OUP user menu

Molecular Systematics of the Peromyscus truei Species Group

Nevin D. Durish , Kristina E. Halcomb , C. William Kilpatrick , Robert D. Bradley
DOI: http://dx.doi.org/10.1644/BER-115.1 1160-1169 First published online: 21 December 2004

Abstract

DNA sequences from the mitochondrial cytochrome-b gene were used to examine the composition and phylogenetic relationships of the Peromyscus truei species group. Thirty-one individuals from the southwestern United States and Mexico were examined. Results indicated that 6 Peromyscus species (attwateri, difficilis, gratus, nasutus, pectoralis, and truei) compose the P. truei species group and that the group should be divided into 2 species assemblages. The relationship of the P. truei species group to the P. boylii and P. aztecus species groups was unresolved, indicating a close association of these 3 groups. In addition, levels of sequence divergence between sister taxa were higher than those reported for other sister species of Peromyscus.

Key words
  • cytochrome-b gene
  • DNA sequences
  • Peromyscus
  • Peromyscus truei species group

The Peromyscus truei species group is 1 of 11 species groups of the subgenus Peromyscus (Carleton 1989; Musser and Carleton 1993). The truei species group described originally by Osgood (1909) contained 5 Peromyscus species (bullatus, difficilis, nasutus, polius, and truei). Since Osgood (1909) revision, several taxa have been added or removed from the truei group. Most notable has been the removal of P. polius (Hoffmeister 1951; Kilpatrick and Zimmerman 1975; Schmidly et al. 1985; Zimmerman et al. 1975, 1978) and the addition of Peromyscus attwateri (DeWalt et al. 1993; Janecek 1990; Sullivan et al. 1991; Tiemann-Boege et al. 2000). More recently, Tiemann-Boege et al. (2000) suggested that Peromyscus pectoralis, P. polius, and Peromyscus sagax might be affiliated with the truei species group, but they conservatively placed them as incertae sedis. In addition, Tiemann-Boege et al. (2000) questioned the validity of a monophyletic truei species group. They based this premise on the lack of support between a “truei clade” and a “difficilis clade.” Although, their position was tenuous, they suggested that a difficilis assemblage (attwateri, difficilis, and nasutus) and a truei assemblage (gratus and truei) be recognized.

Several taxonomic revisions over the last 100 years have been suggested that, although systematically important, have not affected the overall composition of the P. truei group as envisioned by Osgood (1909). First, Blair (1943) described Peromyscus comanche as a new species in the truei species group. This position was supported by Johnson and Packard (1974); however, Hoffmeister (1951), Hoffmeister and de la Torre (1961), Schmidly (1973) and Modi and Lee (1984) relegated P. comanche to subspecific status within P. nasutus, P. difficilis, and P. truei, respectively. Second, Hoffmeister (1951) relegated P. dyselius and P. montiponoris to subspecific level within P. truei. Third, Hoffmeister and de la Torre (1961) and Janecek (1990) suggested that P. nasutus be placed in synonomy with P. difficilis. However, the karyotype of P. nasutus with a fundamental number (FN) of 58 is differentiated from that of P. difficilis with FN = 56 (Hsu and Arrighi 1968). Consequently, Zimmerman et al. (1975, 1978) and Avise et al. (1979) recommended that P. nasutus be recognized as a separate species, whereas the study by DeWalt et al. (1993) was inconclusive. Fourth, Modi and Lee (1984) recognized the southern cytotype of P. truei as being distinct from a northern cytotype and suggested that P. truei be split into 2 species (P. truei and P. gratus). Zimmerman et al. (1978) concurred; however, they referred the southern cytotype to P. gentilis.

The phylogenetic relationship of the P. truei species group to other species groups has been poorly understood. The chromosome and allozyme data (Avise et al. 1979; Rogers et al. 1984; Stangl and Baker 1984; Sullivan et al. 1991; Zimmerman et al. 1975, 1978) have been inconclusive or have shown an affiliation of the P. truei group with the P. boylii group. However, most of these studies did not contain a sufficient representation of other species groups. Although suffering from an inadequate representation of species groups, Tiemann-Boege et al. (2000), on the basis of DNA sequence data, depicted an unresolved relationship of the truei species group to both the boylii and aztecus groups.

The objectives of this study were 3-fold. First, reevaluate the systematic status of species within the truei species group. Second, examine the composition of the truei species group in light of the potential diphyletic nature as proposed by Tiemann-Boege et al. (2000). Third, determine the phylogenetic relationship of the truei species group.

