction. In R.M. Nowak (edi-tor), Walker’s bats of the World: 1–46. Balti-more, MD: Johns Hopkins University Press.

Menzies, J.I. 1977. Fossil and subfossil bats fromthe mountains of New Guinea. AustralianJournal of Zoology 25: 329–336.

Miller, G.S., Jr. 1907. The families and genera ofbats. Bulletin of the United States NationalMuseum 57: 1–282.

Miller, G.S., Jr., and N. Hollister. 1921. Twentynew mammals collected by H. Raven in Celebes.Proceedings of the Biological Society of Wash-ington 34: 93–104.

Miller-Butterworth, C.M., W.J. Murphy, S.J.O’Brien, D.S. Jacobs, M.S. Springer, and E.C.Teeling. 2007. A family matter: conclusiveresolution of the taxonomic position of thelong-fingered bats, Miniopterus. Molecular Bi-ology and Evolution 24: 1553–1561.

Paguntalan, L.M.J., M.dG. Pedregosa, and M.J.C.Gadiana. 2004. The Philippine bare-backed fruitbat Dobsonia chapmani Rabor, 1952: rediscov-ery and conservation status on Cebu Island.Silliman Journal 45(2): 114–122.

Porter, C.A., M. Goodman, and M.J. Stanhope.1996. Evidence on mammalian phylogeny fromsequences of exon 28 of the von Willerbrandfactor gene. Molecular Phylogenetics and Evo-lution 5: 89–101.

Romagnoli, M.L., and M.S. Springer. 2000.Evolutionary relationships among Old Worldfruitbats (Megachiroptera: Pteropodidae) basedon 12S rRNA, tRNA valine and 16S rRNAgene sequences. Journal of Mammalian Evolu-tion 7: 259–284.

Ronquist, F. 1997. Dispersal–vicariance analysis: anew approach to the quantification of historicalbiogeography. Systematic Biology 46: 195–203.

Simmons, N.B. 2005. Order Chiroptera. In D.E.Wilson and D.M. Reeder (editors), Mammalspecies of the world: a taxonomic and geo-graphic reference, 3rd ed.: 312–529. Baltimore,MD: Johns Hopkins University Press.

Slaughter, B.H. 1970. Evolutionary trends inchiropteran dentitions. In B.H. Slaughter andW.D. Walton (editors), About bats: 51–83.Dallas, TX: Southern Methodist UniversityPress.

Springer, M.S., L.J. Hollar, and J.A. Kirsch. 1995.Phylogeny, molecules versus morphology andrates of character evolution among fruit bats

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:20 200 Cust # 582-6

200 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

(Chiroptera: Megachiroptera). Australian Jour-nal of Zoology 43: 557–582.

Stamatakis, A. 2006. RAxML-VI-HPC: Maxi-mum likelihood based phylogenetic analyseswith thousands of taxa and mixed models.Bioinformatics 22: 2688–2690.

Stamatakis, A., P. Hoover, and J. Rougemont.2008. A rapid bootstrap algorithm for theRAxML web servers. Systematic Biology57(5): 758–771.

Tate, G.H.H. 1951. Harpyionycteris, a genus of rarefruit bats. American Museum Novitates 1522:1–9.

Teeling, E.C., M. Scally, D. Kao, M. Romagnoli,M.S. Springer, and M.J. Stanhope. 2000. Mo-lecular evidence regarding the origin of echolo-cation and flight in bats. Nature 403: 188–192.

Teeling, E.C., M.S. Springer, O. Madsen, P. Bates,S.J. O’Brien, and W.J.A. Murphy. 2005. Amolecular phylogeny for bats illuminates bioge-ography and the fossil record. Science 307:580–584.

Thomas, O. 1896. Mammals from Celebes, Bor-neo, and the Philippines recently received at theBritish Museum. Annals and Magazine ofNatural History 18(6): 241–250.

Varon, A., L.S. Vinh, I. Bomash, and W.C.Wheeler. 2007. POY 4.0 Beta 2398. New York:American Museum of Natural History (avail-able at http://research.amnh.org/scicomp/projects/poy.php)

Wheeler, W.C. 1996. Optimization alignment: theend of multiple sequence alignment in phyloge-netics? Cladistics 12: 1–9.

