Journal of Mammalogy, 84(3):791±808, 2003

SYSTEMATICS OF ROUND-EARED ( AND ) BASED ON NUCLEAR AND MITOCHONDRIAL DNA SEQUENCES

CALVIN A. PORTER,* STEVEN R. HOOFER,RONALD A. VAN DEN BUSSCHE, THOMAS E. LEE,JR., AND ROBERT J. BAKER Department of Biological Sciences, Texas Tech University, Lubbock, TX 79409, USA (CAP, RJB) Department of Zoology, Collection of Vertebrates, and Oklahoma Cooperative Fish and Wildlife Research Unit, Oklahoma State University, Stillwater, OK 74078, USA (SRH, RAVDB) Department of Biology, Box 27868, Abilene Christian University, Abilene, TX 79699, USA (TEL) Museum of Texas Tech University, Lubbock, TX 79409, USA (RJB) Present address of CAP: Department of Biology, Xavier University of Louisiana, 1 Drexel Drive, Box 85B, New Orleans, LA 70125, USA

We examined the systematics of round-eared bats (Tonatia and Lophostoma) using sequence data from the nuclear recombination activator gene-2 (Rag2) gene and the mitochondrial valine transfer RNA, 12S ribosomal RNA (rRNA), and 16S rRNA. Some analyses of the mitochondrial and combined data sets indicate that round-eared bats may be paraphyletic relative to the genera , Phylloderma, and Mimon. Both nuclear and mito- chondrial data sets indicate substantial genetic divergence between Tonatia and Lophosto- ma, and neither data set unambiguously demonstrates monophyly of round-eared bats. How- ever, our analyses cannot exclude the possibility that round-eared bats form a monophyletic lineage that separated anciently into Tonatia and Lophostoma. Within the Tonatia lineage, T. bidens and T. saurophila are sister taxa but are divergent for both nuclear and mito- chondrial sequences. Nuclear data suggest that L. silvicolum may be paraphyletic, as cur- rently recognized, because L. evotis appears within this clade.

Key words: Lophostoma, Phyllostomidae, Rag2, rDNA, systematics, Tonatia

The round-eared bats include about 7 genus and phylogenetic relationships species of insectivorous and occasionally among species are contentious. This lack of frugivorous bats belonging to the family stability is a result of recent taxonomic Phyllostomidae. The genus Tonatia was es- changes as well as a paucity of phyloge- tablished by Gray (1827) for the species netic studies on all species of round-eared Vampyrus bidens, and despite substantial bats. In addition, studies that did examine ¯uidity in phyllostomid (re- all species did not include representatives viewed by Wetterer et al. 2000), Tonatia of other phyllostomines as in-group taxa or has consistently been associated with the use characters that provided ambiguous res- nominate subfamily, . As olution for intergeneric relationships (Lee et de®ned by most recent authorities (Wil- al. 2002). liams et al. 1995), Tonatia includes 7 spe- Tonatia saurophila was described by cies of round-eared bats (bidens, saurophi- Koopman and Williams (1951) based on la, brasiliense, carrikeri, schulzi, evotis, fossil material of uncertain age collected and silvicola). However, monophyly of the from 2 Jamaican caves. Koopman and Wil- liams (1951) recognized T. bidens as the * Correspondent: [email protected] closest relative of T. saurophila but de-

791 792 JOURNAL OF MAMMALOGY Vol. 84, No. 3 scribed T. saurophila as a distinct species ferred all other species to Lophostoma, the based on skull and tooth morphologies and oldest available genus name. slightly smaller size. Koopman (1976) and The taxonomy of the L. silvicolum com- others (Morgan and Woods 1986) later rel- plex has ¯uctuated over the years. Species egated the fossil T. saurophila to a subspe- in this complex are characterized by a cies of the living taxon T. bidens. Koop- prominent postorbital constriction, and sev- man's (1976) revision was based on a com- eral names are available for this group parison of the fossils with specimens of T. (Koopman 1993). L. silvicolum was de- bidens from middle America and northern scribed in 1836, with 2 similar species be- South America (Williams et al. 1995) but ing described subsequently (L. amblyotis without consideration of T. bidens from the and L. laephotis). The L. silvicolum com- southern part of the range. Williams et al. plex was revised by Cabrera (1958), Good- (1995) examined the morphology of extant win (1942), and Laurie (1955), with all 3 populations and concluded that bats from taxa (amblyotis, laephotis, and silvicolum) the northern part of the range were specif- ®nally being synonymized as T. silvicola. ically distinct from T. bidens from Para- Davis and Carter (1978) described the spe- guay, Argentina, and southern Brazil, with cies L. evotis for the smaller specimens the 2 species being sympatric in eastern from Middle America, with all other forms Brazil. Williams et al. (1995) determined included under L. silvicolum. Subsequent that the southern species is conspeci®c with classi®cations (Koopman 1993; Nowak the type specimen of T. bidens. With T. bi- 1999; Williams et al. 1995) have followed dens being applied to the southern species, this taxonomy. T. saurophila is the oldest available name Recently, Baker et al. (2000) reported on for the northern species, including the con- the systematics of Phyllostomidae based on speci®c Jamaican fossils. In citing previous sequences of the nuclear recombination-ac- studies, we will use current nomenclature tivator gene-2 (Rag2). Rag2 is one of two for T. saurophila in cases where the origi- adjacent genes that code for proteins that nal authors used the name T. bidens. function in the assembly of immunoglobu- Several studies have suggested 2 major lin genes (Oettinger et al. 1990; Shatz et al. clades among the round-eared bats. Immu- 1989). Their study included only a single nologic (Honeycutt and Sarich 1987), allo- species of Lophostoma (L. brasiliense) and zymic (Arnold et al. 1983), and karyotypic did not include the species currently rec- (Baker and Bickham 1980; Patton and Bak- ognized as members of Tonatia; therefore, er 1978) data indicate that both T. schulzi it provides no information on possible pa- and T. saurophila are divergent from most raphyly of the round-eared bats. In this or all other round-eared bats. Honeycutt and study, we examined Rag2 sequences from Sarich (1987) predicted that revision of the additional species of round-eared bats and genus would be required after further study. mitochondrial ribosomal sequences to test Lee et al. (2002) recently examined DNA the taxonomic validity of L. evotis and T. sequence variation in the mitochondrial ri- saurophila and to test the validity of as- bosomal genes and detected signi®cant dif- signment of several Tonatia species to the ferentiation, and possibly paraphyly, within genus Lophostoma (Lee et al. 2002). the round-eared bats. The results of Lee et al. (2002) are not unequivocal by them- MATERIALS AND METHODS selves regarding the paraphyly of the Specimens examined.ÐSpecimens, localities, round-eared bats. However, considering and voucher numbers are listed in Appendix I. their data and the results of previous stud- Except in northeastern Brazil, bats of the genus ies, Lee et al. (2002) retained Tonatia for Tonatia (sensu Lee et al. 2002) can generally be the species bidens and saurophila and re- identi®ed to species based on collecting locality; August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 793 however, our study includes a Tonatia specimen ment with no indels being required. We coded (TK49885) collected from central Bolivia, a re- nucleotides as discrete, unordered characters, gion outside the published range of both T. sau- and analysis was performed using parsimony rophila and T. bidens (Williams et al. 1995). The and distance options in PAUP* 4.0b8 software voucher has been examined by personnel of the (Swofford 2001). The level of phylogenetic sig-

