Morphometric, karyotypic, and molecular evidence for a new species of Peromyscus (Cricetidae: Neotominae) from Nayarit, Mexico

Transcription

1 Journal of Mammalogy, 95(1): , 2014 Morphometric, karyotypic, and molecular evidence for a new species of Peromyscus (Cricetidae: Neotominae) from Nayarit, Mexico ROBERT D. BRADLEY,* NICTÉ ORDÓÑEZ-GARZA, CIBELE G. SOTERO-CAIO, HOWARD M. HUYNH, C. WILLIAM KILPATRICK, L. IGNACIO IÑIGUEZ-DÁVALOS, AND DAVID J. SCHMIDLY Department of Biological Sciences, Flint and Main, Texas Tech University, Lubbock, TX , USA (RDB, NO-G, CGS-C, HMH) Museum of Texas Tech University, th Street, Lubbock, TX , USA (RDB) Department of Biology, Marsh Life Science Building, University of Vermont, Burlington, VT , USA (CWK) Departamento de Ecología y Recursos Naturales, Centro Universitario de la Costa Sur, Universidad de Guadalajara, Autlán, Jalisco 48900, México (LII-D) Department of Biology, University of New Mexico, 167 Castetter Hall, Albuquerque, NM 87131, USA (DJS) * Correspondent: robert.bradley@ttu.edu Historically, specimens representing the Peromyscus boylii species group (from montane regions of western and southwestern Mexico) have been referred to as P. boylii or P. levipes. However, previous studies indicated that specimens from eastern Nayarit possessed a karyotype and mitochondrial DNA haplotype distinct from other members of the P. boylii species group. Together, these data precluded an assignment of these specimens to any currently recognized taxon in the P. boylii species group. Availability of additional specimens from this region provided an opportunity to reevaluate the taxonomic status of this complex. Analyses of 18 morphological characters demonstrated that specimens from eastern Nayarit possessed a significantly longer length of the nasals relative to the other species. Further, phylogenetic analyses (parsimony and likelihood) of DNA sequences obtained from the mitochondrial cytochrome-b gene indicate these specimens form a monophyletic clade embedded within a strongly supported clade containing representatives of P. beatae, P. levipes, and P. schmidlyi. Together, these results indicated that specimens from the Sierra Madre Occidental region of Nayarit represent an undescribed species of Peromyscus. Key words: cytochrome-b gene, karyotype, morphology, Peromyscus, Peromyscus boylii species group Ó 2014 American Society of Mammalogists DOI: /13-MAMM-A-217 Understanding the systematic relationships and taxonomy of members of the Peromyscus boylii species group, especially in western and southwestern Mexico, has been challenging. This region has been impacted greatly by a diverse geological history (Ferrusquía-Villfranca 1993) and includes at least 5 morphotectonic provinces (Northwestern Plains and Sierras, Sierra Madre Occidental, Central Plateau, Trans-Mexican Volcanic Belt, and Sierra Madre del Sur) that interdigitate throughout this area. This diverse terrain and associated habitats have given mammalian populations the opportunity to become isolated, evolve, and speciate under several different environmental conditions. Consequently, it has been difficult to dissect morphological variation associated with speciation events from those linked to geographic, elevational, clinal, and other sources of variation. This especially is true given the general pattern of morphological conservation among members of the P. boylii species group, evidence for cryptic species (Schmidly et al. 1988; Bradley et al. 2004) within the group, and lack of sufficient research material from critical geographic areas. Historically, as many as 7 subspecies of P. boylii (beatae, boylii, glasselli, levipes, madrensis, simulus, andspicilegus) have been recognized from western and southwestern Mexico (Baird 1855; Allen 1893, 1897; Merriam 1898a, 1898b; Thomas 1903; Osgood 1904, 1909; Burt 1932; Hooper 1955; Hall 1981). The complexity of P. boylii was revealed somewhat with the elevation of P. b. madrensis, P. b. simulus, and P. b. spicilegus to species status (Carleton 1977); elevation of P. b. beatae and P. b. levipes to species rank and refinement of their distributions to include only eastern and south-central 176

2 February 2014 BRADLEY ET AL. A NEW SPECIES OF PEROMYSCUS 177 Mexico, respectively (Houseal et al. 1987; Rennert and Kilpatrick 1987; Schmidly et al. 1988); retention of P. b. boylii and P. b. glasselli as subspecies and restriction of their distributions to the Mexican Plateau and southwestern United States and San Pedro Nolasco Island, Sonora, respectively (Houseal et al. 1987; Rennert and Kilpatrick 1987; Schmidly et al. 1988); and the recognition of a cryptic species, P. schmidlyi (Bradley et al. 2004), from western Mexico (Cabrera et al. 2007; Ordóñez-Garza and Bradley 2010; López-González et al. 2013). Despite these taxonomic advances, several lines of evidence (morphology, allozymes, chromosomes, and DNA sequences) suggested the existence of undescribed taxa remaining in western Mexico, especially regarding populations of P. boylii like forms from the states of Jalisco, Michoacán, and Nayarit (Lee et al. 1972; Schmidly and Schroeter 1974; Carleton et al. 1982; Houseal et al. 1987; Tiemann-Boege et al. 2000; Bradley et al. 2004). Specifically, 2 of the unique karyotypic groups (IV and V) proposed by Houseal et al. (1987), based on number of biarmed chromosomes in the karyotype (fundamental number [FN]), remain unassignable to any currently recognized taxon. One chromosomal group (IV) encompassed specimens from Jalisco, Michoacán, and Nayarit that possessed a karyotype with a polymorphic, high FN (65 66, 68) and the 2nd group (V) included specimens from Michoacán that possessed a monomorphic, low FN (54). In this study, efforts were focused on resolving the taxonomy of members of karyotypic group IV (FN ¼ 65 66, 68) that reside in high-elevation (. 