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:20 201 Cust # 582-6

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 201

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:20 202 Cust # 582-6

AP

PE

ND

IX1

Vo

uch

erin

form

ati

on

an

dG

enB

an

ka

cces

sio

nn

um

ber

so

fm

egab

at

spec

imen

su

sed

inth

ep

rese

nt

stu

dy

,li

sted

as

ap

pea

rin

gin

the

tree

of

figu

re4.

Imp

reci

selo

cali

ties

are

qu

ote

d.

Ab

bre

via

tio

ns

of

Inst

itu

tio

ns:

AM

CC

,A

mb

rose

Mo

nel

lC

ryo

Co

llec

tio

n(A

MN

H);

AM

NH

,A

mer

ican

Mu

seu

mo

fN

atu

ral

His

tory

,N

ewY

ork

;C

MN

HC

arn

egie

Mu

seu

mo

fN

atu

ral

His

tory

,P

itts

bu

rgh

;E

BU

Evo

luti

on

ary

Bio

logy

Un

it,

Au

stra

lia

nM

use

um

,S

yd

ney

;F

MN

H,

Fie

ldM

use

um

of

Na

tura

lH

isto

ry,

Ch

ica

go

;M

VZ

,M

use

um

of

Ver

teb

rate

Zo

olo

gy,

Un

iver

sity

of

Cali

forn

ia,

Ber

kel

ey;

US

NM

Sm

ith

son

ian

Inst

itu

tio

n,

Na

tio

nal

Mu

seu

mo

fN

atu

ral

His

tory

,W

ash

ingto

nD

C;

WA

MW

este

rnA

ust

rali

an

Mu

seu

m,

Per

th,

Au

stra

lia

;Z

MA

Zo

olo

gis

chM

use

um

,A

mst

erd

am

.O

ther

ab

bre

via

tio

ns

refe

rto

coll

ecto

r’s

cata

log.

Sp

ecie

sO

rig

in/

vo

uch

erT

issu

eID

Lo

cali

ty

Acc

essi

on

nu

mb

ers

RA

G-1

RA

G-2

vW

Fcy

tb

12

S/

tVal

16S

c-m

os

Art

ibeu

sja

ma

icensi

sG

enB

an

k-

-A

Y8

34

655

AY

83

46

63

AY

83

47

37

AF

06

134

0A

F06

13

40

AF

06

13

40

-

Rh

inop

om

ah

ard

wic

kii

Gen

Ba

nk

--

AF

447

518

AF

44

75

35

AF

44

75

51

AY

62

900

5A

F26

32

32

AF

26

32

32

-

Nycti

mene

alb

ivente

rG

enB

an

k-

-A

Y2

49

870

AF

44

75

31

AF

44

75

49

-U

61

07

7A

F29

36

40

AY

0446

69

Nycti

mene

viz

caccia

AM

NH

PR

S26

36

AM

CC

12

420

8S

olo

mo

nIs

lan

ds,

Wes

tern

Pro

v.,

New

Geo

rgia

Gro

up

,V

ella

Lav

ella

Isla

nd

EU

61

79

59

EU

61

79

04

DQ

44

56

98

DQ

445

71

1-

--

Cy

no

pte

rus

sph

inx

AM

NH

274

354

AM

CC

10

168

8V

ietn

am

,H

a

Gia

ng

Pro

vin

ce,

Vi

Xu

yen

Dis

tric

t,

Cao

Bo

Co

mm

un

e,

Mt

Tay

Co

n

Lin

hII

EU

61

79

47

EU

61

78

97

DQ

44

56

97

DQ

445

70

3-

--

Cy

no

pte

rus

sph

inx

Gen

Ba

nk

--

--

--

EF

13

98

81

EF

13

98

81

-

Pte

no

chir

us

jag

ori

FM

NH

175

395

LR

H6

700

Ph

ilip

pin

es,

Lu

zon

,

Kali

ng

aP

rov

.,

Balb

ala

nM

un

ic.,

Balb

ala

san

g

EU

61

79

60

EU

61

79

10

DQ

44

56

96

FJ2

18

48

0-

--

Pte

ropu

sto

ng

anu

sA

MN

H2

72

873

AM

CC

12

496

2T

on

ga

EU

61

79

76

EU

61

79

24

DQ

44

56

95

--

--

Pte

ropu

sto

ng

anu

sG

enB

an

k-

--

--

AF

04

465

6A

Y0

447

98

AF

04

46

25

AY

0447

20

Pte

ropu

shyp

om

ela

nus

Un

cata

log

edP

444

7C

ap

tiv

ity

Lu

bee

Fo

un

da

tio

n

EU

61

79

65

EU

61

79

14

DQ

44

56

87

--

--

Pte

ropu

shyp

om

ela

nus

Gen

Ba

nk

--

--

-A

B06

247

2U

93

07

3A

F06

95

37

AY

0447

10

Melo

ny

cte

ris

mela

no

ps

US

NM

58

002

9U

SN

M5

800

29

‘‘P

ap

ua

New

Gu

inea

’’

