TEXAS TECH UNIVERSITY Natural Science Research Laboratory Occasional Papers Museum of Texas Tech University Number 345 7 March 2017 Patterns of Morphological and Molecular Evolution in the Antillean Tree Bat, Ardops nichollsi (Chiroptera: Phyllostomidae) Roxanne J. Larsen, Peter A. Larsen, Caleb D. Phillips, HughH. Genoways, GaryG. Kwiecinski, Scott C. Pedersen, CarletonJ. Phillips, and Robert J. Baker Abstract Species endemic to oceanic islands offer unique insights into the mechanisms underlying evolution and have served as model systems for decades. Often these species show phenotypic variation that is correlated with the ecosystems in which they occur and such correlations may be a product of genetic drift, natural selection, and/or environmental factors. We explore the morphologic and genetic variation within Ardops nichollsi , a species of phyllostomid bat endemic to the Lesser Antillean islands. Ardops nichollsi is an ideal taxon to investigate the tempo of evolution in Chiroptera, as it: is a recently derived genus in the family Phyllostomidae; contains intraspecific morphological variation; and has a restricted insular distribution. To evaluate pat¬ terns of evolution in A. nichollsi , we used standard morphological analyses, in addition to ana¬ lyzing Amplified Fragment Length Polymorphisms, mitochondrial cytochrome-/?, and paternal marker zinc finger Y-chromosomal intron DNA sequence data. Our results identified a pattern that consists of two distinct evolutionarily lineages, which correspond to northern and southern islands of the Lesser Antilles. We also describe a new subspecies from the southern island of Saint Vincent. These results indicate gene flow among northern Lesser Antillean populations during the Pleistocene, and local adaptation to individual islands in the southern Lesser Antilles. Our findings can be used to further explore speciation processes within Caribbean bats and, more broadly, within species distributed across other insular systems. Key words: AFLP, Ardops , Ariteus, Caribbean, incipient species, island biogeography, Lesser Antilles, speciation, subspecies Introduction Studies of species endemic to oceanic archipela¬ gos have offered important insights into mechanisms of evolution, and have advanced evolutionary theory (Dar¬ lington 1957; MacArthur and Wilson 1967; Heaney 2007; Losos and Ricklefs 2010). Species complexes such as Caribbean Anolis and Darwin’s Galapagos finches are considered to be model organisms and pro¬ vide a foundation for understanding natural selection, adaptation, colonization, and speciation of island fauna (Roughgarden and Roughgarden 1995; Hedges 1996; Grant and Grant 2008; Pinto et al. 2008). It is within this framework that we approach Ardops nichollsi , 2 Occasional Papers, Museum of Texas Tech University a Lesser Antillean endemic. The monotypic genus Ardops is of relatively recent origin, between 1.8-2.0 million years ago (mya; Rojas et al. 2011; Baker et al. 2012), making it one of the youngest lineages of the phyllostomid bats. Interestingly,^, nichollsi exhibits a relatively large amount of intraspecific morphological variation (Jones and Schwartz 1967; Jones and Geno- ways 1973). Given its insular distribution, the forego¬ ing means that it successfully dispersed, established island populations, and then adapted to local environ¬ ments. Included within the//, nichollsi complex are five morphologically defined subspecies ( montserratensis , annectens , nichollsi , koopmani , luciae\ Fig. 1), which exhibit variation in body size across the 13 Lesser An¬ tillean islands they inhabit (Jones and Schwartz 1967; Masson et al. 1990; McCarthy and Henderson 1992; A. Jones and Schwartz 1967* St. Martin i i i 0 50 100 Saba • St. Eustatius * ^St. Kitts "l Nevis ^^Antigua A. n. montserratensis 4 Montserrat C) Guadeloupe A. n. annectens \J \ Marie-Galante 1 Dominica A. n. nichollsi A. n. koopmani \ Martinique A. n. luciae St. Lucia St. Vincent B. This study St. Martin 0 50 100 Saba St. Eustatius St. Kitts *0 Nevis ^^Antigua ^ Montserrat A. n. montserratensis <*Guadeiou P e . Marie-Galante A. n. nichollsi % Dominica Martinique A. n. koopmani A. n. luciae 0 st Lucia A. n. vincentensis 0 st Vincent c. Figure 1. Map of the Lesser Antilles showing geographic variation in Ardops nichollsi. (A) Distribution based on previous morphological analyses (*sensu Jones and Schwartz 1967; Jones and Genoways 1973; Genoways et al. 2007a; Lindsay et al. 2010). Color coding: A. n. montserratensis - brown; A. n. annectens - yellow; A. n. nichollsi - green; A. n. koopmani - purple; and A. n. luciae - blue. (B) Current taxonomy and distribution based on this study. Color coding: A. n. montserratensis - brown; A. n. nichollsi - green; A. n. koopmani - purple; A. n. luciae - blue; and A. n. vincentensis - black. (C) Photograph of Ardops nichollsi from St. Vincent (by P. A. Larsen). Larsen et al.—Evolutionary Patterns of Ardops nichollsi 3 Pedersen et al. 2003, 2005; Genoways et al. 2001, 2007a, 2007b; Lindsay et al. 2010). Geographically defined phenotypes are typically classified as subspecies, but empirical studies have questioned whether or not such geographically or ecologically defined units represent the initial stages of speciation (Mayr 1954; Pimentel 1958; Lidicker 1962; Baker and Bradley 2006; Johnsen et al. 2006; Patten and Pruett 2009). Here, we consider the varia¬ tion within Ardops as an appropriate system to begin formulating hypotheses regarding incipient speciation given the hypothesized time of origin for Ardops is recent, and the potential for unique evolutionary pres¬ sures associated with an archipelago (i.e., divergent selection, founder effects, and others). Jones and Schwartz (1967) conducted the first de¬ tailed examination of the variation within A. nichollsi. Based on cranial and external measurements, these authors identified a continuum in size and extensive secondary sexual variation within the genus, hypoth¬ esizing the populations adapted independently to envi¬ ronmental conditions on each island. Since then, few studies have specifically explored relationships within A. nichollsi or closely related genera (Greenbaum et al. 1975; Mennone et al. 1986; Carstens et al. 2004; Davalos 2007; Baker et al. 2012), with each of these studies lacking specimens from throughout the known distribution. However, our expeditions to the Lesser Antilles have increased the number of available voucher specimens and tissues of A. nichollsi (Pedersen et al. 2003,2005; Genoways et al. 2001,2007a, 2007b, 2010; Lindsay et al. 2010). To better understand the evolu¬ tionary history of A. nichollsi , we analyzed nuclear, mitochondrial, and Y-chromosomal markers in addition to morphological characters from specimens collected throughout the Lesser Antilles. These data allow for an analysis of genetic and phenotypic variation within the genus and for the formulation of hypotheses concerning the delimitation of subspecies, and incipient speciation in an archipelago. Materials and Methods Molecular methods. —Tissues were collected from natural populations of Ardops nichollsi throughout the Lesser Antilles md Ariteusflavescens from Jamaica (outgroup and sister genus representative; Appendix). Whole genomic DNA was extracted from liver or mus¬ cle tissue for all genetic analyses following standard methods (Longmire et al. 1997), or by using DNeasy Blood and Tissue Kits (Qiagen Inc., Chatsworth, Cali¬ fornia). All tissues used in DNA analyses are archived at the Genetic Resources Collection of the Natural Science Research Laboratory (NSRL) of Texas Tech University. All animals were handled following the guidelines for animal care and use established by the American Society of Mammalogists (Sikes et al. 2011; Texas Tech University IACUC permits #02217-02 and #07083-02). External primers glo7L/glo6H (Hoffmann and Baker 2001) were used to amplify 1,140 base pairs (bp) of the cytochrome b (cyt-/>) gene in three Ariteus and 47 Ardops specimens. The thermal profile consisted of 94°C for 2 min, followed by 35 cycles of denaturation at 94°C for 45 s, annealing at 49°C for 1 min, extension at 72°C for 1 min 15 s, and ended with 72°C for 10 min. All PCR products were purified using QIAquick PCR Purification Kits (Qiagen Inc., Chatsworth, California). Internal primers from Hoffmann and Baker (2001; glolL, glo5L), Smith and Patton (1991; MVZ 04), and Larsen et al. (2007b; ART 16) were used to obtain final sequences of the cyt-b gene. While sequencing cyt -b, we discovered the inadvertent amplification of a pseudogene (translocation of a mitochondrial gene into the nuclear genome) for two specimens (TK 15576, TK 129202). To obtain the full cyt-b gene sequence from the mitochondrial genome for these samples, two long-range primers (Art563_32merF [5'-GGT-ATG- GGC-CCG-ATA-GCT-TAT-TTA-GCT-GAC-CT-3']; Art765_32merR [5'-ATG-ACC-AAC-ATT-CGA- AAA-ACT-CAC-CCC-TTA-TT-3']) were developed (CDP) and used to amplify a 6.3 kilo-base fragment of the mitochondrial genome (NADH dehydrogenase 1 through cyt-/?). The complete cyt -b gene subsequently was sequenced from these amplicons. Primers from Cathey et al. (1998; LGL335F and LGF331R) were used to amplify and sequence 969 bp of an intron of the zinc finger Y-chromosome gene (ZFY) from two male Ariteus flavescens and 22 male 4 Occasional Papers, Museum of Texas Tech University Ardops nichollsi. The thermal profile consisted of 95°C for 3 min, followed by 32 cycles of denaturation at 95°C for 45 s, annealing at 50°C for 30 s, extension at 70°C for 2 min 30 s, and ended with 70°C for 5 min. PCR products were electrophoresed on a 0.8% agarose gel and excised using Qiagen Gel Extraction Kits (Qiagen Inc., Chatsworth, California). Samples were prepared for sequencing with Centri-Sep columns (Princeton Separations, Freehold, New Jersey). DNA sequencing for the cyt-b and ZFY genes was performed using ABI Big Dye v3.1 chemistry, and fragments were electro¬ phoresed on an ABI 3100-Avant Genetic Analyzer (PE Applied Biosystems, Foster City, California). Se¬ quences were verified and assembled using Sequencher v4.10.1 (Gene Codes Corporation, Ann Arbor, Michi¬ gan). To ensure correct open-reading frame, multiple sequence alignments were performed manually and further checked in MacClade v4.08 (Maddison and Maddison 2005) and/or Clustal W (Farkin et al. 2007). AFFPs were generated from three Ariteus and 47 Ardops , following the protocols of Vos et al. (1995) and McDonough et al. (2008). A labeled (6FAM fluo- rophore; Applied Biosystems, Foster City, California) selective EcoRI primer and six non-labeled selective primers (McDonough et al. 2008) were used to generate AFFPs from Ariteus fiavescens and Ardops nichollsi. The labeled fragments were detected using an ABI 3100-Avant genetic analyzer, scored for presence or absence using GeneMapper v4.0 (Applied Biosystems), and converted into a binary data matrix using GenAlEx v6.5 (Peakall and Smouse 2012). Only fragments (50-400 bp in length) with intensities larger than 100 relative fluorescence units were scored as present. Er¬ ror rates (technical error rate and observer error rate) were obtained following Bonin et al. (2004) using 26 replicated samples (approximately 9% of the overall sample size). Molecular analyses .