Materials and Methods

Samples.—Thirty-one specimens representing the P. truei species group were collected from naturally occurring populations in the United States and México (Fig. 1). In addition, DNA sequences from 24 specimens reported in Sullivan et al. (1997), Bradley et al. (2000), Tiemann-Boege et al. (2000), and Bradley et al. (2004a) and 3 previously unreported sequences were included as references. When possible, > 1 specimen per taxon was examined, ensuring that variants from the extremes of the ranges were represented; for polytypic species, multiple subspecies were examined. Specimen identification numbers and collection localities are listed in Appendix I.

Fig. 1

Distribution of members of the Peromyscus truei species group in southwestern United States and México. Filled circles and filled squares represent collecting localities. Localities and GenBank accession numbers are listed in Appendix I.

Sequence data.—Mitochondrial DNA was extracted from frozen liver samples (0.1 g) following the methods of Smith and Patton (1999). The complete cytochrome-b gene, 1,143 base pairs (bp), was amplified from all individuals. The following polymerase chain reaction (PCR) parameters were modified from those described by Saiki et al. (1988): 28 cycles at 95°C denaturation (1 min), 50°C annealing (1 min), 72°C extension (1 min 10 s), and 1 final 72°C extension cycle (7 min). Primers used in PCR reactions were MVZ05 (Smith and Patton 1993) and H15915 (Irwin et al. 1991), and PCR products were purified with the QIAquick PCR purification kit (Qiagen, Valencia, California). Seven primers were used in cycle sequencing reactions to amplify 400-bp fragments on forward and reverse strands: CWEl (Edwards et al. 2001), WDRAT400F, Pero3′, WDRAT400R, Neo 700L (Tiemann-Boege et al. 2000), SIG610 (Peppers and Bradley 2000), and 752R (Bradley et al. 2000). Cycle sequencing was conducted with the ABI Prism dRhodamine terminator ready reaction mix (Applied Biosystems, Foster City, California) or ABI Big Dye version 3.0 ready reaction mix (Applied Biosystems, Foster City, California) and approximately 60–80 ng of PCR product. Sequencing conditions included 29 cycles at 94°C for 30 s (denaturing), 50°C for 20 s (annealing), and 60°C for 3 min (extension). Reactions were then purified with ethanol and sodium acetate (5 M). Samples were analyzed (both strands) on an ABI Prism 310 automated sequencer (Applied Biosystems). Sequencher 3.1 software (Gene Codes, Ann Arbor, Michigan) was used to align and proof nucleotide sequences. All cytochrome-b sequences obtained in this study were deposited in GenBank; accession numbers are listed in Appendix I.

Data analyses.—On the basis of phylogenetic relationships presented in Tiemann-Boege et al. (2000) and Bradley et al. (2004a), Reithrodontomys fulvescens, Ochrotomys nuttalli, and Onychomys arenicola were used as outgroup taxa in all analyses. In addition, Osgoodomys banderanus and 22 species of Peromyscus were included as reference samples. Variable nucleotide positions within the data set were treated as unordered, discrete characters with 4 possible states: A, C, G, or T.

Parsimony trees (PAUP*—Swofford 2002) were constructed with the use of equally weighted characters. The heuristic search and tree-bisection-reconnection options were used to obtain the most parsimonious tree(s). All phylogenetically uninformative characters were excluded from these analyses. Bootstrap analysis (Felsenstein 1985) with 1,000 iterations and the Bremer decay index (Bremer 1994; Eriksson 1997) were used to evaluate nodal support.

Fifty-six models of molecular evolution were examined in a maximum likelihood framework with Modeltest (Posada and Crandall 1998) to determine the model of DNA evolution best fitting the data. The GTR+I+G model was identified as being most appropriate for this dataset. This model generated significantly better likelihood scores than all other models and included the following parameters: base frequencies (A = 0.3343, C = 0.3067, G = 0.1124, T = 0.2466), rates of substitution (A-C = 1.4710, A-G = 7.8624, A-T = 1.8092, C-G = 0.3354, C-T=18.8236, G-T = 1.00), proportion of invariable sites (I = 0.4697), and gamma distribution (F = 0.8805). An optimal likelihood tree, using the above parameters, was obtained with the heuristic search option in PAUP* (Swofford 2002).