Museum of Southwestern Biology (Albuquer- nal was assessed using the g1 statistic (Hillis and que, New Mexico) and by CAP, who concurred Huelsenbeck 1992) with 100,000 randomly on its identi®cation as T. saurophila according drawn trees. Parsimony analyses consisted of to the morphological criteria of Williams et al. heuristic searches with 25 random additions of (1995). input taxa and tree bisection±reconnection Another specimen has been examined in past branch swapping (Swofford and Olsen 1990). studies without reference to the taxonomic re- Stability or support of inferred topologies was vision of Williams et al. (1995). Teeling et al. assessed via bootstrapping (Felsenstein 1985) (2000) reported Rag2 and ribosomal DNA with 1,000 iterations and Bremer support (Bre- (rDNA) data for T. bidens but did not provide mer 1988) using Autodecay 3.0.3 software (Er- information on the specimen or the source of iksson 1997). We conducted parsimony analyses their sample. Lee et al. (2002) analyzed the se- with equal weighting and various weighting sce- quence data of Teeling et al. (2000) and found narios, including weighting transversions over that it fell within the T. saurophila clade. M. J. transitions (3:1), weighting 1st- and 2nd-codon Stanhope (pers. comm.) has recently con®rmed positions over 3rd-codon positions (3:3:1), and the identity of the specimen examined by Teel- successive weighting using the rescaled consis- ing et al. (2000). This specimen (TK10418 in tency index (CIÐCarpenter 1994; Farris 1969). Appendix I) was collected from Suriname, and Designated out-groups were Desmodontinae on the basis of locality, it is referable to T. sau- ( rotundus, Diaemus youngi, and Di- rophila and not to T. bidens. Based on this iden- phylla ecaudata), Micronycterinae (Lampronyc- ti®cation, none of the specimens examined by teris brachyotis and schmidto- Lee et al. (2002) is T. bidens; therefore, we have rum), and Macrotinae ( waterhousii). It obtained samples of T. bidens from Brazil and is generally agreed (Baker et al. 2000; Wetterer Paraguay. Our analysis includes the mitochon- et al. 2000) that all these taxa are appropriate as drial DNA (mtDNA) data (GenBank accession out-groups. We used Macrotus to root our phy- number AF179288) of Teeling et al. (2000), but logenetic trees because other studies (Baker et we omitted their RAG2 data (GenBank accession al. 2000; R. J. Baker et al., in litt.) have already number AF203763) because these were incom- established that the Rag2 and mtDNA sequences plete relative to our other sequences. used in this study support a basal position for According to Reid (1997), the distributions of Macrotus. Morphological studies (Wetterer et al. L. evotis and L. silvicolum extend into Honduras, 2000) have reached different conclusions re- but the 2 species are allopatric. Our Honduran garding Macrotus, but the rooting of the tree specimen of L. evotis was identi®ed on morpho- should not in¯uence the topology of the in- logical criteria and was collected within the pub- group. Using PAUP*, we constructed a neigh- lished range (Reid 1997) of L. evotis. bor-joining tree based on corrected genetic dis- Rag2 analysis.ÐSequences of 14 individuals tances (Tamura and Nei 1993). from Baker et al. (2000) were obtained from Mitochondrial DNA analysis.ÐWe down- GenBank (accession numbers in Appendix I). loaded, from GenBank, 21 sequences of the mi- Using the methods of Longmire et al. (1997), tochondrial genes of valine transfer RNA we isolated DNA from 11 additional individuals (tRNA), 12S ribosomal RNA (rRNA), and 16S representing 5 species. Polymerase chain reac- rRNA (accession numbers in Appendix I), tion (PCR) protocols and DNA sequencing of which were examined by Lee et al. (2002), Teel- the Rag2 gene followed Baker et al. (2000). ing et al. (2000), and Van Den Bussche and Overlapping sequences for each individual were Hoofer (2000). In addition, we isolated genomic assembled into a contiguous sequence using Se- DNA from representatives of T. bidens from quencher software, version 3.1.1 (Gene Codes Brazil and Paraguay and L. silvicolum from Par- Corporation, Ann Arbor, Michigan), and we aguay (Longmire et al. 1997). Conditions for manually generated a multiple sequence align- PCR ampli®cation, DNA sequencing, and as- 794 JOURNAL OF MAMMALOGY Vol. 84, No. 3 sembling overlapping fragments were as de- acters were at the 1st-codon position, 10 at scribed by Van Den Bussche and Hoofer (2000). the 2nd, and 99 at the 3rd. The sequence Mitochondrial rDNA sequences were aligned alignment is available on request from CAP. using Clustal X (Thompson et al. 1997), with Analysis of 100,000 randomly generated the resulting alignment re®ned based on the sec- trees produced a g value of Ϫ0.8987, sug- ondary structure of the RNA gene products (An- 1 derson et al. 1982; de Rijk et al. 1994; Springer gesting a phylogenetic structure in the data and Douzery 1996). Alignment and re®nement (Hillis 1991; Hillis and Huelsenbeck 1992; of rDNA sequences were done by CAP without Huelsenbeck 1991; for criticisms of this knowledge of or reference to the alignment of statistic, see KaÈllersjoÈ et al. 1992). Lee et al. (2002). Nucleotides were coded as un- Unweighted parsimony analysis of the ordered, discrete characters, and phylogenetic Rag2 data produced 20 equally most-parsi- analyses were performed as described for the monious trees with a length of 414 (retention RAG2 data, with gaps being treated as missing index [RI] ϭ 0.735, consistency index [CI] data. ϭ 0.732). Four of the 20 equally parsimo- Combined analysis.ÐBefore combining Rag2 nious trees support monophyly of the round- and mtDNA sequences, we tested for character congruence using the partition homogeneity test eared bats, whereas the strict consensus tree (incongruence length difference testÐFarris et depicts an unresolved polytomy involving al. 1995; Mickevich and Farris 1981) with 1,000 Tonatia, Lophostoma, Phyllostomus, and a random partitions. The combined analysis in- clade consisting of Phylloderma and Mimon cluded 14 specimens for which we had both nu- (Fig. 1A). The node containing this polyto- clear and mitochondrial data from the same in- my has a Bremer support of 4 and was de- dividual (Appendix I). For the following addi- tected in 86% of the bootstrap iterations. tional taxa, we combined data from different in- Monophyly of Lophostoma is strongly sup- dividuals of the same species: Chrotopterus ported (Bremer support ϭ 9, bootstrap ϭ auritus (specimens from Suriname and Mexico), 100%), as is the T. bidens±T. saurophila L. evotis (Mexico and Honduras), L. schulzi (Guyana and French Guiana), M. waterhousii (2 clade (Bremer support ϭ 7, bootstrap ϭ individuals from the same locality in Cuba), Mi- 99%). L. evotis is the sister taxon to L. sil- mon crenulatum (Venezuela and Trinidad), Phyl- vicolum from French Guiana (Bremer sup- lostomus hastatus (2 individuals from nearby lo- port ϭ 2, bootstrap ϭ 93%), which in turn calities in Venezuela), and T. saurophila (1 con- is the sister group to the 2 specimens of L. catenated sequence from 2 nearby localities in silvicolum from Paraguay. Increasing the Guyana and a 2nd sequence concatenated from weight of transversions or 1st- and 2nd-co- specimens from Peru and Ecuador). With the ex- don positions did not in¯uence tree topology ception of C. auritus, we combined specimens and generally did not improve statistical sup- that were collected from relatively close locali- ties. We eliminated L. carrikeri from the com- port for nodes on the tree. Successive bined analysis because there was no Rag2 se- weighting produced 6 most-parsimonious quence. All 22 specimens remaining in the com- trees (Fig. 1B) with topologies similar to that bined analysis were complete for both nuclear produced by the unweighted analysis. and mitochondrial sequences. We performed the Appendix II shows the percent sequence same phylogenetic analyses on the combined difference for all pairwise comparisons of data set as were described for the RAG2 data. individuals, corrected for multiple substi- tutions. Regarding the 2 main objectives of RESULTS this study, mean sequence divergence be- Rag2 sequence analysis.ÐThe alignment tween T. bidens and T. saurophila is 2.83%, of 25 Rag2 sequences resulted in 1,363 among Lophostoma species it is 0.99%, and sites, of which 1,082 (79%) were constant between Tonatia and Lophostoma it is and 123 (9%) were phylogenetically infor- 3.26%. In addition, mean sequence diver- mative. Fourteen of the informative char- gence between L. evotis and L. silvicolum August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 795