—Phylogenetic analyses of cyt-b sequence data were performed using MEGA v5.2 (Tamura et al. 2011), MrBayes v3.2 (Ronquist et al. 2012), and Garli v2.0 (Zwickl 2006) software. Maximum-parsimony, Bayesian, and maximum-like¬ lihood analyses were used to infer cyt-b phylogenies. Maximum-parsimony analyses were performed using heuristic searches, 25 replicates of the random taxon addition option, each with random starting trees, and tree-bisection-reconnection branch swapping. Sub¬ stitution models of DNA evolution were analyzed in MEGA to determine the appropriate model for the cyX-b gene sequence data and the HKY+G model was chosen. Maximum-likelihood analyses were performed in Garli using 100 search replicates, with searches being terminated after the last topological improvement fol¬ lowing 5 x 10 6 generations. Bootstrap support values for maximum-parsimony and maximum-likelihood analyses were calculated based on 500 iterations and values > 75% were considered statistically supported. Bayesian analyses were performed using 5 million generations (1 cold and 3 incrementally heated Markov chains, random starting trees for each chain), and trees were sampled every 100 generations with a final 25% burn-in (convergence was confirmed using Tracer vl .5; Rambaut and Drummond 2007). Posterior probabilities >0.95 were considered statistically supported. Genetic distance values for cyX-b were calcu¬ lated in MEGA using the Kimura 2-parameter model (Kimura 1980), which allowed for comparisons with previous molecular studies of Ardops (Carstens et al. 2004) and with other mammals (Bradley and Baker 2001; Baker and Bradley 2006). Based on previous molecular studies, A. fiavescens is an appropriate out¬ group for the phylogenetic analyses (Baker et al. 2000, 2003,2012; Davalos 2007). Additional cyX-b sequences of Ardops were obtained from GenBank and included in the analyses (Carstens et al. 2004; see Appendix: HapA-I). The paternal ZFY gene sequence data was highly conserved (see Results; Table 1) and as such was analyzed via sequence alignment. Bayesian clustering of AFEP data was performed using STRUCTURE v2.3 (Pritchard et al. 2000) and GENELAND v4.0 (Guillot et al. 2005). STRUCTURE analyses were performed using ten iterations of200,000 replications with a burn-in of50,000 for each iteration, allowing for admixture with correlated allele frequen¬ cies. Analyses were streamlined using the StrAuto v0.3 python utility by Chhatre (2012). Our sample of Ardops consisted of individuals collected from eight islands, thus our K (cluster) values ranged from 1 to 8 for each STRUCTURE iteration. Structure Harvester v0.6 (Earl and vonHoldt 2011) was used to implement the delta K procedure of Evanno et al. (2005), where the estimated number of clusters was chosen based on the greatest Pr(X|K). GENELAND was used for spatial analyses of AFLP data with 1,000,000 itera- Table 1. Characters in the ZFY sequence data. Base pair references are given above the sequences. 44 ... ” indicates removed sequence positions that contain no polymorphisms. For Ardops nichollsi , islands are listed from north to south (the line delineates northern from southern islands). Ariteus flavescens is only known from Jamaica (delineated by double line). Each island group (north and south) is represented by a single haplotype. Sample sizes (n) from each island are listed. Larsen et al.—Evolutionary Patterns of Ardops nichollsi < H H o < H O H o H < H H o < H o H H < U =5 W •'tn 00 a ^ c g < < o o U U & t> c o u c < < < H H <3 H < o H < < .2 B CO w .'tn CO 5^ ^ 2 g H H H H 3 .-J > cn (N O o a\ o o H O H o H o H o o H u o < o .2 B CO w .'tn CO S o Q < C H H S3 i> < Jamaica 6 Occasional Papers, Museum of Texas Tech University tions, 100 thinning intervals, and a burn-in of 25%. The correlated allele frequencies model was used and the maximum population number was set to eight. We performed three runs to test for consistency for the number of populations estimated by GENELAND and we selected the run with the highest average posterior probability. Additional analyses ofAFLPdata included a principal coordinates analysis (PCoA) and an analysis of molecular variance (AMOVA), both of which were conducted using GenAlEx v6.5 (Peakall and Smouse 2012). PhiPT analyses were conducted in GenAlEx with 9,999 permutations, with PhiPT (analogous to F st ) representing the proportion of variance among populations relative to the total variance. Morphological methods. —Voucher specimens are located at the following institutions: American Museum of Natural History (AMNH); British Museum of Natural History (BMNH); Royal Ontario Museum (ROM); Texas Tech University (TTU); and University of Nebraska State Museum (UNSM). A total of 52 females and 48 males of Ardops nichollsi (indicated by 5 and $ symbols in Appendix) were used in mor¬ phological analyses. Following the definitions and methods of Hall (1946) and Genoways et al. (2007b), seven cranial measurements were recorded from adult specimens of Ardops nichollsi from throughout their geographic distribution (based on criteria from Kunz and Anthony 1982). Measurements were taken from museum specimens using digital calipers and are given in millimeters to the nearest 0.01 mm. Measurements include: GLS = greatest length of skull; CBL = con- dylobasal length; ZB = zygomatic breadth; POC = postorbital constriction; MB = mastoid breadth; MTR = maxillary toothrow length; and MM = breadth across upper molars. Statistical analyses tested for secondary sexual dimorphism (one-way multivariate analysis of variance, MANOVA) and evaluated the extent of morphologi¬ cal variability in our sample of the Ardops nichollsi complex (principal component analysis). Since the number of subspecies in A. nichollsi has been debated (Jones and Genoways 1973), a conservative approach of analyzing the data as one unit was utilized. Prin¬ cipal component analysis (PCA) does not take into account any difference between groups based on apriori classification of the sample. Overall, variation in cranial size was summarized by the first axis of the PCA (PCI). Loadings are reported to describe the direction and magnitude of measurements with their respective axis. Descriptive statistics (mean, standard deviation, and range) were obtained for all individuals from each island. Statistical analyses were performed in R statistical software (2014). Results Fifty sequences of the mitochondrial cyt-Z> gene and 289 AFLP bands were generated from Ardops and Ariteus (outgroup and sister genus representative from Jamaica). For each AFLP primer pair, an aver¬ age of 50 bands were scored. An error rate of 1.0% (3 bands of298) was estimated, with these discrepancies originating from poor amplification. Additionally, ZFY sequence data were generated from 24 males (2 Ariteus and 22 Ardops ). Cyt-b and ZFY sequences were deposited in GenBank and accession numbers for all DNA sequence data herein are presented in the specimens examined (Appendix). Phylogenetic analyses. —Sequence alignment of the complete cyt-6 gene from 47 Ardops (in addition to the nine individuals of Ardops from Carstens et al. 2003) and three Ariteus was unequivocal and without stop codons. Including the outgroup, 110 sites were variable with 79 being parsimony-informative with seven at codon position 1, three at position 2, and 69 at position 3. Maximum-likelihood analyses resulted in a single optimal tree (-InL = 2311.91); nucleotide frequencies of A = 0.292, C = 0.326, G = 0.123, and T = 0.259; a transition/transversion ratio of 10.09; and an alpha shape parameter of the gamma distribution of 0.867. Tree topologies resulting from maximum- parsimony, Bayesian, and maximum-likelihood analy¬ ses were largely congruent with differences arising at unsupported nodes. The monophyly of A. nichollsi was statistically supported; however only four clades within the species had statistical support and these did not strictly correspond to island occurrence (Fig. 2). The level of intraspecific variation among cyt-b sequences was found to be < 1.0% in A. nichollsi , whereas the Larsen et al.—Evolutionary Patterns of Ardops nichollsi 1 1 100 99 i-TK 129110 St. Eustatius - TK 161563 St. Eustatius -TK 161564 St. Eustatius -TK 161565 St. Eustatius —TK 161566 St. Eustatius -TK 161567 St. Eustatius -HapF St. Kitts -HapG Nevis i- TK 15709 Montserrat - TK 15711 Montserrat - TK 15712 Montserrat - TK 117451 Saba - TK 129126 St. Eustatius - TK 129171 Montserrat — TK 148687 Antigua - TK 161562 St. Eustatius -TK 161575 St. Eustatius - TK 161582 Montserrat - HapA St. Eustatius - HapH Nevis _r HapC St. Kitts '—HapD St. Kitts i-TK 117570 St. Eustatius -TK 129039 St. Martin —TK 165227 St. Kitts LHapE St. Kitts 0 99 r TK 129127 St. Eustatius 0.01 77 85 X HapB St. Kitts I-h -Hapl Nevis 1 100 99 ■ TK 15574 Dominica — TK 15575 Dominica ■ TK 15576 Dominica 0.97 81 84 0.99 88 90 L TK 129202 Montserrat |—TK 129062 St. Martin -TK 151290 St. Lucia TK 151307 St. Lucia -TK 161343 St. Lucia — TK 161548 St. Lucia ■TK 15602 Dominica -TK 151306 St. Lucia i-TK 151358 St. Lucia -TK 161549 St. Lucia -TK 151291 St. Lucia I-TK 161340 St. Lucia _r-TK 151308 St. Lucia '-TK 161547 St. Lucia — TK 128317 St. Vincent -TK 128445 St. Vincent -TK 144588 St. Vincent TK 144593 St. Vincent I—TK 144663 St. Vincent r TK 128314 St. Vincent -TK 128334 St. Vincent -TK 128500 St. Vincent -TK 144661 St. Vincent LTK 144670 St. Vincent Figure 2. Bayesian phylogram of 1140 base pairs of the cyt-Z> gene from 59 individuals. (Ariteusflavescens was used as an outgroup but is not shown.) Top score = Bayesian posterior probability, middle score = maximum likelihood bootstrap, bottom score = maximum parsimony bootstrap. Bootstrap support values are percentages of 500 iterations. Values for unsupported nodes are not shown. 