A Bayesian model (MrBayes—Huelsenbeck and Ronquist 2001) was used for comparison to the likelihood method and to generate support values (clade probabilities). A GTR+I+G model with a site-specific gamma rate was used with the following options: 4 Markov-chains, 2 million generations, and a sample frequency of every 100th generation. After a visual inspection of likelihood scores, the first 330 trees were discarded, and the model was rerun with 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 following the criteria outlined in Bradley and Baker (2001) and to allow comparison with published estimates of genetic divergence among species of Peromyscus.

Results

Complete nucleotide sequences (1,143 bp) from the mitochondrial cytochrome-b gene were obtained for 31 samples representing the P. truei species group; 26 reference samples from the genus Peromyscus, 1 sample from the genus Osgoodomys, and 3 outgroup taxa. Base frequencies were similar to those reported for other species of rodents with A = 33.4%, C = 30.7%, G = 11.2%, and T = 24.7%. In describing results and topologies obtained from the various analyses, an emphasis was placed on relationships for members of the P. truei species group; relationships of other taxa were described only if relevant to the objectives of this study.

The parsimony analysis with equally weighted characters generated 12 equally parsimonious trees with a length of 2,332 steps, a consistency index (CI) of 0.257, and a retention index (RI) of 0.641. A strict consensus tree was generated (bootstrap and Bremer support values provided above and below branches, respectively) and is shown in Fig. 2. Putative members of the P. truei species group (attwateri, difficilis. gratus, nasutus, pectoralis, and truei) were arranged into 4 well-supported clades (I–IV). Clade I contained all samples of P. attwateri, P. difficilis, and P. nasutus; clade II contained samples of P. pectoralis; clade III contained all samples of P. gratus; and clade IV contained only samples of P. truei. Although the membership within each clade received strong support (bootstrap and Bremer), the association of the 4 clades was not supported. Similarly, the remaining reference samples formed small clades that primarily reflected species groups as defined by Carleton (1989); however, the association between clades received little or no support. P. pectoralis, P. polius, and P. sagax, indicated by Tiemann-Boege et al. (2000) as possible members of the P. truei species group, received no support for association with taxa of the truei group.

Fig. 2

Strict consensus tree representing the 12 most parsimonious trees obtained from a parsimony analysis of unweighted DNA sequence characters from the mitochondrial cytochrome-b gene obtained from representative members of Peromyscus. Roman numerals refer to clades discussed in the text. Numbers above branches represent bootstrap values, and numbers below branches refer to Bremer support values.

In the Bayesian analysis (Fig. 3), P. attwateri, P. difficilis, and P. nasutus formed a well-supported clade (I, clade probability value = 100), with P. difficilis and P. nasutus as sister taxa. Within the difficilis and nasutus clade, 3 subclades (A–C) were depicted. Subclade A contained samples of P. difficilis from southern México, subclade B contained samples of P. nasutus and a sample of P. difficilis from Durango, and subclade C contained the sample of P. difficilis from Aguascalientes. Clade probability values were high (100) for subclades A and B; however, little support was provided for the association of subclades A–C. Samples of P. pectoralis formed a separate clade (II) that was basal to clade I (difficilis, nasutus, and attwateri); however, this relationship was not well supported by clade probability value (84). Clade III contained P. gratus; however, this clade was not well supported (clade probability = 68). Clade IV contained samples of P. truei and was strongly supported (clade probability = 100). Clade III and IV (P. gratus and P. truei) then joined to form a well-supported clade (clade probability = 100).

Fig. 3

Tree generated by Bayesian methods (MrBayes; Huelsenbeck and Ronquist 2001) and the GTR+I+G model of evolution obtained from representative members of Peromyscus. Roman numerals and capital letters refer to clades discussed in the text. Clade probability values are shown above branches.

The aztecus species group was depicted as the sister group to the “truei species group” (difficilis, nasutus, attwateri, pectoralis, gratus, and truei); this clade, containing the 2 species groups, was then sister to a clade containing members of the boylii species group (Fig. 3). However, placement of the aztecus and boylii clades were not supported by clade probability values.

The maximum likelihood analysis produced an identical topology (not shown) as the Bayesian analysis, with 3 exceptions. First, the sample of P. difficilis from Aguascalientes was basal to a clade containing all the samples of P. difficilis, other than the Durango sample. Second, the P. aztecus group was sister to the P. boylii species group. Third, P. crinitus was basal to a clade containing O. banderanus, P. eremicus, P. leucopus, P. maniculatus, and P. melanotis. However, the clade probabilities were low for all of these differences in topology.