FIG. 1.ÐA) Strict consensus of 20 most-parsimonious trees produced from an unweighted analysis of Rag2 sequence data. Numbers above each node represent percentage of 1,000 bootstrap iterations for which each clade was detected, whereas those below are Bremer support values. B) Strict con- sensus of 6 most-parsimonious trees produced from a successive approximations analysis of Rag2 sequence data. Bootstrap and Bremer support values are as illustrated in A. is 0.44%, as compared with a mean of sponding to the bracketed regions in ®g. 3) 1.13% for other pairwise comparisons of and 5 segments totaling 92 sites in the 16S Lophostoma species. A neighbor-joining rRNA gene. Therefore, the ®nal alignment tree based on Tamura and Nei (1993) cor- used for phylogenetic analysis consisted of rected distances revealed the same general 2,513 sites, of which 1,545 (61%) were topology as 1 of the 20 equally most-par- constant and 686 (27%) were phylogeneti- simonious trees (Fig. 2). The branch lengths cally informative. With the removal of the on the neighbor-joining tree (Fig. 2) re¯ect 217 ambiguously aligned sites, we are con- Tamura and Nei (1993) genetic distances ®dent that homologous nucleotides are and reveal substantial genetic divergence aligned in the analyzed data. The g1 value between Tonatia and Lophostoma and be- for the mtDNA data was Ϫ0.907, and com- tween the 2 species of Tonatia. parison with critical values indicated that Mitochondrial DNA sequence analy- there was a nonrandom structure in the data sis.ÐAlignment of mtDNA from 24 taxa (Hillis and Huelsenbeck 1992). resulted in 2,730 aligned sites. Our align- Unweighted parsimony analysis of 2,513 ment is 28 sites shorter than the alignment aligned mtDNA sites resulted in a single of 21 taxa generated independently by Lee most-parsimonious tree (length ϭ 3,130, RI et al. (2002). Because the sequence align- ϭ 0.462, CI ϭ 0.445; Fig. 3) that strongly ment was questionable for some loop re- supports a clade consisting of T. bidens and gions, we designated 217 sites as being am- T. saurophila,aLophostoma clade, and a biguous and removed them from all analy- clade consisting of Chrotopterus and Vam- ses. These 217 sites include 9 segments to- pyrum. Phylogenetic analysis based on suc- taling 125 aligned sites in the 12S rRNA cessive weighting produced a single most- gene (Springer and Douzery 1996Ðcorre- parsimonious tree (length ϭ 767.51, CI ϭ 796 JOURNAL OF MAMMALOGY Vol. 84, No. 3

FIG. 2.ÐNeighbor-joining tree based on Tamura and Nei (1993) genetic distances calculated from Rag2 sequence data.

0.735, RI ϭ 0.665) with the same topology mean of 6.35% between other species pairs that resulted from the unweighted analysis within Lophostoma and 1.67% between in- with moderately strong support (Bremer dividuals of L. silvicolum. support ϭ 2, bootstrap ϭ 88%) for the as- Combined sequence analysis.ÐWe de- sociation of Lophostoma with Phyllosto- tected no signi®cant heterogeneity (P ϭ mus, Phylloderma, and Mimon, a relation- 0.917) in phylogenetic pattern between ship not strongly supported by the un- the Rag2 and mtDNA data, and the g1 sta- weighted analysis. tistic (P ϭϪ0.961) suggested the phylo- Similar to results based on Rag2 data, genetic structure in the data. Unweighted percent sequence divergence (Tamura and analysis of the combined data set pro- Nei 1993) for all pairwise comparisons of duced a single most-parsimonious tree taxa (Appendix III) documented high levels (3,457 steps, RI ϭ 0.457, CI ϭ 0.487; Fig. of genetic differentiation between Tonatia 4), and the same topology was recovered and Lophostoma but low divergence be- when homoplastic characters were down- tween L. evotis and L. silvicolum. Mean se- weighted (length ϭ 1,004.86, RI ϭ 0.699, quence divergence between T. bidens and CI ϭ 0.798). The only difference between T. saurophila is 9.36%, as compared with topologies produced by unweighted and 0.08% among Brazilian and Paraguayan successive weighting was that additional specimens of T. bidens and a mean of support for some nodes was detected with 1.34% among individuals of T. saurophila. successive weighting. For example, boot- Mean sequence divergence among species strap support for Phyllostomus±Phyllo- of Lophostoma is 5.75%, and between Ton- derma±Mimon as the sister group to Lo- atia and Lophostoma it is 10.83%. Between phostoma increased with successive L. evotis and L. silvicolum, mean sequence weighting (from 50% to 93% of bootstrap divergence is 2.13%, as compared with a replications). August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 797