1,900 m) pine oak forests (Pinus sp. and Quercus sp.) of Jalisco, Michoacán, and Nayarit and on determining their phylogenetic relationship to other members of the P. boylii species group. Even though Houseal et al. (1987) treated these populations as a single karyotypic group (IV), examination of DNA sequence data (Tiemann- Boege et al. 2000; Bradley et al. 2004) suggested that 2 of the populations (Ocota, Nayarit and Dos Aguas, Michoacán) were genetically divergent. Although not a focal point of their research, Tiemann-Boege et al. (2000) and Bradley et al. (2004) indicated that these populations were assignable to the P. boylii species group; however, they were not conspecific and were not sister taxa. Therefore, we focus on the taxonomic status and relationships of the following populations: Ocota Airstrip, Nayarit (FN ¼ 66 Carleton et al. 1982), Dos Aguas, Michoacán (FN ¼ 65, 66, and 68 Houseal et al. 1987), and new populations reported herein. MATERIALS AND METHODS Samples. Thirty-two individuals representing a new species of Peromyscus were collected from a naturally occurring population in Nayarit, Mexico (70 km N Santa María del Oro; 2,014 m; UTM 13Q E N; Fig. 1), on 17 July These individuals were examined in the morphological analysis (n ¼ 32) and DNA sequencing (n ¼ 12) portions of this study. In addition, DNA sequences from 26 individuals, representing 14 species and 3 presumably undescribed taxa, reported in Bradley et al. (2000, 2004, 2007), Tiemann-Boege et al. (2000), and this study, were included as internal references and outgroup comparisons. Specimens were collected following methods outlined in guidelines of the American Society of Mammalogists (Sikes et al. 2011) and approved by the Texas Tech University Animal Care and Use Committee. Specimen numbers and collection localities are listed in Appendix I. Morphologic characters. Eighteen cranial measurements (defined in Carleton et al. 1982) were recorded in millimeters (mm) from adult individuals or were obtained from previous studies (Carleton et al. 1982; Schmidly et al. 1988; Schmidly and Bradley 1995). Adult individuals (age classes IV VI) were judged as such based on patterns of molar tooth wear (Schmidly 1973). Measurements are as follows: greatest length of skull, length of auditory bulla, postpalatal length, length of mesopterygoid fossa, palatal length, length of incisive foramen, length of molar toothrow, greatest zygomatic breadth, mastoidal breadth, greatest breadth across molars, postdental palatal breadth, greatest width of mesopterygoid fossa, depth of braincase, breadth of braincase, least interorbital width, rostral breadth, nasal length, and rostral length. Statistical analyses of morphologic data. For the 18 cranial characters listed above, specimens from Santa María del Oro, Nayarit (n ¼ 32), were compared to populations delimited by previous allozyme, karyotypic, and morphologic analyses (Houseal et al. 1987; Rennert and Kilpatrick 1987; Schmidly et al. 1988) to represent samples of P. beatae (n ¼ 39), P. boylii (n ¼ 12), P. levipes (n ¼ 42), P. schmidlyi (n ¼ 10), and P. simulus (n ¼ 16). For all analyses, sexes were combined following Schmidly et al. (1988). For descriptive and comparative purposes, means, ranges, and standard errors were calculated for each character and species; for all further analyses, characters were log-transformed (natural log). Specimens with missing measurements were excluded from analyses. Statistical significance for comparison of variation among the 18 individual characters was estimated using analysis of variance (ANOVA). Levels of statistical significance for variation among species were estimated using multivariate analysis of variance (MANOVA). Canonical variate analysis, of the 6 species discussed above, was used to produce a scatter plot of specimens along the first 2 canonical axes producing maximal and 2nd-to-maximal separation among all groups derived from multigroup discriminant function. All variable loadings are expressed as a product of correlation coefficients of the extracted components of canonical variates with the log-transformed cranial measurements. All statistical tests were evaluated at a ¼ 0.05 and were performed using PAST (Hammer et al. 2001). Karyotypic data. Karyotypes representing the Ocota, Nayarit, population were obtained from Carleton et al. (1982). For comparison, 5 individuals of Peromyscus from Santa María del Oro, Nayarit, were karyotyped following the bone marrow method of Baker and Qumsiyeh (1988). At least 9 metaphase spreads were examined and photographed for each individual. Karyograms were constructed based on

3 178 JOURNAL OF MAMMALOGY Vol. 95, No. 1 FIG. 1. Distribution of selected populations and species of the Peromyscus boylii species group from western Mexico and surrounding states. Emphasis was placed on depicting the newly described species and its closest phylogenetic allies or taxa with similar karyotypes. Closed circles represent collecting localities and numbers refer to samples listed in Appendix I. chromosomal morphology presented in Committee for Standardization of Chromosomes of Peromyscus (1977) and Greenbaum et al. (1994) and were compared to karyotypes and FNs previously discussed by Carleton et al. (1982), Houseal et al. (1987), and Smith (1990). Sequence data. Mitochondrial DNA was isolated from approximately 0.1 g of frozen liver tissue using the DNeasy kit (Qiagen, Valencia, California) or the Wizard Miniprep kit (Promega, Madison, Wisconsin). The entire cytochrome-b gene (Cytb, 1,143 base pairs [bp]) was amplified using the polymerase chain reaction method (Saiki et al. 1988) and the following primers: MVZ05 (Smith and Patton 1993) and PERO3 0 (Tiemann-Boege et al. 2000). Thermal profiles for polymerase chain reaction were as follows: initial denaturation at 958C for 2 min, followed by 35 cycles of denaturation at 958C for 1 min, annealing at 518C for 1 min, and extension at 728C for 2 min, with a final extension at 728C for 7 min. Most polymerase chain reaction products were purified with either an ExoSap or a QIAGEN Gel Extraction Kit (Qiagen). Primers used to cycle sequence the products included: WDRAT1100, 400R, 700H, and NEO700L (Peppers and Bradley 2000) and 400F (Tiemann-Boege et al. 2000). Cycle sequencing reactions were purified using isopropanol cleanup protocols. Purified products were sequenced with an ABI 3100-Avant automated sequencer and ABI Prism Big Dye version 3.1 terminator technology (Applied Biosystems, Foster