For selected taxa, average genetic distances were obtained from the Kimura 2-parameter model of evolution (Kimura 1980) and are shown in Table 1. Within the “truei species group,” intraspecific comparisons ranged from 1.09% (difficilis) to 4.34% (pectoralis), whereas interspecific comparisons ranged from 7.73% (difficilis-nasutus) to 14.21% (difficilis-truei).

View this table:
Table 1

Average genetic distances (Kimura 1980) for select taxa and comparison of samples examined in this study.

Peromyscus comparisonAverage (%)
Within species
P. attwateri1.69
P. difficilis1.09
P. nasutus2.53
P. pectoralis4.34
P. truei2.61
P. gratus2.82
P. g. gratus versus P. g. gentilis1.76
P. g. gratusP. g. gentilis versus P. g. zapotecae4.53
P. t. Comanche versus P. t. truei1.20
P. t. truei (westem) versus P. t. truei (eastern)4.50
P. difficilis (Durango) versus P. difficilis8.19
P. difficilis (Durango) versus P. nasutus3.62
P. difficilis (Aguascalientes) versus P. difficilis7.38
P. difficilis (Aguascalientes) versus P. nasutus7.54
Between species
P. attwateri versus P. difficilis8.89
P. attwateri versus P. nasutus8.62
P. attwateri versus P. pectoralis10.54
P. attwateri versus P. truei13.85
P. attwateri versus P. gratus11.09
P. difficilis versus P. nasutus7.73
P. difficilis versus P. pectoralis12.07
P. difficilis versus P. truei14.21
P. difficilis versus P. gratus10.93
P. nasutus versus P. pectoralis11.39
P. nasutus versus P. truei13.41
P. nasutus versus P. gratus11.33
P. pectoralis versus P. truei13.15
P. pectoralis versus P. gratus10.53
P. truei versus P. gratus10.63
Among species
P. truei versus P. boylii13.11
P. truei versus P. aztecus13.75
P. aztecus versus P. boylii11.19

Discussion

Given the lack of resolution. of deeper nodes in the parsimony analysis and similarity of topologies produced by Bayesian and likelihood analyses, the latter 2 analyses are used to discuss phylogenetic relationships and taxonomic implications of the deeper nodes. However, all 3 analyses were relatively well supported at the tips and were used to evaluate relationships at the more shallow nodes. Taxonomic implications were addressed by the phylogenetics species concept (Cracraft 1983) and genetic species concept (Dobzhansky 1950) as modified by Bradley and Baker (2001).

Relationships within species.—Five taxa (dijficilis, nasutus, gratus, truei, and pectoralis) possessed sufficient sample sizes to tentatively address either geographic variation or taxonomy. First, within P. pectoralis, the 2 samples of P. p. pectoralis formed a sister clade to the sample of P. p. laceianus. Although this relationship was expected, the magnitude of genetic differentiation (5.98%) between the 2 subspecies was not anticipated. This genetic distance is greater than that typically seen between sister species within the genus Peromyscus (Bradley and Baker 2001; Bradley et al. 2004a, 2004b). Kilpatrick and Zimmerman (1976) reported levels of genetic similarity among subspecies of P. pectoralis below the average range of values for conspecific populations and suggested that this genetic differentiation resulted from events associated with the formation of 3 proposed Pleistocene refugiai populations. Patterns of heterozygosity and of occurrence of hemoglobin genotypes, however, were interpreted as the products of gene flow among populations recolonized from these refugiai populations and the absence of reproductive isolation (Kilpatrick and Zimmerman 1976). Further comment on the significance of this observed sequence divergence is reserved until additional samples of P. pectoralis from the eastern portion of their range can be examined.