8 Occasional Papers, Museum of Texas Tech University sequences of Ariteus averaged ~ 5.0% divergent from Ardops. The ZFY intron sequenced from male Ardops nichollsi was highly conserved across all individuals (Table 1). However, specimens collected from St. Lu¬ cia and St. Vincent were found to have the same ZFY intronic sequence as the outgroup Ariteus flavescens (currently distributed only on Jamaica). Males of A. nichollsi from northern Lesser Antillean islands (St. Eustatius, St. Kitts, Montserrat, and Dominica) dif¬ fered from the outgroup at seven nucleotide positions (5 transitions and 2 transversions; Table 1). AFLP analyses. —Our STRUCTURE analyses of 47 Ardops resulted in the identification of two clusters or groups within the AFLP sample of A. nichollsi (Figs. 3 A, B). The blue group (Fig. 3B) included individuals collected from throughout the northern Lesser Antilles (Montserrat, Saba, St. Eustatius, St. Kitts, St. Martin, and Dominica), whereas the red group (Fig. 3B) was comprised of individuals collected from St. Lucia and St. Vincent. Alternatively, GENELAND identified three groups within the AFLP data (Fig. 3C: corre¬ sponding to northern Lesser Antilles, St. Lucia, and St. Vincent, respectively). The principal coordinates analysis (PCoA) of AFLP genetic distance (Fig. 4) reinforced the GENELAND results, with three groups also corresponding to the northern Lesser Antilles, St. Lucia, and St. Vincent. The first, second, and third prin¬ cipal coordinates accounted for 39.32%, 31.91%, and 11.76% of the total variation, respectively. An analysis of molecular variance (AMOVA) of the three groups identified with GENELAND and PCo A resulted in 51% of the total variance being observed among populations and 49% within (Table 2). Pairwise PhiPT values were 0.53 (northern Lesser Antilles versus St. Lucia), 0.55 (northern Lesser Antilles versus St. Vincent), and 0.35 (St. Lucia versus St. Vincent). Morphological analyses. —Secondary sexual variation was significant (P < 0.001) for each vari¬ able among the sampled Ardops; therefore the sexes were separated in further statistical analyses. The PCA of females revealed overlap among individuals from Montserrat, St. Kitts, St. Eustatius, Guadeloupe, Martinique, Antigua, and St. Lucia, whereas Dominica and St. Vincent were separate and distinct along the first principal component. For females, PCI explained 81.7% of the variation and PC2 explained 8.3% of the variation (Fig. 5A). The variables that had the highest component loading on PCI for females were ZB, MB, MTR, and MM (Table 3); however, all seven variables were correlated with PC 1 to a similar extent and in the same direction (Fig. 5A). POC showed a very high component loading with PC2. The PCA of males revealed overlap among individuals from Montserrat, St. Kitts, St. Eustatius, Guadeloupe, Martinique, Nevis, Antigua, and St. Lucia, whereas Dominica and St. Vin¬ cent were separate and distinct along the first principal component. For males, PCI explained 87.2% of the variation and PC2 explained 5.0% of the variation (Fig. 5B). The variables that had high component loadings on PCI for males were ZB, MTR, and MM (Table 3); however, all seven variables were correlated with PC 1 to a similar extent and in the same direction just as was seen in females (Fig. 5B). PC2 had high component loadings, but they were mixed (in opposite directions) for POC and MB (Table 3). Overall, the variance along PC 1 in males and females was primarily due to cranial size, and all measurements from individuals from the northern end of the range and from St. Lucia were larger, whereas those from Dominica and St. Vincent were generally smaller. On PC2, POC appeared to explain most of the variation within males and within females; however MB is also important in the variation seen in males on PC2. Mean, standard deviation, and range are reported for individuals by island for each sex (Tables 4, 5). Larsen et al.—Evolutionary Patterns of Ardops nichollsi 9 A. St. Lucia ( St. Vincent Q .13 0 0 c < 1— 0 0 0 0 -c o CD 'o rs CO £= 0 o c CO Figure 3. (A) Bathymetric map of the Lesser Antilles. Dark grey shading represents potential extent of exposed land at last glacial maximum (sea levels -130 m below current). (B) Results of STRUCTURE analyses of AFLP data. Statistical support was recovered for two groups corresponding to the northern (blue) and southern (red) Lesser Antilles. (C) Results of GENELAND analyses of AFLP data overlaid on inset from A with black dots identifying main collecting localities. Three groups were recovered, corresponding to the northern Lesser Antilles, St. Lucia, and St. Vincent. Light colors indicate high posterior probability of group membership and dark colors indicate low posterior probability. 10 Occasional Papers, Museum of Texas Tech University O Ariteus (Jamaica) St. Martin Saba St. Eustatius St. Kitts Montserrat Dominica St. Lucia St. Vincent Figure 4. Principal coordinates analysis (PCoA) of 289 AFLP bands scored from Ariteus flavescens and Ardops nichollsi. Table 2. Summary of analyses of molecular variance (AMOVA) for AFLPs in three populations of Ardops nichollsi. Populations defined in Figures 3C and 4. Degrees of freedom (df), sum of squares (SS), means squares (MS). Significance level (P < 0.001) is based on 9,999 permutations. Source of Variation df SS MS Variation Total variation (%) PhiPT Among populations 2 157.846 78.923 5.293 51% 0.512 Within populations 44 221.729 5.039 5.039 49% Larsen et al.—Evolutionary Patterns of Ardops nichollsi 11 Table 3. PCA loadings along the first two principal compo¬ nents for Ardops nichollsi. If all loadings were equal, each would be (± 0.378). Only loadings higher than this value are bolded. ZB, MB, MTR, and MM had the highest component loadings on PCI, and POC was the highest on PC2 for female Ardops nichollsi. ZB, MTR, and MM had the highest load¬ ings on PCI, and POC and MB were highest on PC2 for male Ardops nichollsi. Character Females Males PC 1 PC 2 PC 1 PC 2 GLS -0.329 0.152 -0.343 0.000 CBL -0.352 0.000 -0.345 0.170 ZB - 0.398 0.152 - 0.391 0.000 POC -0.304 - 0.932 -0.262 0.379 MB - 0.409 0.231 -0.327 - 0.891 MTR - 0.434 0.163 - 0.469 0.151 MM - 0.401 0.000 - 0.464 0.000 Females PCI (81.7%) B. Males Figure 5. Principal component analysis (PCA) of seven morphological characters from (A) 51 female and (B) 47 male Ardops nichollsi. Polygons surround individuals from an island. The first four letters of the islands are used as labels: St. Martin (StMa), Saba (Saba), St. Eustatius (StEu), St. Kitts (StKi), Nevis (Nevi), Antigua (Anti), Montserrat (Mont), Guadeloupe (Guad), Dominica (Domi), Martinique (Mart), St. Lucia (StLu), and St. Vincent (StVi). Table 4. Summary statistics (mean and standard deviation followed by range and sample size) of skull measurements taken from 52 female Ardops nichollsi from 12 Lesser Antillean islands (Appendix). Islands are in order from north to south. Occasional Papers, Museum of Texas Tech University in ci Ov in p^ 00 o’ ci vo S oo (o' c- S vo o' in p o' p in Cl o' o C cn o' o oo cn o o' o in o o cn o o p o CN 2 -H cl o' Cl p -H ov o' +1 o o 1 o vo -H t—s o’ -H in o 1 41 p o' cn P o in 4H vo o' 41 CN O OV Ov | ci o' 1 o OV o' Cl o' 1 Cl ov in o' o o o' OV ov 1 o o Cl o’ Cl o o 9.75- ov o' o' 1 c- ov vo 00 o' ov o' ov o' OV 00 (o ci ci 00 p 00 ci in' CN (o' s in c~ in in cn p m y—\ vo cn o y—s Cl Cl o y - s y - s o 00 o p rc. o’ Cl o' cn o’ 00 o’ o o 00 n4 o' c- o’ ov H -H 00 in 41 00 +1 00 o -H o' -H 00 41 K 1 o 41 t> -H vd 2 P 1 p ov 1 00 1 p cn 1 ov 1 Cl O C ; Cl 1 in 1 c- Cl c- 00 Cl OV o 00 VD cn 00 p in c C in VD 00 00 c-' P c- c o' VD c-' 'it c- o c 04 c-' cn vd vo r-' o' c o' o’ K c VD (o ri c p* ci vo' (o' in '—' Cl '—'■ OV '— / Cl '—' . — 1 s —' p ^^ Cl v — / p C- VD in Cl o o 00 cn o cn in CN o' O o' . o' Cl o' o' p o ( — ^ n p ci o vo ci o OS o o ci ci ci ci ci l-H ■ —i (o ci ov ci vo p ci S (o' s in cn in t> cn cn y^ l -1 o 1 o y-~* CN c- Cl o i~ 1 y—y 1—1 o i- 1 in o o' cn Ci o' Cl o' Cl w o’ 00 o’ o' ( — i o w w o' o’ c- o -H vd o -H VD +1 vd O 4-1 in’ 4i vd 41 vd 1 O O 41 vd 41 in CLh 00 1 ’-i VO 1 c 1 p Cl 1 o 1 p o ,_i 1 c- 1 o cn vd ov i o c o p c- p VD 00 in' vd p o in vo o' 00 in' p vd oo in’ p m’ in in’ p in' r- m p in’ in’ in’ in’ in’ in in in’ (o' cl c p' ci (o' vo m cn 1 p in 00 VD vq p in’ 00 n o’ p o’ cn o’ >n o' in O o +1 ci 1 ov -n VO ci 1 +1 o Cl 1 © p -H P o’ Cl 1 4i P Cl 1 41 cn Cl Cl t> o t> 41 o o’ CN | 41 o ov 1 vo Ov cn p cn 1 00 o’ cn 00 o ci o 00 Cl o' Cl 00 Cl Cl o c Cl o' Cl cn Cl Cl o t> o' Cl o 00 Cl o' CN o 00 00 p OV o' o' o o OV o oo Cl Cl Cl Cl CN VD ci ci p' ci (o' (o' >n (N '—' in '—' OV '—' i '—' .— 1 % o s_y C- '—' ov N —' r-~ cn ov Cl o o ov cn 00 in ( _^ Cl cn cn C1 t/3 hP p o' in o' p p. o' cn o' p o' 00 p. o' p o’ o -H p Cl p -H P Cl -H p CM | o -H cn Cl 41 p CN 1 41 p Cl Cl o 41 cn CN 41 ci CN O 00 1 ’— _ ,_ i 1 vo p 00 1 Ov p in o | r- 1 in p CI p cn O P 1 cn CN p cn' cn 00 c- Cl P Cl Cl p p Cl o p CM cn Cl c- cn P Cl o vo cn Cl O OO Cl Cl cn Cl vo ov CN CN O ci cn' p cn cn ci Cl ci Cl Cl Cl Cl Cl Cl CN Island St. Martin Saba St. Eustatius St. Kitts Nevis Antigua Montserrat Guadeloupe Dominica Martinique St. Lucia St. Vincent Table 5. Summary statistics (mean and standard deviation followed by range and sample size) of skull measurements taken from 48 male Ardops nichollsi from 10 Lesser Antillean islands (Appendix). Islands are in order from north to south. Larsen et al.—Evolutionary Patterns of Ardops nichollsi 13 CM co 4^ o * 11 cc7 00 in in CM o' O CM o O , — s ,, — s 14 .