FJ2

184

65

FJ2

184

61

FJ2

184

69

--

--

Melo

ny

cte

ris

mela

no

ps

Gen

Ba

nk

--

--

-A

F04

464

5A

Y0

447

39

AF

04

46

14

AY

0446

81

Melo

ny

cte

ris

fard

ouli

siA

MN

H2

75

744

AM

CC

12

427

9S

olo

mo

nIs

lan

ds,

Wes

tern

Pro

v.,

New

Geo

rgia

Gro

up

,

Vel

la

Lav

ella

Isla

nd

EU

61

79

57

EU

61

79

07

DQ

44

56

99

FJ2

18

47

8-

--

Melo

ny

cte

ris

fard

ouli

siG

enB

an

k-

--

--

-U

93

05

6A

F29

36

44

AY

0446

82

202 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:22 203 Cust # 582-6

Sp

ecie

sO

rig

in/

vo

uch

erT

issu

eID

Lo

cali

ty

Acc

essi

on

nu

mb

ers

RA

G-1

RA

G-2

vW

Fcy

tb

12S

/tV

al

16S

c-m

os

Syco

nycte

ris

au

stra

lis

MV

Z14

02

65

MV

Z1

402

65

Pa

pu

aN

ewG

uin

ea,

Wes

tern

Hig

hla

nd

s

Pro

v.,

Tra

un

aV

all

ey,

Ba

iyer

Riv

er

FJ2

18

46

6F

J218

46

2F

J218

470

FJ2

18

479

--

-

Syco

nycte

ris

au

stra

lis

Gen

Ban

k-

--

--

-U

930

60

AF

2936

50

AY

044

677

Macro

glo

ssus

min

imus

AM

NH

27

576

1A

MC

C1

242

83

So

lom

on

Isla

nd

s,

Wes

tern

Pro

vin

ce,

New

Geo

rgia

Gro

up

,

Vel

laL

av

ella

Isla

nd

EU

61

79

55

EU

617

90

5D

Q4

45

693

--

--

Macro

glo

ssus

min

imus

Gen

Ban

k-

--

--

AF

14

40

67

U9

30

62

AF

2936

49

AY

044

679

Dob

son

iam

ino

rM

VZ

14

02

08

MV

Z1

402

08

Pa

pu

aN

ewG

uin

ea,

Mad

an

gP

rov.,

Sem

pi

FJ2

18

46

7F

J218

46

3D

Q4

45

701

DQ

445

705

FJ2

18

47

7F

J21

847

7-

Dob

son

iam

olu

ccensi

sA

MM

207

35

EB

U2

57

57

Pa

pu

aN

ewG

uin

ea,

Sid

eia

Mis

sio

n,

Mil

ne

Ba

yP

rov

ince

EU

61

79

49

EU

617

89

9E

U6

17

930

FJ2

18

484

FJ2

18

47

2-

-

Dob

son

iam

olu

ccensi

sG

enB

an

k-

--

--

--

AF

1792

90

-

Dob

son

iain

erm

isA

MN

H27

573

0A

MC

C1

244

28

So

lom

on

Isla

nd

s,

Wes

tern

Pro

v.,

New

Geo

rgia

Gro

up

,

Vo

nav

on

aL

ag

oo

n

EU

61

79

48

EU

617

89

8D

Q4

45

686

DQ

445

704

FJ2

18

47

6F

J21

847

6-

Dob

son

iap

ero

nii

WA

MM

30

34

1M

30

34

1In

do

nes

ia,

Su

mb

a-

--

Un

sub

mit

ted

a-

--

Ap

rote

les

bu

lmer

ae

AM

26

66

0A

M2

666

0N

ewG

uin

ea-

--

Un

sub

mit

ted

aU

930

66

AF

293645

AY

044702

Bon

eia

bid

ens

ZM

A2

310

0U

nca

talo

ged

Ind

on

esia

,‘‘

Su

law

esi’