Second, 8 samples of P. difficilis were examined, 6 of which formed a well-supported subclade (A) in the parsimony, likelihood, and Bayesian analyses (Fig. 3). Samples making up this clade were from south-central México and unequivocally can be referred to as P. difficilis. These 6 samples were represented by 3 Peromyscus subspecies (amplus, difficilis, and saxicola) and possessed an average genetic distance of 1.09%, indicating littie variation among these subspecies. The remaining 2 samples, from north-central México (Aguascalientes and Durango), failed to form a monophyletic clade with other samples ofP. difficilis (Figs. 2 and 3). The sample from Durango was sister to P. nasutus (subclade B), with an average genetic distance of 3.62%, a value slightly higher than the average genetic distance value (2.53%) obtained from intraspecific comparisons of P. nasutus. Clearly, this sample is affiliated with P. nasutus instead of P. difficilis, from which it differs by 8.19% (Table 1). The occurrence of P. nasutus in northern México is not surprising because Diersing (1976) commented on the lack of a clear designation of the geographic distribution of P. difficilis and P. nasutus. Robbins and Baker (1981) reported a nasutus-like karyotype (FN = 58) from Durango, and their data, in conjunction with the DNA data, suggest that the distribution of P. nasutus extends along the Sierra Madre Occidental into northwestern México, The sample from Aguascalientes (subclade C) formed a separate clade that was either basal to the P. difficilis clade (likelihood analysis) or to a clade that was basal to the P. difficilis and P. nasutus clade (subclade C, Fig. 3). Compared with other samples of P. difficilis or P. nasutus, this sample possessed an average genetic distance of 7.38% and 7.54%, respectively. These values are comparable to the level of genetic divergence observed between P. difficilis and P. nasutus (7.70%). At this time, too few samples exist for any valid taxonomic suggestion; however, further studies are warranted to ascertain whether an undescribed taxon exists in this region.

Seven samples of P. gratus representing 3 subspecies (gentilis, gratus, and zapotecae) were examined (clade III, Fig. 3). The samples of P. g. gentilis formed a monophyletic clade that was sister to the sample representative of P. g. gratus. Although this larger clade was then sister to the 2 samples of P. gratus zapotecae, the average genetic distance separating the samples of P. g. gratus and P. g. gentilis from P. g. zapotecae was 4.53% compared with 1.76% for samples of P. g. gratus and P. g. gentilis. This is in agreement with Hooper (1957) assessment that P. g. zapotecae was unique from P. g. gratus and P. g. gentilis. Given the potential of geographical isolation of P. g. zapotecae (south of the Trans-Mexican Volcanic Zone) and a high level of genetic divergence compared with values generated for other species of Peromyscus, we suggest that further studies are warranted.

Nine samples of P. truei representing 4 subspecies were examined (clade IV, Fig. 3). Two clades were apparent: 1 representing an eastern group (comanche and truei) and a 2nd representing a western group (gilberti and montipinoris). Although the average genetic distance separating the 2 clades was 4.50%, indicating a level of genetic divergence that approached values observed between other sister species of Peromyscus, additional sampling from intermediate areas (Fig. 1) are needed before any valid taxonomic conclusion can be reached. The low level of genetic divergence between P. t. truei and P. t. comanche (1.20%) confirms the assessment of Schmidly (1973) and Modi and Lee (1984) that comanche should be recognized as a subspecies of P. truei.

Relationships between species.—On the basis of the Bayesian and likelihood analyses, 4 clades were formed, with clade I containing attwateri, difficilis, nasutus; clade II containing pectoralis; clade III containing truei; and clade IV containing gratus. Although the sister relationship of truei and gratus (clades III and IV) are well supported (clade probability = 100), phylogenetic relationships within clade I are more enigmatic. First, P. difficilis and P. nasutus are sister species (Figs. 2 and 3), supporting the contentions of Zimmerman et al. (1975, 1978) and Avise et al. (1979). However, it is unclear where the geographic boundaries separating the 2 species lie. Additionally, it should be noted that the taxonomic affinity of the sample from Aguascalientes (subclade C) could complicate this arrangement; especially if it is an undescribed taxon. The arrangement of P. attwateri as sister to the clade containing P. difficilis and P. nasutus is well supported (clade probability = 100) and is consistent with the conclusions of Janecek (1990), DeWalt et al. (1993), Sullivan et al. (1991), and Tiemann-Boege et al. (2000). The 2nd problem concerns the phylogenetic position of P. pectoralis. Tiemann-Boege et al. (2000) reported the possible association of P. pectoralis with the P. truei species group. Likewise, in both the Bayesian and likelihood analyses, P. pectoralis (clade II) was sister to a clade (I) containing attwateri, difficilis, and nasutus, although a clade probability value of 84 renders no support for this claim. However, the consistent formation of this monophyletic clade relative to the other reference samples of Peromyscus should be considered, and we tentatively conclude that P. pectoralis is likely the sister group to the clade containing P. attwateri, P. difficilis, and P. nasutus.

Relationship of species groups.—One of the goals of this study was to address the composition of the truei species group. Tiemann-Boege et al. (2000) found support for a truei assemblage (gratus and truei) and a difficilis assemblage (attwateri, difficilis, and nasutus) but failed to find support that unified the 2 assemblages into a clade that could be referred to as a species group. Their study was further confounded by the association (albeit poor) of P. pectoralis, P. polius, and P. sagax to members of the truei and difficilis assemblages. It was our hope that additional samples and taxa would assist in resolving this problem.