City, California). Resulting sequences were aligned and proofed using Sequencher 4.0 software (Gene Codes, Ann Arbor, Michigan); chromatograms were examined to verify all base changes. All Cytb sequences obtained in this study were deposited in GenBank and are listed in Appendix I. Based on the phylogenetic relationships of the genus Peromyscus presented in Bradley et al. (2007), P. gratus was utilized as the outgroup taxon in all sequence analyses. Representatives of all members of the P. boylii species group (beatae, boylii, levipes, madrensis, schmidlyi, simulus, and stephani) were included in analyses as internal standards. In addition, members of the P. aztecus species group (aztecus, evides, hylocetes, oaxacensis, spicilegus, and winklemanni) and 3 undescribed taxa with apparent affiliations to the P. boylli species group were included based on their chromosomal similarities or geographic proximity. A parsimony analysis (PAUP* Swofford 2002) was constructed using equally weighted characters. The variable nucleotide positions within the data set were treated as unordered, discrete characters with 4 possible states; A, C, G, or T. All phylogenetically uninformative characters were excluded from these analyses. The heuristic search and treebisection-reconnection options (Swofford 2002) were used to find the most-parsimonious trees. A consensus (strict) was generated from the available trees and the bootstrap analysis (Felsenstein 1985) with 1,000 iterations used to evaluate nodal support. Fifty-six maximum-likelihood models were examined using Modeltest (Posada and Crandall 1998) to determine the model of DNA evolution best fitting the data. The Akaike information criterion identified GTRþIþG as being the most appropriate model for this data set. This model generated significantly better likelihood scores ( lnl ¼ ) than all other

4 February 2014 BRADLEY ET AL. A NEW SPECIES OF PEROMYSCUS 179 TABLE 1. Summary statistics of 18 cranial characters obtained from 6 members of the Peromyscus boylii species group. Abbreviations are: n ¼ sample size and SE ¼ 1 standard error. P. boylii (n ¼ 12) P. beatae (n ¼ 39) P. levipes (n ¼ 42) P. schmidlyi (n ¼ 10) P. simulus (n ¼ 160) P. species novum (n ¼ 32) X Range SE X Range SE X Range SE X Range SE X Range SE X Range SE Greatest length of skull Rostral length Nasal length Postpalatal length Greatest zygomatic breadth Breadth of braincase Mastoidal breadth Least interorbital width Length of molar toothrow Length of incisive foramen Length of auditory bulla Depth of braincase Length of mesopterygoid fossa Palatal length Rostral breadth Greatest breadth across molars Postdental palatal breadth Greatest width of mesopterygoid fossa models and included the following parameters: base frequencies (A ¼ , C ¼ , G ¼ , T ¼ ), rates of substitution (A C ¼ , A G ¼ , A T ¼ , C G ¼ , C T ¼ , G T ¼ 1.00), proportion of invariable sites (I ¼ ), and gamma distribution (G ¼ ). A Bayesian model (MrBayes Huelsenbeck and Ronquist 2001) was used for comparison to the likelihood method and to generate support values (clade probabilities). A GTRþIþG model with a site-specific gamma distribution was used with the following options: 4 Markov chains, 10 million generations, and sample frequency ¼ every 1,000th generation. After a visual inspection of the likelihood scores, the first 1,000 trees were discarded and the model was rerun using the remaining stable likelihood values. A consensus tree (50% majority rule) was constructed from the remaining trees. The Kimura 2-parameter model of evolution (Kimura 1980) was used to calculate genetic distances. These values were then used to assess levels of genetic divergence among species of Peromyscus following the criteria outlined in Bradley and Baker (2001) and Baker and Bradley (2006). RESULTS Morphometric data. Standard sample statistics were derived for all characters and taxa and are provided in Table 1. The ANOVA results indicated that all characters except greatest zygomatic breadth (P ¼ 0.07) were significantly different among the 6 species examined. Results of the MANOVA indicated that specimens from Santa María del Oro, Nayarit, were significantly different (Wilkss k 90,620.6 ¼ ; F ¼ 21.47; P ¼ 4.173E 140) from specimens representing the other taxa (P. beatae, P. boylii, P. levipes, P. schmidlyi, and P. simulus). Because the MANOVA showed significant overall difference between groups, the analysis was preceded by the post hoc analysis by pairwise Hotelling s test comparisons. Hotelling s P indicated that P. schmidlyi and P. simulus were not significantly different relative to P. boylii (P ¼ 0.13 and P ¼ 0.08, respectively), whereas the new species of Peromyscus from Santa María del Oro was significantly different (P, 0.05 for all pair-wise comparisons) from all species compared from western Mexico (P. beatae, P. boylii, P. levipes, P. schmidlyi, and P. simulus). Comparable and unambiguous phenetic structure among the samples of Peromyscus from western Mexico was depicted in bivariate plots of the first 2 canonical variates. Projection of individual scores onto canonical variates 1 and 2 (Fig. 2) revealed nonoverlapping densities corresponding to specimens from Santa María de Oro, Nayarit, and those from other taxa in western Mexico (P. beatae, P. boylii, P. levipes, P. schmidlyi, and P. simulus), suggesting that the major axes of these distributions are phenetically discrete. Karyotypic data. Karyotypes were obtained from 5 individuals collected from the locality near Santa María del Oro, Nayarit, as well as for closely related species and populations from nearby geographic localities (Table 2). The