’F

J218

46

8F

J218

46

4F

J218

471

FJ2

18

481

FJ2

18

47

5F

J21

847

5-

Harp

yio

ny

cte

ris

wh

itehead

i

FM

NH

14

66

46

LR

H4

81

1P

hil

ipp

ines

,

Min

da

nao

,B

uk

idn

on

Pro

v.,

Mo

un

tK

ita

ng

lad

Ran

ge

EU

61

79

54

EU

617

90

4D

Q4

45

690

DQ

445

708

FJ2

18

47

4F

J21

847

4-

Harp

yio

ny

cte

ris

cele

ben

sis

Un

cata

log

edU

nca

talo

ged

Ind

on

esia

,‘‘

Su

law

esi’

’-

--

-F

J21

847

3F

J21

847

3-

Ro

use

ttu

sa

egy

pti

acu

sA

MN

H11

738

6U

nca

talo

ged

Mo

zam

biq

ue,

Za

mb

ezia

,M

t.

Nam

uli

EU

61

79

79

EU

617

92

7D

Q4

45

688

DQ

445

713

--

-

Ro

use

ttu

sa

egy

pti

acu

sG

enB

an

k-

--

--

-A

B205

18

3A

B2051

83

-

Ro

use

ttu

s

am

ple

xic

au

da

tus

Gen

Ban

k-

-A

F44

751

2A

F44

752

9A

Y0

57

83

6A

B04

63

29

U9

30

70

AF

2037

42

AY

044

690

Eon

ycte

ris

spela

eaM

VZ

17

64

87

MV

Z1

764

87

Ch

ina

,‘‘

Yu

nn

an

Pro

vin

ce’’

EU

61

79

51

EU

617

90

1D

Q4

45

684

FJ2

18

482

--

-

AP

PE

ND

IX1

(C

on

tin

ued)

2009 GIANNINI ET AL.: HARPYIONYCTERINE PHYLOGENY 203

Bulletin of the American Museum of Natural History amnb-582-00-06.3d 24/9/09 15:58:22 204 Cust # 582-6

Sp

ecie

sO

rig

in/

vo

uch

erT

issu

eID

Lo

cali

ty

Acc

essi

on

nu

mb

ers

RA

G-1

RA

G-2

vW

Fcy

tb

12S

/tV

al

16

Sc-

mo

s

Eony

cte

ris

spel

aea

Gen

Ban

k-

--

--

-U

93059

AF

203743

AY

04468

5

My

ony

cte

ris

torq

ua

taA

MN

H2

68

36

2A

MC

C1

09

058

Cen

tral

Afr

ican

Rep

ub

lic,

Sa

ng

ha,

Dza

ng

a-S

an

gh

a

EU

617

958

EU

61

79

08

DQ

44

570

0F

J21

848

3-

--

My

ony

cte

ris

torq

ua

taG

enB

an

k-

--

--

-A

Y044744

AF

044619

AY

04468

6

Meg

alo

glo

ssus

woerm

anni

AM

NH

26

835

8A

MC

C1

09

064

Cen

tral

Afr

ican

Rep

ub

lic,

Sa

ng

ha,

Dza

ng

a-S

an

gh

a

EU

617

956

EU

61

79

06

DQ

44

570

2D

Q44

571

0-

--

Meg

alo

glo

ssus

woerm

anni

Gen

Ban

k-

--

--

-U

93055

AF

203741

AY

04468

9

Epom

ops

fran

queti

AM

NH

23

835

6A

MC

C1

09

070

Cen

tral

Afr

ican

Rep

ub

lic,

Sa

ng

ha,

Dza

ng

a-S

an

gh

a

EU

617

952

EU

61

79

02

DQ

44

569

2D

Q44

570

7-

--

Epom

ops

fran

queti

Gen

Ban

k-

--

--

--

AF

044608

AY

04468

7

Epom

oph

oru

sw

ah

lber

gi

AM

NH

11

733

6F

MN

H1

77

209

Mo

zam

biq

ue,

Zam

bez

ia,

Mt.

Nam

uli

UE

617

953

EU

61

79

03

DQ

44

569

1D

Q44

570

6-

--

Epom

oph

oru

sw

ah

lber

gi

Gen

Ban

k-

--

--

-U

93

06

4A

F20

37

44

aS

equ

ence

sp

rov

ided

by

Dea

nn

eB

yrn

es.

AP

PE

ND

IX1

(C

on

tin

ued)

204 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 582