—1 e—\ 1—1 o c- / — s CO o o o' CM o CM o' CO o' op o o’ o' o o -H o' 4 o’ o o 4 o' 4 o’ 4 00 o 4 o' 4 2 | 4 | oo CO o | o o 1 o o 1 CM 00 o’ o’ in 1 p c- 4 o’ o o CO o' «n o o o O o o o o’ o CO 00 o' 00 o' o' o' o' 00 o' in' o Co' 4) 4^ 4 CO 4 00 in c~ o o o ,_ i O t -s /-s ’—i in o o f -s o o o' •n o' o * * ' , o' 4 o’ o CM o’ in 4^ o’ 4 o H 41 o’ 4 o' o CO 4 o 4 O- 4 o' o 4 o' 4 2 4 1 co 1 n o' o' CO o’ 00 vd o' o' o' o’ o' VJD in' 0 s 4 s CO in o vo '—' o s —' o '— / 4 ' _ , 4 '—' CM '— / m CM o CO o CO CO (—s 00 CM 4 CM o' vo o' o P VO o' in o' in o' 4 O pq 41 CM 4 CM o 00 4 CM 4 CM 4 p o 4 CM 4 2 > I | __ | CM CM CM | o 1 1 4 1 >n CO | o \l n CM CO o’ 4 O CM CM O CM O CM CM CM CM CO CM CM CM CM CO CO o’ CM CM CM CM CM CM CM Island St. Eustatius St. Kitts Nevis Antigua Montserrat Guadeloupe Dominica Martinique St. Lucia St. Vincent 20.27-20.82(5) 17.57- 17.93(4) 12.61 -13.54(5) 5.30-5.77(5) 10.34-10.95(5) 6.15-6.34(5) 8.21 -8.44(5) 14 Occasional Papers, Museum of Texas Tech University Taxonomic Review Our evaluation of the morphologic and molecular variation in Ardops nichollsi leads to the recognition of an undescribed subspecies. However, we do not have complete datasets for all known populations of Ardops nichollsi, ; therefore this is not a comprehensive taxonomic review. This review will show where future data can be placed to complete this work. The last revision of bats of the genus Ardops was by Jones and Schwartz (1967) based on 37 specimens from seven of the Lesser Antillean islands, whereas we had 100 specimens from 12 islands available for study. Jones and Schwartz (1967) recognized a single species with five subspecies, one of which they described as new. Family Phyllostomidae Gray, 1825 Subfamily Stenodermatinae Gervais, 1856 Ardops nichollsi vincentensis R. J. Larsen, Genoways, and Baker, new subspecies Ardops nichollsi luciae Jones and Schwartz, 1967, Proceedings of the United State National Mu¬ seum, 124(3634):9-10, report of a single specimen from “St. Vincent: no specific locality.” Holotype. —Adult male, with skin, skull, and tissue samples (TK 144588). TTU 105628, from Colonarie River, 1 km S, 2.4 km W South Rivers, 248 m, Charlotte Parish, island of St. Vincent, St. Vincent and the Grenadines, Lesser Antilles; obtained by Hugh H. Genoways on 29 July 2005, original number 6407A. Deposited at the Natural Science Research Laboratory, Museum of Texas Tech University, Lubbock, Texas. Measurements of holotype. —Total length, 65; length of hind foot, 13; length of ear, 13; weight, 13.6; length of forearm, 40.7; greatest length of skull, 20.4; condylobasal length, 17.6; zygomatic breadth, 12.8; interorbital constriction, 5.5; postorbital constriction, 5.3; mastoid breadth, 10.8; palatal length, 4.6; length of maxillary toothrow, 6.3; and breadth across upper molars, 8.2. Distribution. —Known only from the island of St. Vincent. Diagnosis. — Specimens from St. Vincent form a statistically supported clade based on AFLP data (Figs. 3,4); male and female Ardops nichollsi from St. Vincent are the smallest-sized members of the genus in cranial measurements, approached in size only by individuals of A. n. nichollsi from Dominica (Fig. 5, Tables 4,5). Males possess the southern haplotype for the very conservative ZF Y intronic sequence, which is shared with A. n. luciae (St. Lucia) and Ariteus flave- scens (Jamaica). Remarks. —The newly delineated subspecies is distinguished from other populations by both molecular and morphologic characteristics. Although males from St. Vincent have the southern haplotype of the ZFY intronic sequence (Table 1), males and females have a combination of AFLPs that distinguish them at a mo¬ lecular level from other subspecies of Ardops nichollsi (Figs. 3,4; Table 2). Morphologically this new subspe¬ cies needs only comparison with the geographically adjacent population of A. n. luciae from St. Lucia. In six of the seven cranial measurements for males (except POC) and five of seven for females (except POC and MB) there is no overlap in the range of variation of these two taxa (Tables 4, 5). The principal component analyses (Fig. 5, Table 3) confirm this relationship, with A. n. vincentensis at the furthest right position along PC 1 indicating these bats had the smallest skulls in our study. The one taxon that approaches the position of the sample of A. n. vincentensis is A. n. nichollsi from Dominica, but their projections do not overlap and the cranial measurements confirm this relationship (Tables 4,5). Although our study only included a single female from Dominica, the measurements for this individual are larger and outside the range of variation for three cranial measurements of females from St. Vincent (ZB, MTR, and MM; Table 4). The ranges of measurements for samples of males from the two islands do not overlap for two cranial measurements (MB and MM; Table 5) with those from Dominica being the larger. Although these two taxa are close morphologically, they differ at the molecular level and are separated geographically by two intervening islands (St. Lucia and Martinique; Fig. 1). Jones and Schwartz (1967) had only a single male from St. Vincent available for study and it had a fragmentary skull and both forearms broken. Based on this limited material, they surmised, “bats on St. Larsen et al.—Evolutionary Patterns of Ardops nichollsi 15 Vincent may be smaller than those of any described race of A. nichollsi A Because of this limited informa¬ tion, however, they assigned the specimen tentatively to A. n. luciae. We have surveyed other Lesser Antil¬ lean islands to the east and south of St. Vincent for bats, but Ardops was not recovered from any of these islands, including Barbados (Genoways et al. 2011), The Grenadines (Genoways et al. 2010), and Grenada (Genoways et al. 1998). Given that bats of the genus Ardops appear to have a geographic distribution that follows Koopman’s Line (Genoways et al. 2010), we do not expect members of the genus to be found on The Grenadines or Grenada. Genoways et al. (2011) stated: “it is our working hypothesis that the relatively young geological age of Barbados and the distance separating Barbados from neighboring islands have dually con¬ tributed to the small chiropteran fauna of Barbados.” Among the species of bats “missing” from Barbados was Ardops nichollsi , which we do not expect will be documented from there. Etymology. —It is our pleasure to name this unique new subspecies, vincentensis, in recognition of its home on the beautiful island of St. Vincent. Specimens examined [Type Series](10). —ST. VINCENT: Charlotte Parish: Colonarie River, 1 km S, 2.4 km W South Rivers, 13°14T0.4" N, 61°09'52.7" W, 248 m (28 July 2005: male, TK 144593, TTU 105632; male, TK 144588, TTU 105628 [holotype]); Golden Grove, 1.5 km N, 2.7 km W Mesopotamia, 13°11'58.3" N, 61°11'32.7" W, 410 m (26 May 2006: male, TK 128314, TTU 105316; female, TK 128317, TTU 105319); La Soufriere Trailhead, 3.7 km W Or¬ ange Hill, 13°19'0.2" N, 61°09'9.5" W, 420 m (1 June 2006: female, TK 128445, TTU 105367). St. Andrew Parish: Parrot Lookout, Vermont Nature Trail, 2.3 km N, 1.75 km E Vermont, 13°13'20.2"N, 61°12'43.4" W, 496 m (1 August 2005: male, TK 144661, TTU 105758; female, TK 144663, TTU 105760; female, TK 144670, TTU 105767); Mt. St. Andrew, 0.35 km S, 3 km E Pembroke, 13°11T2.6" N, 61°13'07.0" W, 501 m (4 June 2006: male, TK 128500, TTU 105479). St. David Parish: Morgan Woods, 0.4 km N, 2.4 km E Richmond, 13°18'28.9" N, 61°12'27.9" W, 523 m (27 May 2006: female, TK 128334, TTU 105524). Specimens are in fluid, with skulls removed, and tissue samples (TK). Ardops nichollsi nichollsi (Thomas, 1891) Stenoderma nichollsi Thomas, 1891, Annals and Magazine of Natural History, series, 6, 7:529. Holotype. —Adult female in fluid with skull re¬ moved. BMNH 91.5.14.4, from an unknown locality on Dominica, Lesser Antilles; obtained by H. A. A. Nicholls. Measurements of holotype. —Total length, 58; length of ear, 12; length of forearm, 45.7; greatest length of skull, 22.2; condylobasal length, 18.7; inter¬ orbital constriction, 6.7; postorbital constriction, 5.7; palatal length, 4.8; and length of maxillary toothrow, 7.0. Distribution. —Known only from the island of Dominica. Remarks. —Thomas (1891) described this new species based on a single female with a slightly dam¬ aged skull from the island of Dominica as a new species of genus Stenoderma , but Miller (1906) later created the new genus Ardops to include the Lesser Antillean taxa of this group. The nominate subspecies is characterized by grouping with northern populations of A. nichollsi , in which males possess the northern haplotype of ZFY intronic sequence (Table 1), and in which AFLP analy¬ ses confirm a statistically supported group confined to the Lesser Antilles from Dominica northward (Figs. 3, 4). On the other hand, the Dominican population can be distinguished from the other northern taxon, montserratensis, by its overall small cranial size. In individual cranial measurements, the values for the two taxa do not overlap in five measurements for males and all seven measurements for females (Tables 4, 5). PCI reveals this relationship with A n. nichollsi to the right of the projection, and all other populations except A. n. vincentensis to the left of the projection (Fig. 5). It is not possible to fully assess the relationship of A n. nichollsi with A. n. koopmani from Martinique because of the lack of data for the latter, but comparing cranial measurements, the range of variation in males from Dominica falls below the size of the male from 16 Occasional Papers, Museum of Texas Tech University Martinique for five measurements (except POC and MB), whereas the female from Dominica is smaller in all measurements than the female from Martinique (Tables 4,5). Based on this information, A n. koopmani is likely distinct, at least at the subspecific level, from A. n. nichollsi. The relationship of A. n. nichollsi to the newly described A. n. vincentensis is discussed in the account for the latter subspecies. Ardops nichollsi koopmani Jones and Schwartz, 1967 Ardops nichollsi koopmani Jones and Schwartz, 1967, Proceedings of the United States National Mu¬ seum, 124(3634): 11. Holotype. —Adult female in fluid with skull removed. AMNH 213951, from near Balata, Fort-de- France, Martinique, France, Lesser Antilles; obtained by Harry Beatty and Peter Martin on 18 March 1967, original no. 656. Measurements of holotype. —Total length, 65; length of hind foot, 14; length of ear, 16; length of forearm, 50.1; greatest length of skull, 23.5; condylo- basal length, 20.7; zygomatic breadth, 15.9; interorbital constriction, 7.0; postorbital constriction, 6.0; mastoid breadth, 12.6; palatal length, 5.2; length of maxillary toothrow, 7.7; and breadth across upper molars, 10.5. Distribution. —Known only from the island of Martinique. Remarks. —Because we lacked tissues from members of this subspecies, we were unable to per¬ form analyses of ZFY intronic sequence or AFLPs, leaving us only limited morphologic data to assess the relationships of A. n. koopmani. Jones and Schwartz (1967) in their original description of A. n. koopmani examined four individuals and presented the cranial measurements of the same two individuals presented in our Tables 4 and 5. Jones and Schwartz (1967) stated: “Ardops nichollsi koopmani differs from populations of the species on adjacent islands (A. n. nichollsi to the north on Dominica and A. n. luciae to the south on St. Lucia) in being considerably larger.” Other character¬ istics that they cited include well-developed sagittal crest, relatively narrow skull, and narrow molariform teeth. Examination of the PC A (Fig. 5) reveals that the specimens from Martinique fall with a group of bats with large-sized skulls from St. Lucia and islands from Guadeloupe northward. Compared to Dominica, the male from Martinique falls above the range of varia¬ tion for five measurements and within for POC and MB (Table 5), whereas the female from Martinique is larger in all measurements (Table 4); compared to St. Lucia, the Martinique male falls within the range for four measurements and below the range for CBL, POC, and MB (Table 5), whereas the Martinique female falls within the range for only three measurements and above the range for CBL, ZB, MB, and MM (Table 4). These results indicate that specimens of A. n. koopmani are larger than the individuals of A. n. nich¬ ollsi from Dominica, which is in agreement with the PCA(Fig. 5) and the description by Jones and Schwartz (1967). The morphologic results for A. n. luciae are not so clear, however, because our male specimen from Martinique falls within or below the range of values of males from St. Lucia, whereas the female from Martinique falls within or above the range of values of females from St. Lucia. These discordant results are undoubtedly, at least in part, because of having only two individuals from Martinique to analyze (Tables 4, 5). Due to the lack of molecular data and the ambiguous morphologic results, we tentatively continue to recog¬ nize this subspecies. As future data become available, the relationship between A n. koopmani and other taxa will need to be further evaluated. Ardops nichollsi luciae (Miller, 1902) Stenoderma luciae Miller, 1902, Proceedings of the Academy of Natural Sciences of Philadelphia, 54:407. Holotype. —Adult female in fluid with skull removed. NMNH 110921, from an unknown locality on the island of St. Lucia, Lesser Antilles; obtained by H. S. Branch on 4 February 1901. Measurements of holotype. —Total length, 65; length of hind foot, 12.6; length of ear, 18; length of forearm, 48.0; greatest length of skull, 23.2; condylo- basal length, 20.3; zygomatic breadth, 14.8; interorbital constriction, 6.6; postorbital constriction, 5.7; mastoid breadth, 12.0; palatal length, 5.5; length of maxillary toothrow, 7.5; and breadth across upper molars, 10.4. Larsen et al.—Evolutionary Patterns of Ardops nichollsi 17 Distribution. —Known only from the island of St. Lucia. Remarks. —This subspecies easily is diagnosed based on males possessing the southern haplotype of the ZFY intronic sequence, which it shares with A. n. vincentensis and Ariteus flavescens (Jamaica; Table 1), in addition to the AFLP analyses where St. Lucia specimens are isolated or grouped with St. Vincent specimens (Figs. 3,4). The large skull size of individu¬ als from St. Lucia, easily distinguishes them from A. n. vincentensis on the island of St. Vincent to the south. In six measurements for males (except POC) and five for females (except POC and MB), there is no overlap in the range of measurements of these two subspecies (Tables 4, 5). As discussed above, the relationships of A. n. koopmani on Martinique to A. n. luciae are not clear at present because of the limited data available from Martinique. If these two subspecies ultimately prove to be indistinguishable, A. n. luciae would be the senior synonym. Ardops nichollsi montserratensis (Thomas, 1894) Stenoderma montserratense [sic] Thomas, 1894, Proceedings of the Zoological Society of London, 1894:132-133. Ardops annectens Miller, 1913, Proceedings of the Biological Society of Washington, 26:33. Holotype. —Adult male in fluid with skull re¬ moved. BMNH 94.1.9.1, from an unknown locality on the island of Montserrat, Lesser Antilles; obtained by Joseph Sturge. Measurements of holotype. —Total length, 69; length of ear, 16.5; length of forearm, 51.5; greatest length of skull, 23.8; condylobasal length, 20.8; zy¬ gomatic breadth, 15.8; interorbital constriction, 6.9; postorbital constriction, 6.0; mastoid breadth, 10.7; palatal length, 5.1; length of maxillary toothrow, 7.5; and breadth across upper molars, 10.2. Distribution. —Known from the islands of Anti¬ gua, Guadeloupe, Marie-Galante, Montserrat, Nevis, Saba, St. Eustatius, St. Kitts, and St. Martin/St. Maarten in the Lesser Antilles. Remarks. —This subspecies may be distinguished from others in the species complex based on males possessing the northern haplotype of the ZFY intronic sequence (shared by Ardops populations from Domi¬ nica northward; Table 1) and AFLP analyses indicating the northern populations were statistically significant when compared to St. Lucia and St. Vincent (Figs. 3, 4). Therefore, based on molecular data from our study, A. n. montserratensis and A. n. nichollsi cannot be distinguished; however, based on the morphologic data the two can be easily separated as shown in the PC A (Fig. 5). In individual cranial measurements the values for the two taxa do not overlap in five mea¬ surements for males and all seven measurements for females as follows (smallest montserratensis vs. largest nichollsi , males followed by females): greatest length of skull, 21.3 vs. 20.0, 22.8 vs. 21.9; condylobasal length, 18.9 vs. 18.2,19.8 vs. 18.7; zygomatic breadth, 14.2 vs. 14.1, 14.9 vs. 14.3; postorbital breadth, only female comparisons 5.6 vs. 5.5; mastoid breadth, only female comparisons 11.9 vs. 11.4; length of maxillary toothrow, 6.7 vs. 6.5; and breadth across upper molars, 9.3 vs. 8.8, 9.8 vs. 9.4. Although we do not have molecular data for Ardops from Guadeloupe, the position of the island between Dominica and the northern Lesser Antilles leads us to believe the male bats from Guadeloupe will possess the northern haplotype for the ZFY intronic sequence and males and females will have the northern Lesser Antillean AFLP pattern. The PCA results for A. n. annectens reveals that both males and females clus¬ tered to the left side of PC 1 overlapping broadly with samples from islands to the north of Guadeloupe and from St. Lucia (Fig. 5). These samples have individuals with an overall larger skull size, whereas those samples to the right side of the projection from Dominica and St. Vincent contain the individuals with an overall smaller skull size. The samples from Guadeloupe ex¬ tended further to the right of the projection than other northern samples, but the Guadeloupe samples were clearly positioned with the northern group. Examina¬ tion of Tables 4 and 5 showed that mean values for the samples from Guadeloupe had, or were among those with, the lowest mean values. Individual tree bats on Guadeloupe were slightly smaller than bats from other islands in the northern Lesser Antilles, but as the PCs showed the major morphological break was between 18 Occasional Papers, Museum of Texas Tech University Guadeloupe and Dominica. These insights lead us to place A. n. annectens , originally described by Miller (1913) from Guadeloupe, as a junior synonym of A. n. montserratensis, which was described 19 years earlier by Thomas (1894) based on a specimen from the island of Montserrat. There is a population of A. nichollsi on the small island of Marie-Galante situated 27 km south-southeast of Guadeloupe and 30 km northeast of Dominica, which places the population geographical between A. n. montserratensis and A. n. nichollsi. We did not have specimens from Marie-Galante available for our study, but six cranial measurements (four males and two females) were recorded by McCarthy and Henderson Genetic variation in Ardops nichollsi.—We compared our cyt-b results with a previous molecu¬ lar analysis of northern Lesser Antillean A. nichollsi (Carstens et al. 2004) who found support for genetically distinct lineages from northern Lesser Antillean islands; however, their sample consisted of only the widespread subspecies A. n. montserratensis from three of these islands (St. Eustatius, St. Kitts, and Nevis). With their findings, Carstens et al. (2004) suggested this northern population of Ardops was the result of a single founding event and the individual island populations had com¬ pleted lineage sorting. With a much broader molecular sample (10 of 13 islands and four of five subspecies), we found that the islands in the north shared mitochon¬ drial haplotypes (suggesting incomplete lineage sorting and/or maternal gene flow, Fig. 2), whereas populations from Dominica, St. Lucia, and St. Vincent did not share mitochondrial haplotypes (suggesting complete lineage sorting). Our ZFY intron sequence data from male A. nichollsi indicated a northern-southern island split, where Ardops specimens from St. Fucia and St. Vincent share ZFY intronic sequence with Ariteus flavescens and those from the northern islands have their own distinct sequence (Table 1). These data could be interpreted as either the ancestral sequence for the ZFY intron has been conserved within southern Fesser Antillean Ardops , or alternatively, it diverged but then returned to the ancestral nucleotide sequence. Finally, our nuclear AFFP data also support the observation that there is a distinct cluster in the northern Fesser Antilles (1992) in their initial report of Ardops from the island. The range of measurements for the four males and the measurements of the two females are as follows: g reat- est length of skull, 21.95-22.85,22.5,23.3; zygomatic breadth, 14.6-15.2, 14.65, 15.5; postorbital constric¬ tion, 5.4-5.8, 5.3, 5.8; mastoid breadth, 11.7-12.1, 12.05, 12.75; length of maxillary toothrow, 7.0-7.35, 7.55,7.55; and breadth across upper molars, 9.2-9.65, 9.95,10.1. All of these values, except for POC, exceed the values of our sample from Dominica (Tables 4, 5) and fall within or near the range of our sample from Guadeloupe. We assign the specimens from Marie- Galante to A. n. montserratensis , which makes the population the southern-most for this subspecies. as well as two separate clusters in the south, one from St. Fucia and one from St. Vincent (Figs. 3,4). We have identified a consistent division between northern and southern lineages with our multiple molecular marker approach (paternal, maternal, and nuclear markers), thus evidence for current and potentially distinct evo¬ lutionary trajectories within A. nichollsi is high. We did not have genetic samples from annectens (Guade¬ loupe, Marie-Galante) or koopmani (Martinique), and representatives from both islands will be needed before a complete taxonomic revision of A. nichollsi can be made and to determine if there are other signatures of genetic isolation. Morphological variation in Ardops nichollsi.