5 180 JOURNAL OF MAMMALOGY Vol. 95, No. 1 FIG. 2. Projection of the first 2 canonical variates obtained from analysis of the 18 cranial characters examined in this study. Symbols are as follows: crosses (Peromyscus beatae), open circles (Peromyscus boylii), closed squares (Peromyscus levipes), closed diamonds (Peromyscus schmidlyi), closed triangles (Peromyscus simulus), and open squares (Peromyscus species novum). karyotypes of the specimens from Santa María del Oro, Nayarit, display a similar number and morphology of biarmed chromosomes (10 pairs) to that reported for specimens collected at a nearby locality (Ocota, Nayarit Carleton et al. 1982). However, the diploid numbers differ considerably because the 5 karyotypes obtained from the specimens collected near Santa María del Oro, Nayarit, possess diploid numbers ranging from 2n ¼ 42 to 2n ¼ 44. At least 30 other individuals, representing 7 other species of Peromyscus were karyotyped on the same collecting trip; several individuals were karyotyped simultaneously as the specimens from Santa María del Oro. All individuals from localities other than Santa María del Oro possess a 2n ¼ 48, making it unlikely that karyotyping methods produced erroneous results. In addition, several species of bats and rodents were karyotyped during this collecting trip and all karyotypes correspond to published results, further discounting the possibility of erroneous results. To eliminate the possibility of errors associated with labeling specimen tubes or transcription of data, we obtained Cytb gene sequences from 5 of the chromosome preparations. Specifically, a 150-ll aliquot of the chromosome suspension was washed (2 times) with ethanol (200 ll) and 1 3 phosphate-buffered saline (200 ml) to remove the chromosome fixative (Carnoy s methanol and acetic acid solution). DNA was then isolated using the DNeasy kit (Qiagen) and a 500-bp region of the Cytb gene was amplified using combinations of the primers MVZ04 and MVZ05 (Smith and Patton 1993), and LGL765 (Bickham et al. 1995). These sequences were obtained using separate primers and reagents and all methods were completed in a different laboratory to rule out contamination. Resultant DNA sequences were identical to those obtained from the tissue samples, confirming the authenticity of the results. This interesting result, relative to the occurrence of an abnormal diploid number for Peromyscus, cannot be explained at this time. Currently, attempts are underway to determine if the TABLE 2. Comparison of karyotypes for members of the Peromyscus boylii species group examined in this study. All chromosomal assessments are based on nondifferentially stained karyotypes as interpreted from comparisons to data presented in or cited by Houseal et al. (1987) and Smith (1990). Only chromosomes that have been identified as biarmed for the P. boylii species group are included. All karyotypes possessed a biarmed condition for chromosomes 1, 22, and 23 (except in some populations of P. beatae [see Davis et al. 1986]). Abbreviations are: a ¼ acrocentric, b ¼ biarmed, and p ¼ polymorphic. References: 1 ¼ Lawlor (1971), 2 ¼ Lee et al. (1972), 3 ¼ Schmidly and Schroeter (1974), 4 ¼ Carleton et al. (1982), 5 ¼ Houseal et al. (1987), 6 ¼ Smith (1990), 7 ¼ Bradley et al. (2004), and 8 ¼ this study. We refrain from reporting the fundamental number (FN) for the samples from Santa María del Oro until the discrepancy in FN can be resolved. Chromosome Taxon FN Reference P. beatae p a a a a a a a a a 5, 6 P. boylii 52 a a a a a a a a a a 5 P. levipes b p a a p a a b a a 5, 6 P. madrensis 54 a a a a a a a b a a 4 P. schmidlyi p a a a a a a p a a 2, 3, 7 P. simulus 52 a a a a a a a a a a 4 P. stephani 52 a a a a a a a a a a 1 P. sp. nov. (Ocota, Nayarit; n ¼ 3) 66 b b a b b b a b b a 4 P. sp. nov. (Santa María del Oro, Nayarit; n ¼ 5) * b b a b b b a b b a 8 P. sp. 1 (Dos Aguas, Michoacán; n ¼ 5) 65, 66, 68 b b a b b p a b p p 5 P. sp. 2 (Zitácuaro, Michoacán; n ¼ 1) 56 b a a a b a a b a a 8 P. sp. 3 (Zinapécuaro, Michoacán; n ¼ 4) 68 b b a b b b a b b b 8 P. sp. (Patzcuaro, Michoacán; n ¼ 1) 56 b a a a a a a b a a 5 P. sp. (Los Reyes, Michoacán; n ¼ 1) 66 b b a b b b a b b a 3 P. sp. (Volcan de Colima, Jalisco; n ¼ 4) 68 b b a b b b a b b b 5 P. sp. (Los Reyes, Michoacán; n ¼ 1) 54 b a a a a a a a a a 2, 3 P. sp. (Jiquilpan Reyes, Michoacán; n ¼ 4) 54 b a a a a a a a a a 2, 3 P. sp. (Los Azufres, Michoacán; n ¼ 3) 56 b a a a a a a b a a 5

6 February 2014 BRADLEY ET AL. A NEW SPECIES OF PEROMYSCUS 181 FIG. 3. Phylogenetic tree of the Peromyscus boylii species group generated using Bayesian methods (MrBayes Huelsenbeck and Ronquist 2001) and the GTRþIþG model of evolution. Clade probability values (. 0.95) are indicated by an asterisk (*) and are shown above branches; bootstrap support values obtained from the parsimony analysis are below branches. missing chromosomal material was translocated (or shuffled by other mechanisms) onto other chromosomes, or if it was truly lost from the genome. Sequence data. Sequences from 14 individuals representing the new species were combined with 23 samples representing members of the P. boylii species group, other closely related species groups, and the outgroup taxon. In all analyses (parsimony, likelihood, and Bayesian), the P. boylii species group was monophyletic, in agreement with the findings of Tiemann-Boege et al. (2000) and Bradley et al. (2004). Relationships of taxa included as reference samples were peripheral to this study and are not discussed in detail (see Fig. 3). The maximum-likelihood analysis produced a topology (Fig. 3) in which the 14 individuals, representing the undescribed taxon (samples from Santa María del Oro and Ocota, Nayarit), formed a strongly supported clade (clade probability value [CPV] ¼ 1.00). This clade, in turn, was embedded in a larger unresolved clade containing representatives of P. beatae, P. levipes, P. schmidlyi, and P. sp. 1 3 from Michoacán. This large clade was strongly supported (CPV ¼ 1.00); however, the sister group relationship of the undescribed taxon to other members of the P. boylii species group could not be determined. The remaining samples of the P. boylii species group (boylii, madrensis, simulus, and stephani) were added to form a monophyletic P. boylii species group (CPV ¼ 1.00). The parsimony analysis generated 372 equally mostparsimonious trees (length ¼ 650). A majority rule consensus tree was generated (not shown) that was similar in topology to the tree obtained in the Bayesian analysis. Bootstrap support values were obtained and superimposed onto Fig. 3. In this analysis, the 14 samples from Santa María del Oro and Ocota formed a strongly supported monophyletic clade (bootstrap [BS] ¼ 96). This clade was embedded within a moderately supported (BS ¼ 76) but unresolved clade containing members of P. levipes and an undescribed species of Peromyscus (P. sp. 2 from Michoacán). This clade then was joined with samples of P. beatae and P. schmidlyi to form a strongly supported monophyletic clade (BS ¼ 96). This large clade was joined in a stepwise fashion by clades containing the remaining members of the P. boylii species group (boylii, madrensis, simulus, and stephani). The genetic divergence values (Table 3), estimated using the Kimura 2-parameter model of evolution (Kimura 1980), among samples representing the new species averaged 0.65%. Collectively, these samples differed from their closest relatives