DISCUSSION data consistent with the Lophostoma±Ton- atia dichotomy. The majority of 20th century authors rec- Our separate and combined analyses of ognized a single genus (Tonatia) for the mitochondrial and Rag2 data (Figs. 1±4) round-eared bats (Arnold et al. 1983; Good- provide support for the recognition of 5 win 1942; Hall 1981; Hall and Kelson distinct lineages among the phyllostomi- 1959; Jones and Carter 1976; Miller 1907; nes: Tonatia, consisting of T. bidens and Nowak 1999; Simpson 1945; Wetterer et al. T. saurophila; Lophostoma; Chrotopterus 2000; Williams et al. 1995). However, sev- and Vampyrum; Trachops and Macro- eral studies have provided evidence that the phyllum; and a clade consisting of Phyl- T. bidens complex (recognized until 1995 loderma, Phyllostomus, and Mimon.Of as the single extant species T. bidens) is di- these 5 clades, the ®rst 2 are the most vergent from the other round-eared bats. strongly supported by all analyses and all Although the karyotype of T. bidens (sensu data sets. The association of Lophostoma Williams et al. 1995) has not been de- with Phyllostomus, Phylloderma, and Mi- scribed, the G-banded karyotype of T. sau- mon receives only moderate support in the rophila, a member of the T. bidens com- unweighted analyses. Our data show a plex, is radically reorganized from that substantial genetic distance (Appendices found in other round-eared bats and in other II and III; Fig. 2) between the 2 major related genera (Baker and Bickham 1980; clades of round-eared bats. This distance Patton and Baker 1978). Additional support is comparable with that separating other for the uniqueness of the T. bidens complex genera of phyllostomid bats (Appendices comes from allozymic and immunologic II and III) and is in agreement with pre- studies. Arnold et al. (1983) found that L. vious data from Arnold et al. (1983) and schulzi and T. saurophila did not share any Honeycutt and Sarich (1987). Therefore, synapomorphic allozymic alleles with each allozyme, nuclear DNA and mtDNA se- other or with other round-eared bats. Sim- quences, and immunologic data support ilarly, immunologic distances were inter- an ancient separation between Tonatia preted by Honeycutt and Sarich (1987) as and Lophostoma and support recognition indicating that both T. saurophila and L. of Lophostoma. We, therefore, follow Lee schulzi (along with L. carrikeri) were ge- et al. (2002) and continue to recognize netically divergent from other round-eared Lophostoma as a valid genus. bats. Our study includes 2 independent molec- The studies mentioned above detected ular data sets, which strongly support Avilla signi®cant genetic differentiation among et al. (1998) and Williams et al. (1995) in round-eared bats, although the authors did the recognition of T. saurophila as a species not propose any taxonomic revision. In distinct from T. bidens. The 2 species are light of these previous studies (Arnold et al. genetically distant (Appendices II and III; 1983; Baker and Bickham 1980; Honeycutt Fig. 2), with a mean sequence divergence and Sarich 1987; Patton and Baker 1978) (2.83% for Rag2 and 9.36% for mtDNA) and their analysis of DNA sequence varia- that approaches or exceeds the degree of tion in the mitochondrial rRNA genes, Lee genetic divergence found between some et al. (2002) divided the round-eared bats genera (Appendices II and III). Both mo- into 2 genera; T. bidens and T. saurophila lecular biology and classical morphological were retained in the genus Tonatia, whereas studies continue to play important roles in all other species were elevated to generic the discovery of some species that have es- status, with Lophostoma being the oldest caped recognition despite many decades of available name. An abstract of a morpho- study. logical study (Avilla et al. 1998) references The taxonomy of the L. silvicolum com- 798 JOURNAL OF MAMMALOGY Vol. 84, No. 3

FIG. 3.ÐMost-parsimonious tree produced from an analysis of the 12S ribosomal RNA (rRNA), valine transfer RNA, and 16S rRNA genes. Parsimony analyses based on equal and successive weighting produced identical, single most-parsimonious trees. For each node, the 1st number is the percentage of 1,000 bootstrap iterations in which each clade was detected for the unweighted analysis, the 2nd number is the bootstrap percentage for the successive weighting analysis, the 3rd number is the Bremer support value for the unweighted analysis, and the last number is the Bremer support value for the successive weighting analysis, rounded to the nearest whole number. plex has been revised repeatedly during the combined specimens from the remainder past century (Cabrera 1958; Davis and of the range under what is now recognized Carter 1978; Goodwin 1942; Laurie 1955). as L. silvicolum. Rag2 data indicate that L. The most recent revision by Davis and silvicolum from northern South America is Carter (1978) included the description of more closely allied with L. evotis than with L. evotis for the specimens from Mexico, L. silvicolum from the southern part of the Belize, Guatemala, and Honduras and range. Mitochondrial data do not support August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 799