— Earlier authors have attempted to describe the morpho¬ logical variation within Ardops, resulting in a complex taxonomic history fox Ardops nichollsi (Thomas 1891, 1894; Miller 1902, 1906, 1913; Allen 1942; Hall and Kelson 1959; Jones and Schwartz 1967; Jones and Genoways 1973). These studies mainly focused on the size variation within Ardops and the distinctiveness of males and females, with females being significantly larger (Allen 1942; Hall and Kelson 1959; Jones and Schwartz 1967). Our data confirm cranial size varia¬ tion among populations of Ardops , as both males and females from Dominica and St. Vincent are smaller than individuals from the other islands (Fig. 5). Interest¬ ingly, specimens of A. nichollsi luciae from St. Fucia are more similar in morphology to specimens from Larsen et al.—Evolutionary Patterns of Ardops nichollsi 19 the northern Lesser Antilles. Therefore, a taxonomic assessment of the genus based strictly on morphology would result in conflicting assemblages with respect to genetic lineages and island occurrence, and indicates that the morphological variation within Ardops is likely plastic and related to ecological and demographic fac¬ tors. Patterns of evolution in Ardops.—When con¬ sidering the historical diversity of the short-faced bats (» Stenoderma , Phyllops , and Ariteus) in the Greater Antilles (fossil records from Koopman and Williams 1951; Koopman 1968; Steadman et al.1984; Mancina and Garcia-Rivera 2005) and the sister relationship of Ardops to Ariteus (Genoways 2001; Baker et al. 2003; Genoways et al. 2005), it is likely that the most recent common ancestor of Ardops nichollsi was of Greater Antillean origin. This would suggest a stepping-stone colonization pattern of the Lesser Antilles in a general north to south direction by this Ardops ancestor. If this hypothesis were accurate, then it would be expected that the southernmost populations (St. Lucia and St. Vin¬ cent) would be the result of a more recent colonization; however, it appears there has been sufficient time for identifiable genetic lineages to develop, corresponding geographically to St. Lucia and St. Vincent (Figs. 2^1). This pattern of isolation was not found in the north¬ ern Lesser Antilles, where those populations may be relatively older (based on a potential Greater Antillean origin). Indeed, none of the northern Antillean island populations is strictly monophyletic and the genetic data indicate these populations likely have undergone short periods of isolation followed by periods of disper¬ sal and subsequent gene flow (Figs. 2—4 ). However, the congruencies in our molecular datasets provide strong evidence of a period of enhanced geographic isolation whereby gene flow between the northern and southern Lesser Antillean islands was restricted, especially between Dominica and both St. Lucia and St. Vincent. Geographic isolation of Lesser Antillean bats during the Pleistocene epoch has been hypothesized to have contributed to other recent speciation events (Larsen et al. 2010, 2011). Given the genetic patterns observed within Ardops , in combination with a Pleistocene origin of the genus (Rojas et al. 2011; Baker et al. 2012), we hypothesize that Pleistocene environmental conditions likely contributed to the presence/absence of gene flow among island populations of Ardops. For example, at last glacial maximum, sea levels were likely -130 meters lower (Clark and Mix 2002; Clark et al. 2009) than contemporary levels and inter-island distances would have been reduced in most cases (Fig. 3 A). The northern Lesser Antillean islands sit on several large banks and ridges (Barbuda, Saba, St. Kitts, and St. Martin banks; Genoways et al. 2007a, 2007b), much of which would have been exposed during the last glacial maximum (22,000-19,000 years ago; Clark and Mix 2002; Clark et al. 2009; Fig. 3A). Gene flow in the northern Lesser Antilles would have been facilitated by the exposed land during the last glacial maximum, while water gaps in the southern chain of islands (Fig. 3A; Hill 1905; Steadman et al. 1984; Pregill et al. 1994; Morgan 2001; Genoways et al. 2010) may have played a role in the isolation of A. nichollsi and reduced the potential for gene flow in this part of its distribution (supported by AFLP data, Fig. 4). It is also important to consider the unique life his¬ tory traits of Ardops and how these could be reflected in our data. Specifically, Ardops is known to be an obligate tree rooster that typically travels relatively short distances when foraging (when compared to other Caribbean bat species such as Artibeus jamaicensis and Brachyphylla cavernarunr, Jones and Schwartz 1967; Jones and Genoways 1973). Our research has shown that Antillean tree bats may inhabit relatively small forest patches on each island, for example, Saba is 12 square km, but Ardops occupies only about four square km; St. Martin is about 85 square km, Ardops occupies only about 1.5 square km; Antigua is about 279 square km, but only about 22 square km, or about 8% of the island, is potential habitat for Ardops (Genoways et al. 2007a; Lindsay et al. 2010). Thus, fluctuations in avail¬ able forest habitat would impact the viability of Ardops populations on each respective Antillean island. The life history traits of Ardops may contribute to slower recovery time once a population suffers an ecological disturbance, perhaps arising from hurricane activity, droughts, and volcanic eruptions (Pedersen et al. 2010). Additionally, smaller populations of Antillean tree bats could be subject to localized extinction events, with subsequent reinvasion from adjacent islands to maintain viable populations. These events could also contribute to the morphological similarity and low level of genetic diversity observed among the northern Lesser Antillean islands, and separation from the southern 20 Occasional Papers, Museum of Texas Tech University islands. Similar patterns appear in the distribution of other volant species from multiple taxonomic groups throughout the archipelago (e.g., butterflies, Drosophila and birds; Scott 1972; Seutin et al. 1994; Davies and Smith 1998; Hunt et al. 2001; Davies and Bermingham 2002; Wilder and Hollocher 2003). Finally, we have not encountered Ardops on low-lying islands with arid-adapted vegetation, i.e., Anguilla (59 m; Genoways et al. 2007c), Barbuda (42 m; Pedersen et al. 2007), and St. Barts (281 m; Larsen et al. 2007a). We do not expect Ardops to be found on these three islands, as they lack appropriate habitat for Antillean tree bats. Conversely, we have captured Ardops on islands with elevations of at least 250 m that typically include closed canopy evergreen seasonal forests (Beard 1949). Conclusions It is our hypothesis that the populations of Ardops nichollsi evolved primarily on the islands of the Lesser Antillean Faunal Core (Genoways et al. 2001). On these islands from Guadeloupe to St. Vincent, the populations have been differentiating at both molecular and morphologic levels. This level of differentiation was not detected in populations on the islands to the north of Guadeloupe. Collectively, our morphological and molecular data have identified several interesting macroevolutionary patterns within A. nichollsi. In particular, our genome scan data may provide evidence for the initial stages of speciation, with the northern Lesser Antillean populations being on a separate evolutionary trajectory with respect to southern Lesser Antillean populations. Monophyly is not observed in all datasets and thus incomplete lineage sorting may account for some of the patterns observed. Similar conflicting patterns in morphological and genetic datasets have been identified in a number of bat genera, such as Artibeus (Marchan-Rivadeneira et al. 2012; Larsen et al. 2013), Eumops (McDonough et ah 2008; Baker et al. 2009), Platyrrhinus (Velazco and Patterson 2008), and Myotis (Larsen et al. 2012). Such conflict among multi-source data can be attributed to incipient speciation events (Coyne and Orr 2004) and/or incomplete lineage sorting (Funk and Omland 2003; McGuire et al. 2007). Additional research using advanced techniques such as RAD-seq and/or whole genome sequencing are required to further explore the genetics underlying phenotypic plasticity and the speciation dynamics of Ardops. Acknowledgments Many colleagues from a number of institutions assisted with the collection of specimens that were used in this study, and we greatly appreciate their efforts. We would like to thank the following museums for access to specimens: American Museum of Natural History (AMNH), British Museum of Natural History (BMNH), National Museum of Natural History (NMNH), Royal Ontario Museum (ROM), Texas Tech University (TTU), and the University of Nebraska State Museum (UNSM). We especially thank Heath Gamer and Kathy MacDonald (Natural Science Research Laboratory at the Museum of TTU) for assisting in archiving valu¬ able specimens and tissues. Christopher Blair provided insightful suggestions that helped to improve the manu¬ script and J. D. Pampush assisted with writing R code. We appreciate the assistance of the governments and institutions from many localities in the Caribbean for allowing us to obtain tissue and specimen vouchers, as well as conduct research. Lastly, we thank the Texas Tech Association of Biologists Graduate Student Mini¬ grant, the Department of Biological Sciences at Texas Tech University, James Sowell, and the Biological Database for financial support of research and travel. Larsen et al.