7 182 JOURNAL OF MAMMALOGY Vol. 95, No. 1 TABLE 3. Average genetic distances estimated using the Kimura 2- parameter model of evolution (Kimura 1980) for selected comparisons of taxa of Peromyscus. Comparison Average genetic distance Within species P. beatae 2.24% P. boylii 0.56% P. levipes 1.24% P. schmidlyi 0.87% P. species novum 0.65% Between P. species novum and selected species in the P. boylii species group P. species novum P. beatae 5.70% P. species novum P. boylii 7.63% P. species novum P. levipes 3.50% P. species novum P. madrensis 6.55% P. species novum P. schmidlyi 3.40% P. species novum P. simulus 7.37% P. species novum P. stephani 7.77% Between selected species in the P. boylii species group P. beatae P. boylii 8.31% P. beatae P. levipes 5.65% P. beatae P. schmidlyi 5.63% P. boylii P. levipes 8.50% P. boylii P. schmidlyi 7.94% P. boylii P. simulus 2.82% P. levipes P. schmidlyi 3.25% (determined in this study), P. beatae, P. levipes, and P. schmidlyi, by 5.70%, 3.50%, and 3.40%, respectively. Genetic divergence values for other currently recognized species in the P. boylii species group ranged from 2.82% (P. boylii and P. simulus) to 8.50% (P. boylii and P. levipes). The undescribed taxon differs from P. boylii, with which it was considered conspecific, by a genetic divergence value of 7.63%. DISCUSSION Together, the results from the multivariate analyses, DNA sequence data, karyotypic data, and patterns of geographic distribution support the hypothesis that the individuals (previously referred to P. boylii) from Nayarit and possibly surrounding areas in Jalisco and Zacatecas, represent an undescribed species of Peromyscus. Below these Nayarit populations are formally described as a new species. Peromyscus carletoni, species novum Holotype. Museum of Texas Tech University (MoTTU) ; adult male; skin, skull, and skeleton. Original number Robert D. Bradley 2526; TK identifies tissue samples deposited in Natural Sciences Research Laboratory, MoTTU. Type locality. México: Nayarit; 70 km N Santa María del Oro (UTM 13Q ), collected 19 July Paratypes. Skins, skulls, skeletons, and tissues of 15 males (TTU110265, TTU110266, TTU110269, TTU110271, TTU110276, and TTU TTU110132) and 16 females (TTU TTU110264, TTU110267, TTU110268, TTU110270, TTU TTU110275, TTU TTU110280) deposited in the MoTTU. Diagnosis. A species of Peromyscus with the following characters: size medium for the genus; tail slightly longer than head and body; hind foot length medium; ear length medium; dorsal coloration dark (Chaetura Drab at tips, Blackish Slate at base; color terms after Ridgway [1912]); sides Snuff Brown; venter pelage White at tips, Blackish Slate at base; feet with Hair Brown strip extending slightly past ankle; toes White; tail bicolored, Chaetura Drab above and White below, scantily haired at base and tufted at tip; ears Chaetura Black; vibrissae Black; skull elongate, twice as long as wide; braincase rounded; nasal length 101.6% of rostral length and rostral length 37% of greatest skull length; molar toothrow about 16% of skull length; interorbital constriction a smooth outline, not angular; zygomatic arches nearly parallel; auditory bullae median; karyotype with 2n ¼ and 10 pairs of biarmed chromosomes. Distribution. At this time, known only from the type locality and 1 other site (Ocota) in the high-elevation mesic, montane, pine oak forests in the Sierra Madre Occidental of east-central Nayarit, Mexico. It may be that P. carletoni is a relictual species, although it is possible that this species may occur in the pine oak forest of western Zacatecas and northern Jalisco. Additional trapping efforts are needed throughout this region to refine its distribution. Measurements. External measurements of the holotype as taken in the field (in mm) by R. D. Bradley are: total length ¼ 175; tail length ¼ 93; hind foot length ¼ 20; and ear length ¼ 19. Cranial measurements were obtained using dial calipers (in mm) and are: greatest length of skull ¼ 27.1; greatest zygomatic breadth ¼ 13.6; length of auditory bulla ¼ 4.3; postpalatal length ¼ 9.8; length of mesopterygoid fossa ¼ 4.4; palatal length ¼ 4.1; length of incisive foramen ¼ 5.6; length of molar toothrow ¼ 4.4; greatest zygomatic breadth ¼ 13.6; mastoidal breadth ¼ 12.1; greatest breadth across molars ¼ 5.4; postdental palatal breadth ¼ 4.2; greatest width of mesopterygoid fossa ¼ 2.6; depth of braincase ¼ 9.6; breadth of braincase ¼ 13.1; least interorbital width ¼ 4.5; rostral breadth ¼ 4.6; nasal length ¼ 11.1; and rostral length ¼ 9.5. Mean measurements, ranges, and standard errors for a series of 32 specimens are presented in Table 1. Comparisons. A species of the P. boylii species group, resembling P. beatae, P. levipes, and P. schmidlyi in size and coloration but not in distribution (allopatric); karyotype and DNA sequence divergence (Cytb gene) preclude confusion. Smaller and darker compared to P. levipes and P. beatae and slightly larger in most external and cranial measurements than P. schmidlyi, P. boylii rowleyi, and P. simulus. Length of rostrum relative to greatest length of skull (35.2%) shorter than for P. schmidlyi (39.4%), P. simulus (39.5%), P. b. rowleyi (41.2%), P. levipes (41.1%), and P. beatae (41.6%). Length of nasals, relative to length of rostrum, longer in P. carletoni (115.2%) compared to the other taxa (P. beatae, 89.7%; P. levipes, 90.5%; P. simulus, 88.4%; P. b. rowleyi, 86.4%; and P.