FIG. 4.ÐSingle most-parsimonious tree produced from analysis of combined nuclear and mi- tochondrial sequences. Equally and successive weighting analyses produced the same topology for the most-parsimonious tree in the respective analysis. Support values shown on each node are as in Fig. 3. this relationship, although they do not RESUMEN strongly contradict it. The L. silvicolum complex is widely distributed throughout Se examino la sistemaÂtica de murcieÂlagos much of Latin America, and several nom- de oreja redondeada (Tonatia y Lophosto- inal species have been described. Our ma) usando datos de secuencias nucleares Val study includes representatives from limited del gen Rag2 y mitocondriales de ARNt , parts of the range; therefore, we do not ARNr 12s, y ARNr 16s. Algunos anaÂlisis propose any taxonomic revisions at this de los datos mitocondriales y combinados time. However, our results warrant further indican que los murcieÂlagos de oreja redon- study of morphologic and genetic variation deada pueden ser para®leÂticos con respecto throughout the range of the L. silvicolum a los geÂneros Phyllostomus, Phylloderma,y complex. Mimon. Tanto los datos nucleares como los 800 JOURNAL OF MAMMALOGY Vol. 84, No. 3 mitocondriales indican una divergencia ge- A new species of Tonatia from the Brazilian savan- nas and its phylogenetic interrelationships (Chirop- neÂtica importante entre Tonatia y Lophosto- tera, Phyllostominae). Research News 39:127. ma, y ninguno de los datos demuestra de BAKER,R.J.,AND J. W. BICKHAM. 1980. Karyotypic modo concluyente la mono®lia de los mur- evolution in bats: evidence of extensive and conser- vative chromosomal evolution in closely related cieÂlagos de oreja redondeada. Sin embargo, taxa. Systematic Zoology 29:239±253. nuestros anaÂlisis no excluyen la posibilidad BAKER, R. J., C. A. PORTER,J.C.PATTON, AND R. A. de que los murcieÂlagos de oreja redondeada VAN DEN BUSSCHE. 2000. Systematics of bats of the family Phyllostomidae based on RAG2 DNA se- formen un grupo mono®leÂtico que se separo quences. Occasional Papers, The Museum, Texas tempranamente en Tonatia y Lophostoma. Tech University 202:i ϩ 1±16. En el linaje de Tonatia, T. bidens y T. sau- BREMER, K. 1988. The limits of amino acid sequence rophila son grupos hermanos pero diver- data in angiosperm phylogenetic reconstruction. Evolution 42:795±803. gentes, tanto por las secuencias nucleares CABRERA, A. 1958. CataÂlogo de los mammõÂferos de como por las mitocondriales. Los datos nu- AmeÂrica del Sur. Revista del Museo Argentino Cien- cleares sugieren que L. silvicolum podrõÂa cias Nataturales ``Bernardino Rivadavia'' Ciencias ZooloÂgicas 4:1±307. ser para®leÂtico tal como se le reconoce ac- CARPENTER, J. M. 1994. Successive weighting, reli- tualmente, ya que L. evotis se encuentra ability, and evidence. Cladistics 10:210±220. dentro de dicho clado. DAVIS,W.B.,AND D. C. CARTER. 1978. A review of the round-eared bats of the Tonatia silvicola com- plex, with descriptions of three new taxa. Occasional CKNOWLEDGMENTS A Papers, The Museum, Texas Tech University 53:1± We thank the following for loan of tissues and 12. for access to voucher specimens: M. D. Engs- DE RIJK,P.,Y.VAN DE PEER,S.CHAPELLE, AND R. DE WACHTER. 1994. Database on the structure of large trom of the Royal Ontario Museum; J. L. Patton ribosomal subunit RNA. Nucleic Acids Research 22: and E. Lacey of the Museum of Vertebrate Zo- 3495±3501. ology, University of California, Berkeley; N. ERIKSSON, T. 1997. Autodecay. Version 3.03. Botaniska Simmons of the American Museum of Natural Institution, Stockholm University, Stockholm, Swe- History; T. L. Yates, W. L. Gannon, and J. Sa- den. FARRIS, J. S. 1969. A successive approximations ap- lazar-Bravo of the Museum of Southwestern Bi- proach to character weighting. Systematic Zoology ology, University of New Mexico; and M. R. 18:374±385. Willig, R. R. Monk, R. D. Stevens, and the Nat- FARRIS, J. S., M. KAÈ LLERSJOÈ ,A.G.KLUGE, AND C. ural Science Research Laboratory of Texas Tech BULT. 1995. Testing signi®cance of incongruence. University. M. J. Stanhope provided unpub- Cladistics 10:315±319. FELSENSTEIN, J. 1985. Con®dence limits on phyloge- lished data. F. G. Hoffmann provided assistance nies: an approach using the bootstrap. Evolution 39: with tissue and DNA samples. A. M. Porter crit- 783±791. ically reviewed the manuscript. Funding was GOODWIN, G. G. 1942. A summary of recognizable provided by a grant from National Science species of Tonatia with descriptions of two new spe- Foundation (DEB 98-73657) to R. A. Van Den cies. Journal of Mammalogy 23:204±209. GRAY, J. E. 1827. Synopsis of the species of the class Bussche. Many of the tissues used in this study Mammalia, as arranged with reference to their or- were collected using National Science Founda- ganization, by Cuvier, and other naturalists, with tion support to R. J. Baker for chromosomal speci®c characters, synonyma, etc. In The studies of phyllostomid bats. We thank E. Lessa kingdom arranged in conformity with its organiza- for preparing the Spanish summary. tion, by the Baron Cuvier, with additional descrip- tions of all the species hitherto named and of many not before noticed (E. Grif®th, C. H. Smith, and E. LITERATURE CITED Pidgeon, eds.). G. B. Whittaker, London, United ANDERSON, S., M. H. L. DE BRUILN,A.R.COULSON,I. Kingdom 5:1±391. C. EPERON,F.SANGER, AND I. G. YOUNG. 1982. Com- HALL, E. R. 1981. The of North America. plete sequence of bovine mitochondrial DNA: con- 2nd ed. John Wiley & Sons, Inc., New York 1:1± served features of the mammalian mitochondrial ge- 600 ϩ 90. nome. Journal of Molecular Biology 156:683±717. HALL,E.R.,AND K. R. KELSON. 1959. The mammals ARNOLD, M. L., R. J. BAKER, AND R. L. HONEYCUTT. of North America. Ronald Press, New York 1:1±546 1983. Genic differentiation and phylogenetic rela- ϩ 79. tionships within two New World bat genera. Bio- HILLIS, D. M. 1991. Discriminating between phyloge- chemical Systematics and Ecology 11:295±303. netic signal and random noise in DNA sequences. AVILLA, L. S., L. O. SALLES, AND N. B. SIMMONS. 1998. Pp. 278±294 in Phylogenetic analysis of DNA se- August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 801

quences (M. M. Miyamoto and C. Cracraft, eds.). genes that synergistically activate V(D)J recombi- Oxford University Press, New York. nation. Science 248:1517±1523. HILLIS, D. M., AND J. P. HUELSENBECK. 1992. Signal, PATTON, J. C., AND R. J. BAKER. 1978. Chromosomal noise, and reliability in molecular phylogenetic anal- homology and evolution of phyllostomatoid bats. ysis. Journal of Heredity 83:189±195. Systematic Zoology 27:449±462. HONEYCUTT,R.L.,AND V. M. SARICH. 1987. Mono- REID, F. A. 1997. A ®eld guide to the mammals of phyly and molecular evolution within three phyllo- Central America and southeast Mexico. Oxford Uni- stomid bat genera. Journal of Mammalogy 68:518± versity Press, New York. 525. SHATZ, D. G., M. A. OETTINGER, AND D. BALTIMORE. HUELSENBECK, J. P. 1991. Tree-length distribution 1989. The V(D)J recombination activating gene, skewness: an indicator of phylogenetic information. RAG-1. Cell 59:1035±1048. Systematic Zoology 40:257±270. SIMPSON, G. G. 1945. The principles of classi®cation JONES, J. K., JR., AND D. C. CARTER. 1976. Annotated and a classi®cation of mammals. Bulletin of the checklist, with keys to subfamilies and genera. Pp. American Museum of Natural History 85:1±350. 1±38 in Biology of bats of the New World family SPRINGER, M. S., AND E. DOUZERY. 1996. Secondary Phyllostomatidae. Part I. Special Publications, The structure, conservation of functional sites, and rates Museum, Texas Tech University 10:1±218. of 12S rRNA genes based on sequences from pla- centals, marsupials and a monotreme. Journal of KAÈ LLERSJOÈ , M., J. S. FARRIS,A.G.KLUGE, AND C. Molecular Evolution 43:357±373. BULT. 1992. Skewness and permutation. Cladistics 8:275±287. SWOFFORD, D. L. 2001. PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4. KOOPMAN, K., AND E. E. WILLIAMS. 1951. Fossil Chi- Sinauer Associates, Inc., Publishers, Sunderland, roptera collected by H. E. Anthony in Jamaica Massachusetts. 1919±1920. American Museum Novitates 1519:1± SWOFFORD,D.L.,AND G. J. OLSEN. 1990. Phylogeny 29. reconstruction. Pp. 411±501 in Molecular systemat- KOOPMAN, K. F. 1976. Zoogeography. Pp. 39±47 in ics (D. M. Hillis and C. Moritz, eds.). Sinauer As- Biology of bats of the New World family Phyllosto- sociates, Inc., Publishers, Sunderland, Massachu- matidae. Part I. Special Publications, The Museum, setts. Texas Tech University 10:1±218. TAMURA, K., AND M. NEI. 1993. Estimation of the KOOPMAN, K. F. 1993. Order Chiroptera. Pp. 137±241 number of nucleotide substitutions in the control re- in species of the world: a taxonomic and gion of mitochondrial DNA in humans and chim- geographic reference. 2nd ed. (D. E. Wilson and D. panzees. Molecular Biology and Evolution 10:512± M. Reeder, eds.). Smithsonian Institution Press, 526. Washington, D.C. TEELING, E. C., M. SCALLY,D.J.KAO,M.L.ROM- LAURIE, E. M. O. 1955. Notes on some mammal of AGNOLI,M.S.SPRINGER, AND M. J. STANHOPE. 2000. Ecuador. Annals and Magazine of Natural History, Molecular evidence regarding the origin of echolo- Series 12, 8:268±276. cation and ¯ight in bats. Nature 403:188±192. LEE, T. E., JR., S. R. HOOFER, AND R. A. VAN DEN THOMPSON, J. D., T. J. GIBSON,F.PLEWNIAK,F.JEAN- BUSSCHE. 2002. Molecular phylogenetics and taxo- MOUGIN, AND D. G. HIGGINS. 1997. The ClustalX nomic revision of the genus Tonatia (Chiroptera: windows interface: ¯exible strategies for multiple Phyllostomidae). Journal of Mammalogy 83:49±57. sequence alignment aided by quality analysis tools. LONGMIRE, J. L., M. MALTBIE, AND R. J. BAKER. 1997. Nucleic Acids Research 24:4876±4882. Use of ``lysis buffer'' in DNA isolation and its im- VAN DEN BUSSCHE, R. A., AND S. R. HOOFER. 2000. plication for museum collections. Occasional Papers, Further evidence for inclusion of the New Zealand The Museum, Texas Tech University 163:1±3. short-tailed bat (Mystacina tuberculata) within Noc- MICKEVICH,M.F.,AND J. S. FARRIS. 1981. The impli- tillionoidea. Journal of Mammalogy 81:865±874. cations of congruence in Menidia. Systematic Zo- WETTERER, A. L., M. V. ROCKMAN, AND N. B. SIM- ology 27:143±158. MONS. 2000. Phylogeny of phyllostomid bats (Mam- MILLER, G. S., JR. 1907. The families and genera of malia: Chiroptera): data from diverse morphological bats. Bulletin of the United States National Museum systems, sex chromosomes, and restriction sites. 57:1±282. Bulletin of the American Museum of Natural His- tory 248:1±200. MORGAN, G. S., AND C. A. WOODS. 1986. Extinction and the zoogeography of West Indian mammals. Bi- WILLIAMS, S. L., M. R. WILLIG, AND F. A. REID. 1995. Review of the Tonatia bidens complex (Mammalia: ological Journal of the Linnaean Society 28:167± Chiroptera), with descriptions of two new subspe- 203. cies. Journal of Mammalogy 76:612±626. NOWAK, R. M. 1999. Walker's mammals of the world. 6th ed. John Hopkins University Press, Baltimore, Submitted 17 January 2002. Accepted 14 February Maryland. 2003. OETTINGER, M. A., D. G. SHATZ,C.GORKA, AND D. BALTIMORE. 1990. RAG-1 and RAG-2, adjacent Associate Editor was Mark D. Engstom. 802 JOURNAL OF MAMMALOGY Vol. 84, No. 3 Rag2 AF442082* Val tRNA rDNA and AF411532 AF442092* AF442081* AF411528 GenBank accession number N, Ј 22 Њ W AF316442 Ј 12 Њ N, 56 Ân on Highway 200 AF263228 AF316444 Ј 44 Њ Ârida AF411533 AF316447 Ânamo Nava1 BaseÂnamo Naval Base AF263229 AF316461 I Ân: 1 km N Me PPENDIX W Ј A 29 Њ 60 S headquarters kari N, 1.2 km W Compamento Âxico: Campeche: 44 km S Constitucion AF411529 catalog number Locality Museum Tissue number TK17104 CM68638 Suriname: Saramacca: Raleigh Falls, 4 TK34625TK13514 TTU62792 Unknown El Salvador: La Paz: 1 Mexico: mi Yucata N La Herradura AF411534 AF316445 TK70457TK4764 USNM TTU35582 Peru Mexico: Guerrero: 24.1 mi N Rio La Unio AF411538 TK56716 TTU Paraguay: Departamento San Pedro: Yaguarete Forests, 1.5 km TK25238TK18834 TCWC55445TK49900 AMNH267103 ROM107391 Trinidad: Mayaro: French Guayagayare Guiana: Paracou: near Sinnamary Guyana: Potaro-Siparuni, Buro Buro R, 25 km WNW Kurupu- AF411544 AF316489 AF411536 AF316463 TK56635 TTUTK17946TK18832 CM77174TK19119 AMNH267107TK32030 CM78289TK32021 French TTU52481 Guiana: Paraguay: Paracou:TK70447 Suriname: Departamento near Marowijne, San Sinnamary TTU52478 Oelomarie Pedro: Yaguarete Forests, 2.5 MUSM13737 km Venezuela: Bolivar: 8 Cuba: km Guantanamo: S, Guanta 5 Cuba: Peru: km Guantanamo: Camisea E Guanta El Manteco AF411540 AF263230 AF316458 AF442083* AF411535 AF316470 TK49870 ROM95626 Me TK40341TK18833 TTU61070TK49888 AMNH267106 ROM101128 French Guiana: Honduras: Paracou: Atlantida: near Lancitilla Sinnamary Guyana: Northwest District, Baramita, Old World, 7 AF442079* AF442080* .ÐAccession numbers marked with asterisks are new sequences. Other sequences were obtained from GenBank and were Taxon Specimens examined Voucher specimens are deposited in mammal collections at the Natural Science Research Laboratory at Texas Tech University (TTU), the Royal Chrotopterus auritus used in Bakerwere et assigned al. a (2000), TexasMSB70286. Lee Tech Specimens University et identi®ed tissue al. only (TK) (2002), with number. Teeling a University et museum ofOntario al. acronym New Museum, (2000), have Mexico Canada not and tissue (ROM),at yet Van number the the been Den NK30034 Carnegie cataloged University Bussche also Museum orthe identi®es of and of the University specimen California, Hoofer Natural catalog number of Berkeley (2000). History, number NewNatural (MVZ), All is Pittsburgh, Mexico at the original unknown. Pennsylvania (MSB), the American museum (CM), the Universidad Museum specimens the Nacional Texas Cooperative of Mayor Museum Wildlife de Natural of Collection San History Vertebrate at Marcos, Zoology (AMNH), Lima, Texas the A&M Peru Museum University (MUSM). Individual (TCWC), of specimens and Southwestern also the Biology are Museo at cross-referenced de by Hisoria TK numbers. Diphylla ecaudata Diaemus youngi Desmodus rotundus Chrotopterus auritus Lophostoma silvicolum Lophostoma brasiliense Lophostoma carrikerri Lampronycteris brachyotis Lophostoma silvicolum Macrophyllum macrophyllum Macrotus waterhousii Lophostoma silvicolum Lophostoma silvicolum Macrotus waterhousii Micronycteris schmidtorum Lophostoma evotis Lophostoma evotis Lophostoma schulzi Lophostoma schulzi August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 803 Rag2 AF442089* Val tRNA rDNA and AF442090* AF442087* AF442091* AF442088* AF411530 AF411531 AF179288 GenBank accession number W AF411542 AF316480 Ј 12 Њ N, 56 Ј 44 Њ Ântida: Lancitilla AF411537 AF316495 IÐContinued. PPENDIX Â: Quebrado, Aguas Negras, Cocha Zoraida AF442085* S, 2.2 Km E Compamento reia-Itatins reia-Itatins bur Hill Sur km S Brownsberg. A catalog number Locality Museum number Tissue TK15121TK25230 TTU33287TK10201 CM25230 CM63614 Venezuela: Guarico: 45 km S Trinidad Calabozo and Tobago: Trinidad, 1 Suriname: mi Saramacca: Raleigh S, Falls, 2 4 mi W Guayaguayare AF411543 AF316472 TK19289TK19243 CM19289TK56633 CM78333 TTU Venezuela: Bolivar: 28 km Venezuela: E Bolivar: El 8 Palmar, km Rio W Grande El Manteco Paraguay: Departamento San Pedro: Yaguarete Forests, 1.2 km AF411541 AF316479 TK56519 MVZ185673 Brazil: Sao Paulo: Base do Rio Verde, Estacao Ecologica Ju- TK56520 MVZ185959TK49889TK49890 ROM103210 Brazil: Sao ROM103401 Paulo: Base do Rio Verde, Guyana: Estacao Upper Ecologica Takutu±Upper Ju- Essequiba: Guyana: 5 Upper km Demerara-Berbice: E Tropenbos, Surama 20 km SSE Ma- AF442084* TK46028TK49892 USNM ROM104459 Ecuador: Napo: Parque Nacional Peru Yasuni, 37 km S Pompeya TK49885TK10418 MSB70286 CM63675 Bolivia: Cochabamba: Valle de Sagta Suriname: Brodopondo, Brownsberg Nature Park, 2 km W, 8 AF442086* TK18829TK40370 AMNH267129 TTU61070 French Guiana: Paracou: near Sinnamary Honduras: Atla AF411539 AF316490 Taxon Mimon crenulatum Phylloderma stenops Mimon crenulatum Phyllostomus hastatus Tonatia bidens Phyllostomus hastatus Tonatia bidens Tonatia saurophila Tonatia bidens Tonatia saurophila Tonatia saurophila Tonatia saurophila Tonatia saurophila Tonatia saurophila Vampyrum spectrum Trachops cirrhosus 804 JOURNAL OF MAMMALOGY Vol. 84, No. 3

APPENDIX II Tamura and Nei (1993) genetic distances calculated from the Rag2 data and expressed as percent sequence divergence. Numbers in column headings correspond to the specimens as numbered in the 1st column.

Specimen 1 2 3 45678910 1. Macrotus TK32030 Ð 2. Diaemus TK34625 4.73 Ð 3. Diphylla TK13514 3.70 3.47 Ð 4. Desmodus TK4764 4.57 2.24 3.48 Ð 5. Lampronycteris TK25238 4.16 4.49 3.32 4.34 Ð 6. Micronycteris TK70447 4.81 5.30 4.99 5.23 4.75 Ð 7. Macrophyllum TK19119 3.79 4.26 3.08 4.57 3.39 4.82 Ð 8. Trachops TK18829 4.42 4.26 3.63 4.34 4.10 5.13 2.70 Ð 9. Chrotopterus TK17104 3.78 4.26 3.40 4.43 3.71 4.75 2.85 3.318 Ð 10. Vampyrum TK40370 3.39 3.55 2.70 3.79 3.09 4.27 2.17 2.702 1.79 Ð 11. Phylloderma TK10201 4.10 4.17 3.16 4.18 3.47 4.51 3.01 3.241 2.93 2.32 12. Phyllostomus TK19243 5.06 5.14 4.34 5.06 4.42 5.31 3.87 4.265 3.79 3.09 13. Mimon TK15121 5.15 5.39 4.51 5.31 5.08 6.23 4.59 4.833 4.43 3.72 14. Lophostoma brasiliense TK18834 4.91 4.99 4.27 5.00 4.52 5.18 3.96 4.039 3.89 3.10 15. L. schulzi TK18833 4.18 4.26 3.72 4.43 3.96 4.77 3.25 3.406 3.01 2.40 16. L. silvicolum TK18832 4.26 4.17 3.79 4.50 4.04 4.37 3.01 3.476 3.08 2.32 17. L. silvicolum TK56716 4.34 4.18 3.71 4.42 4.11 4.77 3.24 3.400 3.17 2.39 18. L. silvicolum TK56635 4.26 4.10 3.63 4.34 4.03 4.69 3.16 3.321 3.09 2.32 19. L. evotis TK40341 4.34 4.10 3.71 4.42 4.11 4.34 3.09 3.399 3.01 2.24 20. Tonatia saurophila TK49885 5.06 5.21 4.98 5.21 4.81 5.77 4.35 4.262 4.25 3.63 21. T. saurophila TK46028 4.98 5.13 4.90 5.13 5.05 5.69 4.27 4.182 4.17 3.55 22. T. saurophila TK49889 5.13 5.28 5.05 5.29 5.05 5.85 4.26 4.342 4.49 3.86 23. T. bidens TK56519 4.74 4.66 4.27 4.66 4.19 5.31 3.25 3.800 3.79 3.01 24. T. bidens TK56520 4.67 4.67 4.28 4.67 4.19 5.31 3.26 3.806 3.79 3.02 25. T. bidens TK56633 4.74 4.66 4.27 4.66 4.19 5.31 3.25 3.800 3.79 3.01 August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 805

APPENDIX II.ÐExtended.

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Ð 2.48 Ð 2.79 4.21 Ð 2.71 3.90 4.21 Ð 2.17 3.18 3.73 1.27 Ð 2.24 3.25 3.80 1.65 0.81 Ð 2.17 3.17 3.73 1.57 0.74 0.67 Ð 2.09 3.09 3.65 1.49 0.66 0.59 0.07 Ð 2.17 3.17 3.73 1.57 0.74 0.22 0.59 0.52 Ð 3.40 4.02 4.76 4.67 3.95 3.86 3.95 3.87 3.79 Ð 3.32 4.10 4.68 4.60 3.87 3.79 3.87 3.79 3.71 0.22 Ð 3.63 4.26 5.00 4.92 4.19 4.10 4.18 4.10 4.02 0.66 0.44 Ð 2.94 3.41 4.61 3.90 3.18 3.09 3.18 3.10 3.02 2.70 2.93 2.93 Ð 2.94 3.41 4.53 3.90 3.19 3.10 3.18 3.10 3.02 2.63 2.86 2.86 0.00 Ð 2.94 3.41 4.61 3.90 3.18 3.09 3.18 3.10 3.02 2.70 2.93 2.93 0.15 0.15 Ð 806 JOURNAL OF MAMMALOGY Vol. 84, No. 3

APPENDIX III Tamura and Nei (1993) genetic distances calculated from the mitochondrial 12S rDNA, tRNAVal, and 16S rDNA data and expressed as percent sequence divergence. Numbers in column headings correspond to the specimens as numbered in the 1st column.

Specimen 12345678 1. Macrotus TK32021 Ð 2. Diaemus TK32625 16.75 Ð 3. Diphylla TK13514 18.14 17.23 Ð 4. Desmodus TK4764 18.89 12.38 19.95 Ð 5. Lampronycteris TK25238 16.71 16.01 18.47 17.97 Ð 6. Micronycteris TK70447 14.37 15.26 17.49 17.11 12.85 Ð 7. Macrophyllum TK19119 16.45 17.65 17.83 19.86 15.49 14.49 Ð 8. Trachops TK18829 15.83 16.68 17.48 18.77 15.17 14.67 13.37 Ð 9. Chrotopterus TK70457 15.79 16.96 18.20 18.86 16.21 14.35 14.48 13.06 10. Vampyrum TK40370 16.18 15.71 17.61 16.95 14.53 13.28 14.47 13.37 11. Phylloderma TK10201 15.75 15.98 16.90 18.27 15.51 13.12 14.45 12.99 12. Phyllostomus TK19289 14.15 15.38 16.41 16.73 14.20 12.65 13.12 10.88 13. Mimon TK25230 15.63 18.41 17.64 19.28 15.99 14.68 14.08 12.74 14. Lophostoma carrikari TK49900 14.40 15.18 16.65 16.84 14.76 12.11 13.10 11.34 15. L. brasiliense TK18834 13.77 16.50 16.68 17.23 14.87 12.92 12.72 11.32 16. L. schulzi TK18833 14.00 15.61 16.16 17.77 15.03 13.54 12.89 11.77 17. L. silvicolum TK17964 13.99 15.80 16.68 16.84 14.40 12.40 12.40 10.67 18. L. silvicolum TK56716 13.82 15.28 16.20 16.70 14.50 12.22 12.70 10.86 19. L. evotis TK49870 14.11 15.53 16.68 16.83 14.69 12.35 12.95 11.35 20. Tonatia saurophila TK49892 14.76 15.85 17.05 17.33 15.12 13.35 13.37 11.32 21. T. saurophila TK49889 14.81 15.73 16.49 17.42 15.63 13.37 13.67 12.09 22. T. saurophila TK10418 15.07 15.98 16.80 17.80 15.83 13.56 13.91 12.33 23. T. bidens TK56519 14.28 14.90 16.37 17.38 15.33 14.46 13.84 12.01 24. T. bidens TK56633 14.29 14.95 16.26 17.33 15.22 14.35 13.83 12.02 August 2003 PORTER ET AL.ÐSYSTEMATICS OF ROUND-EARED BATS 807

APPENDIX III.ÐExtended.

9 1011121314151617

Ð 11.09 Ð 13.86 13.00 Ð 11.73 11.60 10.14 Ð 14.52 14.38 12.76 10.50 Ð 13.33 12.52 11.61 9.80 12.05 Ð 12.65 12.98 11.25 9.68 11.74 6.32 Ð 13.32 12.44 12.06 10.70 12.36 6.32 6.01 Ð 12.19 12.31 10.22 9.79 11.42 6.84 6.59 5.78 Ð 12.39 12.51 10.90 9.50 11.54 6.93 6.62 5.44 1.67 13.16 12.96 11.30 10.09 11.42 6.65 6.83 5.89 2.07 13.01 13.41 11.67 11.24 12.76 11.24 11.46 11.63 10.68 13.33 13.70 12.26 11.50 12.60 11.39 12.05 12.08 11.06 13.58 13.90 12.41 11.74 12.95 11.63 12.29 12.32 11.29 13.48 12.98 12.05 9.73 12.33 10.62 10.58 10.08 9.15 13.38 12.87 11.95 9.73 12.34 10.58 10.53 10.04 9.05 808 JOURNAL OF MAMMALOGY Vol. 84, No. 3

APPENDIX III.ÐExtended.

Specimen 18 19 20 21 22 23 24 1. Macrotus TK32021 2. Diaemus TK32625 3. Diphylla TK13514 4. Desmodus TK4764 5. Lampronycteris TK25238 6. Micronycteris TK70447 7. Macrophyllum TK19119 8. Trachops TK18829 9. Chrotopterus TK70457 10. Vampyrum TK40370 11. Phylloderma TK10201 12. Phyllostomus TK19289 13. Mimon TK25230 14. Lophostoma carrikari TK49900 15. L. brasiliense TK18834 16. L. schulzi TK18833 17. L. silvicolum TK17964 18. L. silvicolum TK56716 Ð 19. L. evotis TK49870 2.19 Ð 20. Tonatia saurophila TK49892 11.04 11.47 Ð 21. T. saurophila TK49889 11.11 11.60 1.76 Ð 22. T. saurophila TK10418 11.34 11.83 1.97 0.29 Ð 23. T. bidens TK56519 8.69 9.76 9.06 9.42 9.60 Ð 24. T. bidens TK56633 8.70 9.66 9.06 9.43 9.61 0.08 Ð