—Evolutionary Patterns of Ardops nichollsi 21 Literature Cited Allen, G.M. 1942. Extinct and vanishing mammals of the western hemisphere: with the marine species of all the oceans. New York: Special Publication, American Committee for International Wild Life Protection 11:1-620. Baker, R. J., O. R. P. Bininda-Emonds, H. Mantilla-Meluk, C. A. Porter, and R. A. Van Den Bussche. 2012. Molecular timescale of diversification of feeding strategy and morphology in New World leaf-nosed bats (Phyllostomidae): a phylogenetic perspective. Pp. 385-409 in Evolutionary history of bats: fossils, molecules and morphology (G. F. Gunnell and N. B. Simmons, eds.). New York: Cambridge University Press, 572 pp. Baker, R. J., and R. D. Bradley. 2006. Speciation in mam¬ mals and the genetic species concept. Journal of Mammalogy 87:643-662. Baker, R. J., S. R. Hoofer, C. A. Porter, and R. A. Van Den Bussche. 2003. Diversification among New World leaf-nosed bats: an evolutionary hypothesis and classification inferred from digenomic congruence of DNA sequence. Occasional Papers, Museum of Texas Tech University 230:1-32. Baker, R. J., M. M. McDonough, V. J. Swier, P. A. Larsen, J. P. Carrera, and L. K. Ammerman. 2009. New spe¬ cies of bonneted bat, genus Eumops (Chiroptera: Molossidae) from the lowlands of western Ecuador and Peru. Acta Chiropterologica 11:1-13. Baker, R. J., C. A. Porter, J. C. Patton, and R. A. Van Den Bussche. 2000. Systematics of bats of the family Phyllostomidae based on RAG2 DNA sequences. Occasional Papers, Museum of Texas Tech Uni¬ versity 202:1-16. Beard, J. S. 1949. The natural vegetation of the Windward and Leeward islands. Oxford Forestry Memoir 21:1-192. Bonin, A., E. Bellemain, P. B. Eidesen, F. Pompanon, C. Brochmann, and P. Taberlet. 2004. How to track and assess genotyping errors in population genetics studies. Molecular Ecology 13:3261-3273. Bradley, R. D., and R. J. Baker. 2001. A test of the genetic species concept: cyt-b sequences and mammals. Journal of Mammalogy 82:960-973. Carstens, B. C., J. Sullivan, L. M. Davalos, P. A. Larsen, and S. C. Pedersen. 2004. Exploring population and genetic structure in three species of Lesser Antillean bats. Molecular Ecology 13:2557-2566. Cathey, J. C., J. W. Bickham, and J. C. Patton. 1998. Intro- gressive hybridization and nonconcordant evolu¬ tionary history of maternal and paternal lineages in North American deer. Evolution 52:1224-1229. Chhatre, V. E. 2012. StrAuto vO.3.1: a python program - automation of structure analysis. Clark, P. U., A. S. Dyke, J. D. Shakun, A. E. Carlson, J. Clark, B. Wohlfarth, J. X. Mitrovica, S. W. Hostetler, and A. M. McCabe. 2009. The last glacial maximum. Science 325:710-714. Clark, P. U., and A. C. Mix. 2002. Ice sheets and sea level of the last glacial maximum. Quaternary Science Reviews 21:1-7. Coyne, J. A., and H. A. Orr. 2004. Speciation. Sinauer As¬ sociates, Sunderland, MA. 545 pp. Darlington, P. J., Jr. 1957. Zoogeography: the geographical distribution of animals, 1 st edition. John Wiley and Sons, New York. 675 pp. Davalos, L. M. 2007. Short-faced bats (Phyllostomidae: Stenodermatina): a Caribbean radiation of strict frugivores. Journal of Biogeography 34:364-375. Davies, N., and E. Bermingham. 2002. The historical bioge¬ ography of two Caribbean butterflies (Lepidoptera: Heliconiidae) as inferred from genetic variation at multiple loci. Evolution 56:573-589. Davies, N., and D. S. Smith. 1998. Monroe revisited: a survey of West Indian butterfly faunas and their species-area relationship. Global Ecology and Biogeography Letters 7:285-294. Earl, A., and B. M. vonHoldt. 2011. STRUCTURE HAR¬ VESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conservation Genetics Resources 4:359-361. Evanno, G., R. Regnaut, and J. Goudet. 2005. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecol¬ ogy 14:2611-2620. Funk, D. J., and K. E. Omland. 2003. Species-level para- phyly and polyphyly: frequency, causes, and con¬ sequences, with insights from animal mitochondrial DNA. Annual Review of Ecology, Evolution, and Systematics 34:397-423. Genoways, H. H. 2001. Review of Antillean bats of the genus Ariteus. Occasional Papers, Museum of Texas Tech University 206:1-11. 22 Occasional Papers, Museum of Texas Tech University Genoways, H. H., R. J. Baker, J. W. Bickham, and C. J. Phil¬ lips. 2005. Bats of Jamaica. Special Publications, Museum of Texas Tech University 48:1-155. Genoways, H. H., G. G. Kwiecinski, P. A. Larsen, S. C. Pedersen, R. J. Larsen, J. D. Hoffman, M. de Silva, C. J. Phillips, and R. J. Baker. 2010. Bats of the Grenadine islands, West Indies, and place¬ ment of Koopman’s Line. Chiroptera Neotropical 16:501-521. Genoways, H. H., P. A. Larsen, S. C. Pedersen, and J. J. Huebschman. 2007a. Bats of Saba, Netherlands Antilles. Acta Chiropterologica 9:97-114. Genoways, H. H., R. J. Larsen, S. C. Pedersen, G. G. Kwie¬ cinski, and P. A. Larsen. 2011. Bats of Barbados. Chiroptera Neotropical 17:1029-1054. Genoways, H. H., S. C. Pedersen, P. A. Larsen, G. G. Kwiecinski, and J. J. Huebschman. 2007b. Bats of Saint Martin, French West Indies/Sint Maarten, Netherlands Antilles. Mastozoologia Neotropical 14:169-188. Genoways, H. H., S. C. Pedersen, C. J. Phillips, and L. K. Gordon. 2007c. Bats of Anguilla, northern Lesser Antilles. Occasional Papers, Museum of Texas Tech University 270:1-12. Genoways H. H., C. J. Phillips, and R. J. Baker. 1998. Bats of the Antillean island of Grenada: a new zoogeo¬ graphic perspective. Occasional Papers, Museum of Texas Tech University 177:1-28. Genoways H. H., R. M. Timm, R. J. Baker, C. J. Phillips, and D. A. Schlitter. 2001. Bats of the West Indian island of Dominica: natural history, aerography, and trophic structure. Special Publications, Museum of Texas Tech University 43:1-43. Grant, P. R., and B. R. Grant. 2008. How and why spe¬ cies multiply: the radiation of Darwin’s finches. Princeton University Press, Princeton, New Jersey. 224 pp. Greenbaum, I. F., R. J. Baker, and D. E. Wilson. 1975. Evolutionary implications of the karyotypes of the stenodermine genera Ardops, Ariteus, Phyllops, and Ectophylla. Bulletin of the Southern California Academy of Sciences 74:156-159. Guillot, G., A. Estoup, F. Mortier, and J. F. Cosson. 2005. A spatial statistical model for landscape genetics. Genetics 170:1261-1280. Hall, E. R. 1946. Mammals of Nevada. University of Cali¬ fornia Press, Berkeley. 736 pp. Hall, E. R., and K. R. Kelson. 1959. The mammals of North America, Volume 1. Ronald Press Co, New York. 1083 pp. Heaney, L. R. 2007. Is a new paradigm emerging for oceanic island biogeography? Journal of Biogeography 34:753-757. Hedges, S. B. 1996. Historical biogeography of West Indian vertebrates. Annual Review of Ecology and Sys- tematics 27:163-196. Hill, R. T. 1905. Pele and the evolution of the windward archipelago. Bulletin of the Geological Society of America 16:243-288. Hoffmann, F. G., and R. J. Baker. 2001. Systematics of bats of the genus Glossophaga (Chiroptera: Phyl- lostomidae) and phylogeography of G. soricina based on the cyt-b gene. Journal of Mammalogy 82:1092-1101. Hunt, J. S., E. Bermingham, and R. E. Ricklefs. 2001. Mo¬ lecular systematics and biogeography of Antillean thrashers, tremblers, and mockingbirds (Aves: Mimidae). Auk 118:35-55. Johnsen, A., S. Andersson, J. Garcia Fernandez, B. Kempen- aers, V. Pavel, S. Questiau, M. Raess, E. Rindal, and J. T. Lifjeld. 2006. Molecular and phenotypic diver¬ gence in the bluethroat (Luscinia svecica ) subspe¬ cies complex. Molecular Ecology 15:4033-4047. Jones, J. K., Jr., and A. Schwartz. 1967. Bredin-Archbold- Smithsonian Biological Survey of Dominica: 6. synopsis of bats of the Antillean genus Ardops. Proceedings of the United States National Museum 124(3634): 1-13. Jones, J. K., Jr., and H. H. Genoways. 1973. Ardops nichollsi. Mammalian Species 24:1-2. Kimura, M. 1980. A simple method for estimating evolution¬ ary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Mo¬ lecular Evolution 16:111-120. Koopman, K. F. 1968. Taxonomic and distributional notes on Lesser Antillean bats. American Museum Novitates 2333:1-13. Koopman, K. F., and E. E. Williams. 1951. Fossil Chiroptera collected by H. E. Anthony in Jamaica, 1919-1920. American Museum Novitates 1519:1-29. Kunz, T. H., and E. L. P. Anthony. 1982. Age estimation and post-natal growth in the bat Myotis lucifugus. Journal of Mammalogy 63:23-32. Larkin, M. A., et al. 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23:2947-2948. Larsen et al.—Evolutionary Patterns of Ardops nichollsi 23 Larsen, P. A., H. H. Genoways, and S. C. Pedersen. 2007a. New records of bats from Saint Barthelemy, French West Indies. Mammalia 70:321-325. Larsen, P. A., S. R. Hoofer, M. C. Bozeman, S. C. Pedersen, H. H. Genoways, C. J. Phillips, D. E. Pumo, and R. J. Baker. 2007b. Phylogenetics and phylogeography of the Artibeusjamaicensis complex based on cyt -b DNA sequences. Journal of Mammalogy 88:712— 727. Larsen, P. A., M. R. Marchan-Rivadeneira, and R. J. Baker. 2010. Natural hybridization generates mammalian lineage with species characteristics. Proceedings of the National Academy of Sciences 107:11447- 11452. Larsen, P. A., M. R. Marchan-Rivadeneira, and R. J. Baker. 2013. Speciation dynamics of the fruit-eating bats (genus Artibeus ): with evidence of ecological divergence in Central American populations. Pp. 315-339 in Bat evolution, ecology, and conser¬ vation (R. A. Adams and S. C. Pedersen, eds.). Springer, New York. 547 pp. Larsen, P. A., L. Siles, S. C. Pedersen, and G. G. Kwiecinski. 2011. A new species of Micronycteris (Chiroptera: Phyllostomidae) from Saint Vincent, Lesser Antil¬ les. Mammalian Biology 76:687-700. Larsen, R. J., M. C. Knapp, H. H. Genoways, F. A. A. Khan, P. A. Larsen, D. E. Wilson, and R. J. Baker. 2012. Genetic diversity of Neotropical Myotis (Chirop¬ tera: Vespertilionidae) with emphasis on South American species. PLoS ONE 7:e46578. Lidicker, W. Z. 1962. The nature of subspecies boundaries in a desert rodent and its implications for subspecies taxonomy. Systematic Zoology 11:160-171. Lindsay, K. C., G. G. Kwiecinski, S. C. Pedersen, J. Bade, and H. H. Genoways. 2010. First record of Ardops nichollsi from Antigua, Lesser Antilles. Mammalia 74:93-95. Longmire, J. L., M. Maltbie, and R. J. Baker. 1997. Use of "‘lysis buffer” in DNA isolation and its implica¬ tion for museum collections. Occasional Papers, Museum of Texas Tech University 163:1-3. Losos, J. B., and R. E. Ricklefs. 2010. The theory of island biogeography revisited. Princeton University Press, 496 pp. MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey. 224 pp. Maddison, D. R., and W. R. Maddison. 2005. MacClade 4: analysis of phylogeny and character evolution. Sinauer Associates, Sunderland, Massachusetts. 492 pp. Mancina, C. A., and L. Garcia-Rivera. 2005. New genus and species of fossil bat (Chiroptera: Phyllostomidae) from Cuba. Caribbean Journal of Science 41:22-27. Marchan-Rivadeneira, M. R., P. A. Larsen, C. J. Phillips, R. E. Strauss, and R. J. Baker. 2012. On the asso¬ ciation between environmental gradients and skull size variation in the great fruit-eating bat, Artibeus lituratus (Chiroptera: Phyllostomidae). Biological Journal of the Linnean Society 105:623-634. Masson, D., M. Breuil, and A. Breuil. 1990. Premier inven- taire des chauves-souris de File de Marie-Galante (Antilles francaises). Mammalia 54:656-658. Mayr, E. 1954. Geographic speciation in tropical echinoids. Evolution 8:1-18. McCarthy, T. J., and R. W. Henderson. 1992. Confirmation of Ardops nichollsi on Marie-Galante, Lesser Antilles, and comments on other bats. Caribbean Journal of Science 28:106-107. McDonough, M. M., L. K. Ammerman, R. M. Timm, H. H. Genoways, P. A. Larsen, and R. J. Baker. 2008. Speciation within bonneted bats (genus: Eumops ): the complexity of morphological, mitochondrial, and nuclear data sets in systematics. Journal of Mammalogy 89:1306-1315. McGuire, J. A., C. W. Linkem, M. S. Koo, D. W. Hutchinson, A. K. Lappin, D. I. Orange, J. Lemos-Espinal, B. R. Riddle, and J. R. Jaeger. 2007. Mitochondrial introgression and incomplete lineage sorting through space and time: phylogenetics of crota- phytid lizards. Evolution 61:2879-2897. Mennone, A., C. J. Phillips, and D. E. Pumo. 1986. Evo¬ lutionary significance of interspecific differences in gastrin-like immunoreactivity in the pylorus of phyllostomid bats. Journal of Mammalogy 67:373-384. Miller, G. S., Jr. 1902. Twenty new American bats. Pro¬ ceedings of the Academy of Natural Sciences of Philadelphia 54:389-412. Miller, G. S., Jr. 1906. Twelve new genera of bats. Proceed¬ ings of the Biological Society of Washington 19:83-86. Miller, G. S., Jr. 1913. Five new mammals from tropical America. Proceedings of the Biological Society of Washington 26:31-34. Morgan, G. S. 2001. Patterns of extinction in West Indian bats. Pp. 369-406 in Biogeography of the West Indies: patterns and perspectives, 2nd edition (C. 24 Occasional Papers, Museum of Texas Tech University A. Woods andF. E. Sergile, eds.). CRC Press, Boca Raton, Florida. 608 pp. Patten, M. A., and C. L. Pruett. 2009. The song sparrow, Melospiza melodia , as a ring species: patterns of geographic variation, a revision of subspecies, and implication for speciation. Systematics and Biodiversity 7:33-62. Peakall, R., and P. E. Smouse. 2012. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research - an update. Bioinformatics 28:2537-2539. Pedersen, S. C., H. H. Genoways, M. N. Morton, J. W. John¬ son, and S. E. Courts. 2003. Bats ofNevis, northern Lesser Antilles. Acta Chiropterologia 5:251-267. Pedersen, S. C., H. H. Genoways, M. N. Morton, G. G. Kwiecinski, and S. E. Courts. 2005. Bats of St. Kitts (St. Christopher), northern Lesser Antilles, with comments regarding capture rates of Neotropical bats. Caribbean Journal of Science 41:744-760. Pedersen, S. C., G. G. Kwiecinski, P. A. Larsen, M. N. Morton, R. A. Adams, H. H. Genoways, and V. J. Swier. 2010. Bats of Montserrat: population fluc¬ tuation and response to hurricanes and volcanoes, 1978-2005. Pp. 302-340 in Island bats: ecology, evolution and conservation (T. H. Fleming and P. Racey, eds.). University of Chicago Press, Chicago, Illinois. 560 pp. Pedersen, S. C., P. A. Larsen, H. H. Genoways, M. N. Morton, K. C. Lindsay, and J. Cindric. 2007. Bats of Bar¬ buda, northern Lesser Antilles. Occasional Papers, Museum of Texas Tech University 271:1-19. Pimentel, R. A. 1958. Taxonomic methods, their bearing on subspeciation. Systematic Biology 7:139-156. Pinto, G., D. L. Mahler, L. J. Harmon, and J. B. Losos. 2008. Testing the island effect in adaptive radiation: rates and patterns of morphological diversification in Caribbean and mainland Anolis lizards. Proceed¬ ings of the Royal Society B: Biological Sciences 275:2749-2757. Pregill, G. K., D. W. Steadman, and D. R. Waters. 1994. Late Quaternary vertebrate faunas of the Lesser Antilles: historical components of Caribbean biogeography. Bulletin of the Carnegie Museum of Natural His¬ tory 30:1-51. Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. In¬ ference of population structure using multilocus genotype data. Genetics 155:945-959. Rambaut, A., and A. J. Drummond. 2007. Tracer vl.4. Rojas, D., A. Vale, V. Ferrero, and L. Navarro. 2011. When did plants become important to leaf-nosed bats? Diversification of feeding habits in the family Phyl- lostomidae. Molecular Ecology 20:2217-2228. Ronquist, F., M. Teslenko, P. Van Der Mark, D. L. Ayres, A. Darling, S. Hohna, B. Larget, L. Liu, M. A. Suchard, and J. P. Huelsenbeck. 2012. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. System¬ atic Biology 61:539-542. Roughgarden, J., and J. Roughgarden. 1995. Anolis lizards of the Caribbean: ecology, evolution, a n d p 1 a t e tectonics. New York: Oxford University Press, 226 pp. Scott, J. A. 1972. Biogeography of Antilles butterflies. Bio- tropica 4:32-45. Seutin, G., N. K. Klein, R. E. Ricklefs, and E. Bermingham. 1994. Historical biogeography of the banana- quit (Coereba flaveola) in the Caribbean region: a mitochondrial DNA assessment. Evolution 48:1041-1061. Sikes, R. S., W. L. Gannon, and The Animal Care and Use Committee of the American Society of Mammalo- gists. 2011. Guidelines of the American Society of Mammalogists for use of wild mammals in research. Journal of Mammalogy 92:235-253. Smith, M. F., and J. L. Patton. 1991. Variation in mito¬ chondrial cyt -b sequence in natural populations of South America akodontine rodents (Muridae: Sigmodontinae). Molecular Biology and Evolution 8:85-103. Steadman, D. W., G. K. Pregill, and S. L. Olson. 1984. Fossil vertebrates from Antigua, Lesser Antilles: evidence for late Holocene human-caused extinctions in the West Indies. Proceedings of the National Academy of Science 81:4448-4451. Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and S. Kumar. 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolu¬ tion 10:2731-2739. Team, R. Core. 2014. R: a language and environment for statistical computing. R Foundation for Statistical Computing; Vienna, Austria. 2013. Thomas, O. 1891. Description of three new bats in the Brit¬ ish Museum collection. Annals and Magazine of Natural History, series 6, 7:527-530. Thomas, O. 1894. Description of a new bat of the genus Stenoderma from Montserrat. Proceedings of the Zoological Society of London, pp. 132-133. Larsen et al.—Evolutionary Patterns of Ardops nichollsi 25 Velazco, P. M., andB. D. Patterson. 2008. Phylogenetics and biogeography of the broad-nosed bats, genus Plat- yrrhinus (Chiroptera: Phyllostomidae). Molecular Phylogenetics and Evolution 49:749-759. Vos, P, R. Hogers, M. Bleeker, M. Reijans, T. van de Lee, M. Homes, A. Frijters, J. Pot, J. Peleman, M. Kui- per, andM. Zabeau. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23:4407-4414. Addresses of authors: Roxanne J. Larsen Department of Evolutionary Anthropology Duke University Durham, NC 27708-9976, USA roxy. larsen@duke. edu Peter A. Larsen Department of Biology Duke University Durham, NC 27708-9976, USA peter. larsen@duke. edu Caleb D. Phillips Department of Biological Sciences and Natural Science Research Laboratory Museum of Texas Tech University Lubbock, TX 79409-3131 USA caleb.phillips@ttu. edu Hugh H. Genoways Emeritus University of Nebraska State Museum Lincoln, NE 68588-0514 USA h. h.genoways@gmail. com Wilder, J. A., and H. Hollocher. 2003. Recent radiation of en¬ demic Caribbean Drosophila of the dunni subgroup inferred from multilocus DNA sequence variation. Evolution 57:2566-2579. Zwickl, D. J. 2006. GARLI: genetic algorithm for rapid likelihood inference. See http://www.bioutexas. edu/faculty/antisense/garli/Garli. html. Gary G. Kwiecinski Biology Department The University of Scranton Scranton, PA 18510-4625, USA gary. kwiecinski@scranton. edu Scott C. Pedersen Department of Biology and Microbiology South Dakota State University Brookings, SD 57007-0011, USA scott.pedersen@sdstate. edu Carleton J. Phillips Department of Biological Sciences Texas Tech University Lubbock, TX 79409-3131 USA carl.phillips@ttu. edu Robert J. Baker Emeritus Department of Biological Sciences and Natural Science Research Laboratory Museum of Texas Tech University Lubbock, TX 79409-3131 USA robert. baker@ttu. edu 26 Occasional Papers, Museum of Texas Tech University Appendix Specimens used in all molecular and morphological (indicated by QlS) analyses. Subspecific names follow our taxo¬ nomic review. A.f =Ariteusflavescens ; A. n. =Ardops nichollsi. An asterisk (*) indicates sequences from Carstens et al. (2004). Species Subspecies Island Museum Catalog # Tissue # Cyt-b GenBank ZFY GenBank AFLP Sex A.f flavescens Jamaica TTU 45290 27695 KJ024702 - A ? A.f flavescens Jamaica TTU 45291 27696 KJ024703 KJ024752 A s Af flavescens Jamaica TTU 45293 27701 KJ024704 KJ024753 A 6 A. n. montserratensis St. Martin TTU 101846 129039 KJ024719 - A % A. n. montserratensis St. Martin TTU 101868 129062 KJ024720 - A A. n. montserratensis Saba TTU 101954 117541 KJ024712 - A ¥ A. n. montserratensis St. Eustatius - - HapA *AY572329 - - A. n. montserratensis St. Eustatius AMNH 3925 - - - -