8 February 2014 BRADLEY ET AL. A NEW SPECIES OF PEROMYSCUS 183 FIG. 4. Bivariate scatter plot depicting the length of rostrum versus length of nasal of samples (mm) representing Peromyscus boylii, Peromyscus beatae, Peromyscus levipes, Peromyscus schmidlyi, Peromyscus simulus, and Peromyscus sp. from Santa María del Oro, Nayarit, Mexico. schmidlyi, 85.8%). A bivariate scatter plot of the rostral length versus nasal length provides the best morphometric separation, with P. carletoni having a longer rostrum and shorter nasals relative to the other taxa (Fig. 4). There is no external or cranial measurement in which P. carletoni differs significantly (ANOVA; P, 0.05) from the other 5 taxa, although the 5 taxa differed significantly when considering all characters. The most distinctive measurement is rostral length, which is significantly (P. 0.05) shorter in P. carletoni than all other taxa but P. schmidlyi (Table 1). P. carletoni is significantly smaller than P. beatae in all 4 external measurements and 8 of 18 cranial measurements (greatest length of skull, rostral length, greatest zygomatic breadth, nasal length, breadth of braincase, length of auditory bulla, depth of braincase, and greatest breadth across molars). P. carletoni is significantly smaller than P. levipes in all external measurements except length of molar toothrow and in 5 cranial measurements (greatest length of skull, rostral length, length of auditory bulla, depth of braincase, and length of mesopterygoid fossa); it is significantly larger than P. levipes in 1 cranial measurement (greatest width of mesopterygoid fossa). From P. b. rowleyi, P. carletoni is significantly smaller in 10 cranial measurements (nasal length, postpalatal length, breadth of braincase, mastoidal breadth, least interorbital width, length of molar toothrow, length of incisive foramen, length of auditory bulla, rostral breadth, and greatest width of mesopterygoid fossa) and significantly larger in 3 others (rostral length, length of auditory bulla and length of mesopterygoid fossa). From P. schmidlyi, P. carletoni is significantly larger in 4 measurements (length of molar toothrow, length of mesopterygoid fossa, rostral breadth, and postdental palatal breadth) and significantly smaller in 3 others (length of hind foot, rostral length, and nasal length). P. carletoni is significantly larger than P. simulus in 10 measurements (nasal length, breadth of braincase, mastoidal breadth, least interorbital width, length of molar toothrow, length of incisive foramen, length of auditory bulla, length of mesopterygoid fossa, postdental palatal breadth, and greatest width of mesopterygoid fossa) and significantly smaller in 3 others (rostral length, length of auditory bulla, and length of mesopterygoid fossa). The MANOVA with pairwise Hotelling s test revealed that P. carletoni differed significantly from the other 5 taxa, and a 2-dimensional plot in multivariate space showed P. carletoni to be more like P. schmidlyi than the other taxa. Axis 1, which is heavily influenced by rostal length, separates P. carletoni and P. schmidlyi from the other 4 taxa, whereas axis 2 distinguishes P. beatae and P. levipes from P. simulus and P. rowleyi. A forthcoming communication will provide a more thorough univariate and multivariate assessment of morphological variation among all of the taxa used in this study. Peromyscus carletoni differs genetically, based on Cytb sequences, from other members of the P. boylii species group (P. beatae, P. boylii, P. levipes, P. madrensis, P. schmidlyi, P. simulus, and P. stephani) by 5.7%, 7.63%, 3.5%, 6.55%, 3.4%, 7.37%, and 7.77% sequence divergence, respectively. At this time, the sister relationship of P. carletoni cannot be determined because it forms an unresolved polytomy with P. beatae, P. levipes, and P. schmidlyi. P. carletoni is slightly smaller and differs cranially (shape of interorbital constriction a smooth outline and not angular; no supraorbital ridge) from P. spicilegus of the P. aztecus species group with which P. carletoni is sympatric in Santa María del Oro, Nayarit. Peromyscus carletoni differs from other members of the genus Peromyscus and the P. boylii species group in having a

9 184 JOURNAL OF MAMMALOGY Vol. 95, No. 1 unique karyotype (2n ¼ 42 44). The karyotype possesses 10 pairs of biarmed chromosomes that are similar in size and morphology (Table 2) to those reported by Carleton et al. (1982). This similarity coupled with a similar cranial morphology and Cytb haplotypes provides support for a conspecific association of these 2 populations. Together, the karyotypes obtained from Santa María del Oro and Ocota are distinguishable from karyotypes reported all other members of the P. boylii species group, except for 4 populations from Jalisco and Michoacán (Table 2); however, differences in DNA sequences exclude an association of the P. carletoni karyotype with the more southern karyotypes. Although we refrain from presenting the FN of these karyotypes until the difference in diploid number can be resolved, the karyotype is distinguishable from other FN ¼ 52 forms (P. boylii and P. simulus), FN ¼ forms (P. beatae), FN ¼ 54 forms (P. madrensis), FN ¼ forms (P. schmidlyi), and FN ¼ forms (P. levipes) by possessing additional biarmed chromosomes (Table 2). Habitat. Found in mesic pine oak forest (Quercus sp. and Pinus sp.) habitat at elevations greater