On the basis of this study, no support was evident for the inclusion of P. polius and P. sagax in the P. truei species group as tentatively suggested by Tiemann-Boege et al. (2000). The P. truei species group appears to be composed of P. attwateri, P. gratus, P. difficilis, P. nasutus, P. pectoralis, and P. truei. The difficulty in recognizing the above taxa as members of the truei species group stems 1st from the association of P. pectoralis to the clade containing P. attwateri, P. difficilis, and P. nasutus (clade probabihty = 84) and 2nd from the dichotomy (clade probability = 86) produced in joining the subclade containing P. attwateri, P. difficilis, P. nasutus, and P. pectoralis and the subclade containing P. gratus and P. truei. Although these clade probability values are lower than those typically considered as evidence for support (Alfaro et al. 2003; Douady et al. 2003), P. pectoralis is aligned consistently within this group of taxa (attwateri, difficilis, nasutus, gratus, truei) that have been included in the truei group on the basis of the analyses of independent data (DeWalt et al. 1993; Janecek 1990; Sullivan et al. 1991). Unfortunately, we were unable to include a sample of P. bullatus, although Hooper (1968) suggested it would prove to be a subspecies of P. difficilis, and its inclusion could help resolve this issue.

Two assemblages (difficilis and truei) are supported as suggested by Tiemann-Boege et al. (2000). The difficilis assemblage contains attwateri, difficilis, nasutus, and pectoralis, although the placement of pectoralis is not well supported by clade probability values. The truei assemblage contains only truei and gratus and is well supported (clade probability value = 100).

As demonstrated by Tiemann-Boege et al. (2000) the relationship of the truei species group to either the aztecus or boylii species groups is unresolved. The Bayesian analysis showed truei and aztecus to be sister groups (clade probability value = 45) followed by the addition of the boylii species group (clade probability value = 69); whereas, the likelihood analysis showed a sister relationship between the aztecus and boylii species group followed by the addition of the truei group. Alternatively, genetic distances were least between aztecus and boylii (11.19%), followed by boylii to truei (13,11%) and aztecus to truei (13.75%). Given the lack of resolution among these 3 species groups, it appears that they had a relatively similar time of divergence. Alternatively, it could be that saturation of transitional substitutions at 3rd positions might produce homoplastic change. To resolve this issue, we eliminated all 3rd positions, used transversions only (eliminated all transitions), and translated nucleotides to amino acids; all attempts produced trees with no support for deep nodes (basically, no resolution beyond the species level). It appears that DNA sequences from other regions might be required to provide resolution of the 3 species groups.

Pairwise comparisons of DNA sequence divergence values, obtained from sister species within the P. truei group, indicated an elevated rate of sequence divergence compared with values obtained from sister species in the P. boylii and P. aztecus species groups (Bradley and Baker 2001; Bradley et al. 2004a; Tiemann-Boege et al. 2000). Typically, levels of sequence divergence for sister species ranged from approximately 3% to 5%, whereas values within the P. truei species group ranged from 7.73% (P. dijficilis and P. nasutus) to 10.63% (P. gratus and P. truei). Whether this is indicative of an increase in mutation rate or an older divergence time for the P. truei species group is unknown.

Resumen

Secuencias de ADN obtenidas del gen cytochrome b fueron usadas para evaluar la composición y las relaciones filogené-ticas del grupo de especies Peromyscus truei. Treinta y un individuos provenientes del suroeste de los Estados Unidos y de México fueron examinados. Los resultados indican que 6 especies (attwateri, difficilis, gratus, nasutus, pectorales, y truei) forman el grupo de especies P. truei y que el grupo debería de ser dividido en 2 conjuntos. La relación entre el grupo de especies P. truei y los grupos de especies P. boylii y P. aztecus quedó inconclusa, lo cual indica una asociación próxima entre estos 3 grupos. Además, los niveles de divergencia entre secuencias provenientes de especies o grupos hermanos fueron más altos que los reportados por otras especies hermanas en el género Peromyscus.

Acknowledgments

We thank B. R. Amman, B. D. Baxter, J. D. Hanson, M. L. Haynie, L. K. Longhofer, and F. Mendez-Harclerode for reviewing earlier versions of the manuscript. Tissue samples were kindly provided by the Natural Science Research Laboratory, Museum of Texas Tech University (R. J. Baker); Zadock-Thompson Natural History Collection, University of Vermont; and Monte L. Bean Museum of Natural History, Brigham Young University (D. S. Rogers). Tissue samples were obtained with the use of guidelines approved by the American Society of Mammalogists (Animal Care and Use Committee 1998). Special thanks goes to the Field Methods classes of 1997 and 2000 for collecting specimens. Support for research was obtained from the Howard Hughes Medical Institute, Undergraduate Biological Sciences Education Program to Texas Tech University (N. D. Durish), and a National Institutes of Health grant (DHHS A141435-01) to R. D. Bradley.

Appendix I

Specimens examined in this study.–For each specimen, the collection locality (see Fig. 1), museum acronym (MVZ, TCWC, TTU, UMMZ, USNM, and ZTNHC, following Hafner et al. [1997]), museum or collector number (CWK, C. William Kilpatrick; DNA, Zadock Thompson Natural History Collection DNA sample; DSR, Duke S. Rogers; GK, Ira I. Greenbaum; TK, K. Nutt and Texas Tech University tissue number), and GenBank accession number (AF, AY, and U) are provided in parentheses. Some sequences were obtained from GenBank and deposited by Sullivan et al. (1997), Bradley et al. (2000), or Tiemann-Boege et al. (2000). Specimens are from the United States unless otherwise noted.

Ochrotomys nuttaUi.—Texas; Wood County, 3.5 miles SE Quitman (TTU: TK 31929; AY 195798).

Onychomys arenicola.—Texas; Presidio County, Big Bend Ranch State Natural Area (TTU: TK 46462; AY195793).

Osgoodomys banderanus.—MEXICO: Jalisco; 6 km SE Chamela (TTU: TK 11796; AF155383).

Peromyscus attwateri.—Oklahoma; Mcintosh County, 3.1 miles E Dustin (TTU: TK 23396; AF 155384); Texas; Knox County, 3 miles E Benjamin (ZTNHC: DNA 27; AF 155385).

Peromyscus aztecus.—MEXICO: Veracruz; 5.5 miles N Huatusco (TCWC: GK4053, U89968).

Peromyscus beatae beatae.—MEXICO: Veracruz; Xometla (TCWC: GK 3954; AF131921).

Peromyscus boylii boylii.—California; Monterey County, Hastings Natural History Reservation (MVZ: Nutt 120; AF155386).

Peromyscus boylii glasselli.—MEXICO: Sonora; Isla San Pedro Nolasco (UMMZ: UMMZ 117347; AF155387).

Peromyscus boylii rowleyi.—MEXICO: Aguascalientes; Rincon de Romos (TCWC: GK 4114; AF131924).

Peromyscus californicus.—California; San Diego County, San Onofre State Beach, 3.5 miles NNE on Christmas Road (TTU: TK 83632; AF155393).

Peromyscus crinitus.—Utah; Vintah County, Cottonwood Canyon, 39°16′51.8″N, 111°10′31.9″W (BYU: DSR 6171; AY376413).

Peromyscus difficilis amplus.—MEXICO: Puebla; 8 miles SE Chignahuapan (ZTNHC: CWK 2770; AY376414); Tlaxcala; Mt. Malinche (TCWC: GK 3904; AY376415); 18 km N, 9 km E Apizaco (TTU: TK 13084; AY387488); 2 km NE Tepetitla (TTU: TK 93120; AY376416).

Peromyscus difficilis difficilis.—MEXICO: Durango; 50 km W Las Herreras (TTU: TK 48580; AY376417); Aguascalientes; 6 miles W Rincon de Romas (TCWC: GK 4129; AY376418).

Peromyscus difficilis saxicola.—MEXICO: Hidalgo; 5.4 miles SE, 3.2 miles S Ixmiquilpan (TCWC: GK 2642; AF 155394); 1.8 miles E Jonacapa (TCWC: GK 3076; AY376419).

Peromyscus eremicus.—Arizona; Yavapai County, Sycamore Station, 34°23′28.2″N, 112°3′1.3″W (TTU: TK 92750; AY195799); California; Los Angeles County, Calabasas Creekside Park (TTU: TK 91184; AY322503).

Peromyscus evides.—MEXICO: Oaxaca; 3.5 miles S Suchixtepec (TCWC: GK 3439, U89970).

Peromyscus gratus gentilis.—MEXICO: Jalisco; 2 km NW Mesconcitos (TTU: TK 93079; AY376420); Durango; 2.2 km S, 2.5 km E Vicente Guerrero (TTU: TK 48800; AF155395; TTU: TK 48799, AY387489); 3.8 miles W Coyotes, Hacienda Coyotes (TTU: TK 72333; AY322507).

Peromyscus gratus gratus.—MEXICO: Michoacán; 4 km E Costzeo (TTU: TK 46354; AY376421).

Peromyscus gratus zapotecae.—MEXICO: Puebla; 5 km SE San Antonia (TTU: TK 93140; AY376422; TTU: TK 93145; AY376423).

Peromyscus hylocetes.—MEXICO: Michoacán; Puerto Gamica (ZTNHC: CWK 2035, U89976).

Peromyscus levipes ambiguus.—MEXICO: Nuevo Leon; Cola de Caballo (TCWC: GK 3840; AF131928).

Peromyscus levipes levipes.—MEXICO: Tlaxcala; 2 km W Teacalco (TCWC: GK 4301; AF131929).

Peromyscus madrensis.—MEXICO: Nayarit; Isla María Madre (USNM: USNM 512599; AF155397).

Peromyscus maniculatus.—Obtained from GenBank, no locality provided (AF119261).

Peromyscus melanophrys.—MEXICO: Jalisco; 30 km W Huejuqilla del Alto (TTU: TK 48638; AY376424).

Peromyscus melanotis.—MEXICO: Durango; 12 km E Ojitos (TTU: TK 70997; AF155398).

Peromyscus mexicanus.—MEXICO: Chiapas; 9 miles N Ocozo-coaulta (TTU: TK 93314; AY376425).

Peromyscus nasutus griseus.—New México; Lincoln County, 4 miles S Carrizozo (TTU: TK 77922; AF155399).

Peromyscus nasutus nasutus.—Texas; Jeff Davis County, Mt. Livermore Preserve (TTU: TK 83576; AY376426).

Peromyscus oaxacensis.—MEXICO: Oaxaca; 0.9 miles N Llano de las Flores (TCWC: GK 3516, U89972).

Peromyscus pectoralis laceianus.—Texas; Kimble County, Walter Buck Wildlife Management Area (TTU: TK 52050; AF155400).

Peromyscus pectoralis pectoralis.—MEXICO: Durango; 1.5 km SE Las Herreras (TTU: TK 48567; AF155401); Jalisco; Huejuquilla del Alto (TTU: TK 48642; AY376427).

Peromyscus polius.—MEXICO: Chihuahua; 3 miles SW Santa Barbara (TTU: TK 47255; AF155403).

Peromyscus sagax.—MEXICO: Michoacán; Puerto Gamica (ZTNHC: CWK 2032; AF155404).

Peromyscus simulus.—MEXICO: Sinaloa; 4 miles E Concordia, Highway 40 (TCWC: GK 3222; AF131927).

Peromyscus spicilegus.—MEXICO: Durango; San Juan de Cama-rones (TTU: TK 70912: AY322512).

Peromyscus stephani.—MEXICO: Sonora; Isla San Esteban (UMMZ: UMMZ 117385; AF155411).

Peromyscus truei comanche.—Texas; Armstrong County, 0.75 miles N 6.25 miles E Wayside (TTU: TK 40211; AY376428; TTU: TK 40215; AY376429); Briscoe County, 3 miles N Quitaque, Caprock Canyon (TTU: TK 21856; AY376430); Caprock Canyon State Park (TTU: TK 54856; AY376431).

Peromyscus truei gilberti.—Califomia; Alameda County, Strawberry Canyon (MVZ: MVZ 157329; AF108703).

Peromyscus truei montipinoris.—Califomia; Los Angeles County, Chatsworth Reservoir Park (TTU: TK 92333; AY376432).

Peromyscus truei truei.—Arizona; Apache County, 35°34′51″N, 109°34′33″W (TTU: TK 92372; AY376433); Navajo County, 3 miles S Woodruff (TTU: TK 77921; AF155412); New México; Socorro County, 32 miles S, 23.5 miles W Socorro (TTU: TK 13474; AY376434).

Peromyscus winkelmanni.—MEXICO: Michoacán; 6.9 miles WSW Dos Aguas (TCWC: GK 3311; AF131930).

Reithrodontomys fulvescens.—Oklahoma; McIntosh County, 3.1 miles E Dustin (TTU: TK 23469; AF176257).

Footnotes

  • Associate Editor was Eric A. Rickart.

Literature Cited

View Abstract