than 2,000 m. Typically associated with rock outcroppings, fallen logs, and moist soils. Collected sympatrically with P. spicilegus and Neotoma mexicana at Santa María del Oro and with N. mexicana and Reithrodontomys sp. at Ocota. Remarks. There are no known morphological characters, including external and cranial measurements, that will uniquely distinguish all specimens of P. carletoni from those of P. beatae, P. levipes, P. schmidlyi, P. b. rowleyi, orp. simulus, although as noted above (Fig. 4), P. carletoni does have on average a shorter rostrum and longer nasals compared to the other taxa. All of these taxa meet the classic definition of cryptic species in that they are genetically distinct but are very difficult to uniquely identify using morphological criteria. Etymology. This species is named in honor of Dr. Michael D. Carleton (National Museum of Natural History, Smithsonian Institution) for his eloquent contributions to the studies of the genus Peromyscus and devotion to rodent systematics and taxonomy. Specimens examined. See Appendix I. RESUMEN Históricamente, los especímenes que representan al grupo de especies de Peromyscus boylii (de regiones montanas del oeste y suroeste de México) han sido identificadas como P. boylii o P. levipes. Sin embargo, estudios previos han indicado que los especímenes del este de Nayarit poseen un cariotipo y un haplotipo del ADN mitocondrial distinto al de otros miembros del grupo de especies de P. boylii. Juntos, estos datos excluyen la asignación de dichos especímenes a cualquier taxón actualmente reconocido en el grupo de especies de P. boylii. La disponibilidad de muestras adicionales de esta región permitió reevaluar la taxonomía de este complejo. El análisis de 18 caracteres morfológicos demostró que los especímenes procedentes del este de Nayarit poseen una longitud significativamente mayor del hueso nasal en relación a otras especies de este complejo. Además, análisis filogenéticos (parsimonia y verosimilitud) de secuencias de ADN obtenidas del gen mitocondrial citocromo-b indican que estos especímenes forman un clado monofilético embebido dentro de otro clado que, con fuerte apoyo estadístico, contiene a representantes de P. beatae, P. levipes, y P. schmidlyi. En conjunto, todos estos resultados indican que los especímenes de la región de Sierra Madre Occidental de Nayarit representan una especie no descrita de Peromyscus. ACKNOWLEDGMENTS We thank C. W. Thompson, M. S. Corley, and M. R. Mauldin for reviewing earlier versions of this manuscript. Special thanks to C. W. Thompson, E. Howell, and M. J. Hamilton for karyotyping specimens. Tissue samples were provided by the Natural Science Research Laboratory, Museum of Texas Tech University, and Zadock- Thompson Natural History Collection, University of Vermont. Collecting permits were provided by Secretaría de Medio Ambiente y Recursos Naturales (FAUT-003). Special thanks to the Field Methods classes of 2006 and 2008 for assistance in collecting specimens. Partial support for this research was provided by a National Institutes of Health grant (DHHS A to RDB). LITERATURE CITED ALLEN, J. A List of mammals collected by Mr. Charles P. Rowley in the San Juan region of Colorado, New Mexico, and Utah, with descriptions of new species. Bulletin of the American Museum of Natural History 5: ALLEN, J. A Further notes on mammals collected in Mexico by Dr. Audley C. Buller, with descriptions of new species. Bulletin of the American Museum of Natural History 9: BAIRD, S. F Characteristics of some new species of North American Mammalia, collected chiefly in connection with the U.S. surveys of a railroad route to the Pacific. Proceedings of the Academy of Natural Sciences of Philadelphia 7: BAKER, R. J., AND R. D. BRADLEY Speciation in mammals and the Genetic Species Concept. Journal of Mammalogy 87: BAKER, R. J., AND M. B. QUMSIYEH Methods in chiropteran mitotic chromosomal studies. Pp in Ecological and behavioral methods for the study of bats (T. H. Kunz, ed.). Smithsonian Institution Press, Washington, D.C. BICKHAM, J. W., C. C. WOOD, AND J. C. PATTON Biogeographic implications of cytochrome b sequences and allozymes in sockeye (Oncorhynchus nerka). Journal of Heredity 86: BRADLEY, R. D., AND R. J. BAKER A test of the Genetic Species Concept: cytochrome-b sequences and mammals. Journal of Mammalogy 82: BRADLEY, R. D., D. S. CARROLL, M.L.HAYNIE, R.MUÑIZ-MARTÍNEZ, M. J. HAMILTON, AND C. W. KILPATRICK A new species of Peromyscus from western Mexico. Journal of Mammalogy 85: BRADLEY, R. D., N. D. DURISH, D.S.ROGERS, J.R.MILLER, M.D. ENGSTROM, AND C. W. KILPATRICK Toward a molecular phylogeny for Peromyscus: evidence from mitochondrial cytochrome-b sequences. Journal of Mammalogy 88: BRADLEY, R. D., I. TIEMANN-BOEGE, C. W. KILPATRICK, AND D. J. SCHMIDLY Taxonomic status of Peromyscus boylii sacarensis: inferences from DNA sequences of the mitochondrial cytochrome-b gene. Journal of Mammalogy 81:

10 February 2014 BRADLEY ET AL. A NEW SPECIES OF PEROMYSCUS 185 BURT, W. H Description of heretofore unknown mammals from islands in the Gulf of California Mexico. Transactions of the San Diego Society of Natural History 7: CABRERA, H., S. T. ÁLVAREZ-CASTAÑEDA, N.GONZÁLEZ-RUIZ, AND J. P. GALLO-REYNOSO Distribution and natural history of Schmidly s deermouse (Peromyscus schmidlyi). Southwestern Naturalist 52: CARLETON, M. D Interrelationships of populations of the Peromyscus boylii species group (Rodentia, Muridae) in western Mexico. Occasional Papers of the Museum of Zoology, University of Michigan 675:1 47. CARLETON, M. D., D. E. WILSON, A.L.GARDNER, AND M. A. BOGAN Distribution and systematics of Peromyscus (Mammalia: Rodentia) of Nayarit. Smithsonian Contributions to Zoology 352:1 46. COMMITTEE FOR STANDARDIZATION OF CHROMOSOMES OF PEROMYSCUS Standardized karyotype of deer mice, Peromyscus (Rodentia). Cytogenetics and Cell Genetics 19: DAVIS, K. M., S. A. SMITH, AND I. F. GREENBAUM Evolutionary implications of chromosomal polymorphisms in Peromyscus boylii from southwestern Mexico. Evolution 40: FELSENSTEIN, J Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: FERRUSQUÍA-VILLFRANCA, I Geology of Mexico: a synopsis. Pp in Biological diversity of Mexico: origins and distribution (T. P. Ramamoorthy, R. Bye, A. Lo, and J. Fa, eds.). Oxford University Press, New York. GREENBAUM, I. F., ET AL Cytogenetic nomenclature of deer mice, Peromyscus (Rodentia): revision and review of the standardized karyotype. Cytogenetics and Cell Genetics 66: HAFNER, M. S., W. L. GANNON, J.SALAZAR-BRAVO, AND S. T. ALVAREZ- CASTAÑEDA Mammal collections in the Western Hemisphere: a survey and directory of existing collections. Allen Press, Lawrence, Kansas. HALL, E. R The mammals of North America. 2nd ed. John Wiley & Sons, Inc., New York 2: þ 90. HAMMER, O., D. A. T. HARPER, AND P. D. RYAN PAST: paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4:1 9. HOOPER, E. T Notes on mammals of western Mexico. Occasional Papers of the Museum of Zoology, University of Michigan 565:1 26. HOUSEAL, T. W., I. F. GREENBAUM, D.J.SCHMIDLY, S.A.SMITH, AND K. M. DAVIS Karyotypic variation in Peromyscus boylii from Mexico. Journal of Mammalogy 68: HUELSENBECK, J. P., AND F. R. RONQUIST MrBayes: Bayesian inference for phylogeny. Biometrics 17: KIMURA, M A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: LAWLOR, T. E Evolution of Peromyscus on northern islands in the Gulf of California, Mexico. Transactions of the San Diego Society of Natural History 16: LEE, M. R., D. J. SCHMIDLY, AND C. C. HUEEY Chromosomal variation in certain populations of Peromyscus boylii and its systematic implications. Journal of Mammalogy 53: LÓPEZ-GONZÁLEZ, C., D. F. GARCÍA-MENDOZA, AND M. M. CORREA- RAMÍREZ Morphologic characterization of Peromyscus schmidlyi (Rodentia: Cricetidae), an endemic of the Sierra Madre Occidental, Mexico. Journal of Mammalogy 94: MERRIAM, C. H. 1898a. Mammals of Tres Marias Islands, off western Mexico. Proceedings of the Biological Society of Washington 12: MERRIAM, C. H. 1898b. Descriptions of twenty new species and a new subgenus of Peromyscus from Mexico and Guatemala. Proceedings of the Biological Society of Washington 12: ORDÓÑEZ-GARZA, N., AND R. D. BRADLEY Peromyscus schmidlyi. Mammalian Species 43(872): OSGOOD, W. H Thirty new mice of the genus Peromyscus from Mexico and Guatemala. Proceedings of the Biological Society of Washington 17: OSGOOD, W. H Revision of the mice of the American genus Peromyscus. North American Fauna 28: PEPPERS, L. L., AND R. D. BRADLEY Cryptic species in Sigmodon hispidus: evidence from DNA sequences. Journal of Mammalogy 81: POSADA, D., AND K. A. CRANDALL Modeltest: testing the model of DNA substitution. Bioinformatics 14: RENNERT, P. D., AND C. W. KILPATRICK Biochemical systematics of Peromyscus boylii. II. Chromosomally variable populations from eastern and southern Mexico. Journal of Mammalogy 68: RIDGWAY, R Color standards and color nomenclature. Privately published, Washington, D.C. SAIKI, R. K., ET AL Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: SCHMIDLY, D. J Geographical variation and taxonomy of Peromyscus boylii from Mexico and the southern United States. Journal of Mammalogy 54: SCHMIDLY, D. J., AND R. D. BRADLEY Morphological variation in the Sinaloan mouse, Peromyscus simulus. Revista Mexicana de Mastozoología 1: SCHMIDLY, D. J., R. D. BRADLEY, AND P. S. CATO Morphometric differentiation and taxonomy of three chromosomally characterized groups of Peromyscus boylii from east-central Mexico. Journal of Mammalogy 69: SCHMIDLY, D. J., AND G. L. SCHROETER Karyotypic variation in Peromyscus boylii (Rodentia: Cricetidae) from Mexico and corresponding taxonomic implications. Systematic Zoology 23: SIKES, R. S., W. L. GANNON, AND THE ANIMAL CARE AND USE COMMITTEE OF THE AMERICAN SOCIETY OF MAMMALOGISTS Guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy 92: SMITH, M. F., AND J. L. PATTON The diversification of South American rodents: evidence from mitochondrial sequence data for the akodontine tribe. Biological Journal of the Linnean Society 50: SMITH, S. A Cytosystematic evidence against monophyly of the Peromyscus boylii species group (Rodentia: Cricetidae). Journal of Mammalogy 71: SWOFFORD, D. L PAUP: phylogenetic analysis using parsimony (*and other methods), version 4.0b10. Sinauer Associates, Inc., Publishers, Sunderland, Massachusetts. THOMAS, O On three new forms of Peromyscus obtained by Dr. Hans Gadow, F. R. S., and Mrs. Gadow in Mexico. Annals and Magazine of Natural History, Series 7, 11: TIEMANN-BOEGE, I., C. W. KILPATRICK, D.J.SCHMIDLY, AND R. D. BRADLEY Molecular phylogenetics of the Peromyscus boylii species group (Rodentia: Muridae) based on mitochondrial cytochrome b sequences. Molecular Phylogenetics and Evolution 16: