Reevaluation of the phylogenetic relationships among Neotomini rodents (Hodomys, Neotoma, and Xenomys) and comments on the woodrat classification

Abstract

The woodrats or packrats of the genus Neotoma have been the subject of a wide array of research including paleoecology, physiology, morphological evolution, systematics, speciation, and hybridization. In recent years, much work has been done to elucidate evolutionary relationships within and between closely related species of the genus; in particular the addition of newly collected specimens from critical geographic regions has provided new opportunities for taxonomic assessment. Given these new data and their potential, parsimony (PARS), maximum likelihood (ML), and Bayesian inference (BI) analyses were conducted on DNA sequences obtained from nine individual genes (four mitochondrial loci: 12S, 16S, CoII, and Cytb; five nuclear loci: AdhI2, BfibI7, En2, Mlr, and Myh6) to estimate the phylogenetic relationships among 23 species of Neotoma. Results of these analyses depicted a wide array of phylogenetic relationships among taxa; with substantial nodal support recovered in both the ML and PARS analyses at some mid-level and terminal positions. Several individual genes, particularly 12S, AdhI2, BfibI7, CoII, and Cytb, provided support at several basal positions; however, phylogenetic resolution was limited in the other genes. A final BI analysis where the nine genes were concatenated into a single data set produced several supported clades that corresponded to previously recognized species groups (floridana, micropus, mexicana, and lepida) and the subgenus Homodontomys. Levels of genetic divergence for within-species comparisons (estimated from the Cytb data set) ranged from 0.88% (N. magister) to 6.82% (N. fuscipes); for between sister species comparisons ranged from 4.68% (N. devia and N. lepida) to 12.70% (N. angustapalata and N. nelsoni); and for members within closely related clades ranged from 8.70% (N. bryanti and N. lepida) to 12.57% (N. goldmani and N. magister). Evaluations of generic, subgeneric, and species group boundaries were explored using phylogenetic principles on the DNA sequence data presented herein, as well as morphological findings from previous studies. Results obtained suggest that the most conservative taxonomic interpretation involves the abandonment of subgeneric delineations and relies on the recognition of eight species groups (cinerea, floridana, fuscipes, lepida, mexicana, micropus, phenax, and stephensi) as the backbone of the woodrat classification.

Woodrats or packrats (genus Neotoma) are members of the New World subfamily Neotominae (Musser and Carleton 2005). Their relatively large body size (for North American rodents), propensity for constructing middens (stick houses), and tendency to collect shiny objects for decoration of middens have made woodrats one of the most easily recognizable rodents in North America. Woodrats are adapted to a variety of habitats and ecological regions and occur throughout southern Canada, most of the continental United States, Mexico, and northern portions of Central America (see Hall 1981). Twenty-three extant species currently are recognized (Musser and Carleton 2005; Patton et al. 2007; Ordóñez-Garza et al. 2014; Bradley and Mauldin 2016; Pardiñas et al. 2017) and three species are presumed to have been extirpated in recent years (Mellink 1992; Smith et al. 1993; Cortés-Calva et al. 2001).

Early alpha-level taxonomic studies (Merriam 1892, 1894; Goldman 1904, 1905, 1909, 1915, 1932; Hall and Genoways 1970; and many others) utilizing morphological data resulted in the description of several species and subspecies. Goldman (1910), in his revision of Neotoma, provided not only a comprehensive synopsis of the taxonomy and systematics for the genus, but formulated a rudimentary organization of relationships through his recognition of subgenera and species groups. Following several decades of morphometric, allozymic, and chromosomal investigations (see citations in Edwards and Bradley 2002b) the first DNA sequence-based phylogeny was developed for the genus Neotoma (Edwards and Bradley 2002b). Although the phylogenetic and taxonomic results presented in Edwards and Bradley (2002b) resulted solely from DNA sequences obtained from the mitochondrial cytochrome-b gene (Cytb) and was limited to 13 species that were available for study at that time, they were able to develop phylogenetic hypotheses concerning the relationships of species within Neotoma, determine the validity and composition of species groups, status of subgenera, and to ascertain the taxonomic status of Hodomys and Xenomys relative to Neotoma.

In an attempt to expand phylogeny reconstruction beyond mitochondrial DNA data, Longhofer and Bradley (2006) used DNA sequences obtained from intron 2 of the nuclear gene alcohol dehydrogenase (Adh1-I2) and a subset of the individuals reported in Edwards and Bradley (2002b). Their findings supported the phylogenetic relationships and topology based on examination and analysis of Cytb sequences (Edwards and Bradley 2002b). Although not a primary focus of their research, phylogenetic analysis of a multigene data set (and morphologic data) by Matocq et al. (2007) further enhanced our knowledge of the phylogenetic relationships of woodrats by: (i) expanding the data set of Edwards and Bradley (2002b) to include additional individuals per taxon, (ii) increasing the number of sequenced genes (three mitochondrial and four nuclear), (iii) expanded the taxonomic coverage by including a newly recognized taxon, N. macrotis (Matocq 2002), and (iv) adding morphologic characters associated with the male genitalia. The topology presented in Matocq et al. (2007), for the most part, was in agreement with the findings of the earlier studies by Edwards and Bradley (2002b) and Longhofer and Bradley (2006) and provided support (posterior probabilities) at additional nodes. In addition, Matocq et al. (2007) provided increased resolution for: (i) composition and interrelationships among species groups and (ii) interpretation and recognition of subgenera.

Several recent taxonomic changes warrant a reevaluation of phylogenetic relationships within Neotoma as well as a reassessment of the overall classification of Neotoma and its allies (Hodomys and Xenomys). First, Patton et al. (2007) provided a much-needed revision of the N. lepida complex. Their findings revealed that four species should be recognized (N. bryanti, N. devia, N. insularis, and N. lepida) and that N. anthonyi, N. bunkeri, and N. martinensis should be subsumed into N. bryanti. Second, Rogers et al. (2011) suggested that although N. angustapalata genetically was indistinguishable from N. leucodon, sufficient morphological differentiation existed to retain N. angustapalata as a species. Third, Edwards and Bradley (2002a) and Ordóñez-Garza et al. (2014) reevaluated the taxonomy of the N. mexicana complex and applied a senior synonym (N. ferrunginea) to woodrats in southern Mexico and portions of northern Central America; resulting in the placement of N. isthmica into synonymy with N. ferrunginea. In addition, Hernández-Canchola et al. (2021) subsumed two additional subspecies of N. mexicana (tropicalis and solitaria) into N. ferrunginea further solidifying the status of that taxon. Fourth, Fernández (2014) reeval- uated the status of N. nelsoni and suggested it was no more than a subspecies of N. leucodon. Fifth, Bradley and Mauldin (2016) recognized N. melanura as a species distinct from N. albigula. Together, these revisions indicate that 23 species reside in Neotoma (see Pardiñas et al. (2017) for a list).

Therefore, our goals were to: (i) develop a molecular phylogeny based on DNA sequences that encompasses numerous individuals per species and that includes a reasonable subsampling of the geographic distribution of each species to determine an approximation of within- and between-species genetic variation, (ii) augment the multigene data set of Matocq et al. (2007) based on recent taxonomic changes, newly available samples, and additional genetic data to examine the phylogenetic relationships within Neotoma, and (iii) produce a classification that addresses the validity of previously described species groups (Goldman 1910; Burt and Barkalow 1942; Birney 1976; Planz et al. 1996; Edwards and Bradley 2002b) and subgenera (Gray 1843; Merriam 1894, 1903; Goldman 1910; Burt and Barkalow 1942; Edwards and Bradley 2002b). To accomplish this task, DNA sequences from five mitochondrial and four nuclear genes were examined as described below.

Materials and Methods

Samples

The sampling scheme involved two stages relative to the goals outlined herein. First, DNA sequences from the entire mitochondrial cytochrome-b gene (Cytb—1,143 bp) were obtained from 691 individuals representing 21 of the 23 recognized species of Neotoma (see Appendix I) and examined to investigate intraspecific variation and to establish broad hypothesized species delineations. Second, DNA sequences were obtained from three additional mitochondrial (small subunit rRNA, 12S—531 bp; large subunit rRNA, 16S—566 bp; and cytochrome-c oxidase subunit II, CoII—633 bp) and five nuclear loci (intron 2 of alcohol dehydrogenase, Adh1-I2—577 bp; intron 7 of B-fibrinogen, BfibI7—668 bp; exon 3 of engrailed 2, En2—146 bp; exon 3 of mineralocorticoid receptor, Mlr—205 bp; and exon 35 and intron 35 of myosin heavy polypeptide 6α, Myh6—238 bp) as presented in Matocq et al. (2007) and Longhofer and Bradley (2006). For the second data set, one to two individuals per species were examined for the 21 Neotoma species. Following Edwards and Bradley (2002b), samples of Xenomys nelsoni and Hodomys alleni were included to establish a genetic benchmark for evaluating generic (Xenomys and Hodomys) and subgeneric (Homodontomys, Neotoma, Teonoma, and Teonopus) level relationships relative to Neotoma.

DNA isolation, polymerase chain reaction, and sequence methods

For most specimens, genomic DNA was isolated from approximately 0.1 g of frozen liver tissue using the DNeasy kit (Qiagen, Valencia, California). The four mitochondrial loci (12S, 16S, CoII, and Cytb) and five nuclear loci (AdhI2, BfibI7, En2, Mlr, and Myh6) were amplified using the polymerase chain reaction (PCR) method (Saiki et al. 1988) following Matocq et al. (2007) and Longhofer and Bradley (2006); specific primers and thermal profiles are listed in Table 1. PCR products were purified with a QIAquick PCR Purification Kit or QIAquick Gel Extraction Kit (Qiagen, Valencia, California). 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 checked by eye using Sequencher 4.10 software (Gene Codes, Ann Arbor, Michigan); chromatograms were examined to verify all base changes. All sequences obtained in this study were deposited in GenBank and museum specimen catalog numbers are listed in Table 2.

Table 1.

Information for specific loci examined in this study: locus being investigated, size of polymerase chain reaction (PCR) amplicons, primers used in PCR and sequencing methods, and thermal profiles for PCR reactions (all profiles included an initial 2–10 min denaturation cycle). Abbreviations for genes are as follows: small subunit rRNA, 12S; large subunit rRNA, 16S; intron 2 of alcohol dehydrogenase, Adh1-I2; intron 7 of B-fibrinogen, Bfib; cytochrome-c oxidase subunit II, CoII; cytochrome-b, Cytb; exon 3 of engrailed 2, En2; exon 3 of mineralocorticoid receptor, Mlr; and exon 35 and intron 35 of myosin heavy polypeptide 6α, Myh6. References for methods are listed in parentheses. The annealing temperature for the Adh1-I2 reaction was ramped from 53°C to 48°C to −53°C to 73°C at a rate of 0.6°C per second.

Locus	Size	Primers	Thermal profile
			Cycles	Denaturation	Annealing	Extension
12S	531 bp	12Sa (Matocq et al. 2007)	33	94°C, 1 min	62°C, 1 min	72°C, 1 min
		12Sb (Matocq et al. 2007)
16S	571 bp	16Sa (Matocq et al. 2007)	33	94°C, 1 min	45°C, 1 min	72°C, 1 min
		16Sb (Matocq et al. 2007)
Adh1-I2	583 bp	EXONII-F (Amman et al. 2006)	30	95°C, 30 s	53°C to 73°C*	73°C, 4 min
		EXONIII-R (Amman et al. 2006)
Bfib	777 bp	β17-mammL (Matocq et al. 2007)	33	94°C, 1 min	60°C, 1 min	72°C, 1 min
		βfib-mammU (Matocq et al. 2007)
CoII	633 bp	COIIa (Atkins and Honeycutt 1994)	33	94°C, 1 min	51°C, 1 min	72°C, 1 min
		COIIb (Atkins and Honeycutt 1994)
Cytb	1,143 bp	LGL765 (Bickham et al. 1995)	33	94°C, 40 s	51°C, 45 s	73°C, 1 min 20 s
		LGL766 (Bickham et al. 2004)
		700H (Peppers and Bradley 2000)
		400F (Edwards et al. 2001)
En2	146 bp	EN2f (Lyons et al. 1997)	33	94°C, 1 min	57°C, 1 min	72°C, 1 min
		EN2r (Lyons et al. 1997)
Mlr	205 bp	MLRf (Lyons et al. 1997)	33	94°C, 1 min	56°C, 1 min	72°C, 1 min
		MLRr (Lyons et al. 1997)
Myh6	236 bp	MYH2f (Lyons et al. 1997)	33	94°C, 1 min	62°C, 1 min	72°C, 1 min
		MYH2r (Lyons et al. 1997)

Table 2.

List of museum catalog and GenBank accession numbers for taxa examined in this study. No sequence data were obtained for N. chrysomelas and N. palatina. Abbreviations for genes are as follows: small subunit rRNA, 12S; large subunit rRNA, 16S; cytochrome-c oxidase subunit II, CoII; cytochrome-b, Cytb; exon 3 of mineralocorticoid receptor, Mlr; exon 35 and intron 35 of myosin heavy polypeptide 6α, Myh6; exon 3 of engrailed 2, En2; intron 7 of B-fibrinogen, Bfib; and intron 2 of alcohol dehydrogenase, Adh1-I2. An N/A indicates that no sequence data were available. Abbreviations for museum catalog numbers follow Hafner et al. (1997) are: University of Kansas, Natural History Museum and Biodiversity Research Center (KU), Louisiana State University Museum of Natural Science (LSUMZ), Brigham Young University, Monte L. Bean Life Science Museum (BYU), University of New Mexico, Museum of Southwestern Biology (MSB), Museum of Texas Tech University (TTU), University of California-Berkeley, Museum of Vertebrate Zoology (MVZ), and University of California, Los Angeles, Dickey Collection (UCLA). A TK number (reference number used by the TTU Robert J. Baker Genetic Resources Collection) was used when museum catalog numbers were unavailable.

	Catalog #	12S	16S	CoII	Cytb	Mlr	Myh6	En2	Bfib	Adh1-I2
H. alleni	TK45042	DQ179660	DQ179710	DQ179760	AF186801	DQ179860	DQ179910	DQ179960	DQ180010	AY817627
X. nelsoni	TTU37790	DQ179663	DQ179713	DQ179763	AF307838	DQ179863	DQ179913	DQ179963	DQ180013	AY817628
N. albigula	TTU78451	DQ179666	DQ179716	DQ179766	AF376477	DQ179866	DQ179916	DQ179966	DQ180016	AY817648
N. albigula	MSB60812	DQ179708	DQ179758	DQ179808	AF186804	DQ179908	DQ179958	DQ180008	DQ180058	AY817651
N. albigula	MSB77371	DQ179707	DQ179757	DQ179807	AF186808	DQ179907	DQ179957	DQ180007	DQ180057	AY817650
N. albigula	TTU76474	DQ179667	DQ179717	DQ179767	AF186803	DQ179867	DQ179917	DQ179967	DQ180017	AY817649
N. angustapalata	BYU27733	N/A	N/A	N/A	HM989966	N/A	N/A	N/A	N/A	N/A
N. angustapalata	KU37062	N/A	N/A	N/A	HM989965	N/A	N/A	N/A	N/A	N/A
N. bryanti	TTU79131	DQ179680	DQ179730	DQ179780	AF307835	DQ179880	DQ179930	DQ179980	DQ180030	AY817633
N. bryanti	TTU79134	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	AY817634
N. bryanti	MVZ159790	DQ179685	DQ179735	DQ179785	DQ179835	DQ179885	DQ179935	DQ179985	DQ180035	N/A
N. cinerea	MSB74610	DQ179705	DQ179755	DQ179805	AF186800	DQ179905	DQ179955	DQ180005	DQ180055	AY817636
N. cinerea	MSB121427	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	AY817635
N. cinerea	BYU17790	DQ179709	DQ179759	DQ179809	DQ179859	DQ179909	DQ179959	DQ180009	DQ180059	N/A
N. devia	MVZ202447	DQ179682	DQ179732	DQ179782	DQ179832	DQ179882	DQ179932	DQ179982	DQ180032	MZ064564
N. devia	MVZ197094	DQ179686	DQ179736	DQ179786	DQ179836	DQ179886	DQ179936	DQ179986	DQ180036	N/A
N. ferruginea	TTU82666	DQ179679	DQ179729	DQ179779	AF305567	DQ179879	DQ179929	DQ179979	DQ180029	AY817631
N. ferruginea	TTU82665	DQ179678	DQ179728	DQ179778	AF329079	DQ179878	DQ179928	DQ179978	DQ180028	AY817630
N. floridana	TTU75413	DQ179669	DQ179719	DQ179769	AF186819	DQ179869	DQ179919	DQ179969	DQ180019	AY817638
N. floridana	TK27751	DQ179670	DQ179720	DQ179770	AF294341	DQ179870	DQ179920	DQ179970	DQ180020	AY817639
N. floridana	TK28244	DQ179671	DQ179721	DQ179771	AF186818	DQ179871	DQ179921	DQ179971	DQ180021	AY817640
N. floridana	MSB74955	DQ179704	DQ179754	DQ179804	AF294335	DQ179904	DQ179954	DQ180004	DQ180054	AY817637
N. fuscipes	MVZ196405	DQ179672	DQ179722	DQ179772	DQ179822	DQ179872	DQ179922	DQ179972	DQ180022	MZ064565
N. fuscipes	MVZ196371	DQ179675	DQ179725	DQ179775	DQ179825	DQ179875	DQ179925	DQ179975	DQ180025	MZ064566
N. goldmani	TTU45227	DQ179677	DQ179727	DQ179777	AF186829	DQ179877	DQ179927	DQ179977	DQ180027	AY817656
N. insularis	UCLA19911	N/A	N/A	N/A	DQ781161	N/A	N/A	N/A	N/A	N/A
N. lepida	MSB77775	DQ179703	DQ179753	DQ179803	AF307833	DQ179903	DQ179953	DQ180003	DQ180053	N/A
N. lepida	MVZ197126	DQ179688	DQ179738	DQ179788	DQ179838	DQ179888	DQ179938	DQ179988	DQ180038	MZ064567
N. leucodon	TTU75440	DQ179689	DQ179739	DQ179789	AF186809	DQ179889	DQ179939	DQ179989	DQ180039	AY817644
N. leucodon	TTU71198	DQ179665	DQ179715	DQ179765	AF186806	DQ179865	DQ179915	DQ179965	DQ180015	AY817643
N. macrotis	TTU81391	DQ179694	DQ179744	DQ179794	AF376479	DQ179894	DQ179944	DQ179994	DQ180044	AY817632
N. macrotis	TTU79134	DQ179693	DQ179743	DQ179793	AF307836	DQ179893	DQ179943	DQ179993	DQ180043	N/A
N. magister	MSB74952	DQ179706	DQ179756	DQ179806	AF294336	DQ179906	DQ179956	DQ180006	DQ180056	AY817641
N. melanura	TTU110072	N/A	N/A	N/A	KM488358	N/A	MZ147829	N/A	N/A	N/A
N. melanura	MVZ147661	MZ027647	MZ020779	MZ064558	AF337750	MZ064571	N/A	MZ064569	MZ064573	MZ064568
N. mexicana	TTU79129	DQ179697	DQ179747	DQ179797	AF294345	DQ179897	DQ179947	DQ179997	DQ180047	AY817646
N. mexicana	TK51346	DQ179696	DQ179746	DQ179796	AF186821	DQ179896	DQ179946	DQ179996	DQ180046	AY817647
N. micropus	TK31643	DQ179698	DQ179748	DQ179798	AF186822	DQ179898	DQ179948	DQ179998	DQ180048	AY817652
N. micropus	TTU80856	DQ179690	DQ179740	DQ179790	AF186827	DQ179890	DQ179940	DQ179990	DQ180040	AY817655
N. nelsoni	LSUMZ36663	KC758853	KC758852	N/A	KC758854	N/A	N/A	N/A	N/A	N/A
N. phenax	TTU6947	N/A	MZ020780 N/A	MK202784	MZ064572	N/A	MZ064570	N/A	N/A
N. picta	TK93390	DQ179701	DQ179751	DQ179801	AF305569	DQ179901	DQ179951	DQ180001	DQ180051	AY817629
N. stephensi	TTU78505	DQ179702	DQ179752	DQ179802	AF308867	DQ179902	DQ179952	DQ180002	DQ180052	AY817642
O. phyllotis	TTU84371	DQ179664	DQ179714	DQ179764	DQ179814	DQ179864	DQ179914	DQ179964	DQ180014	AY817624
T. nudicaudus	TTU62082	DQ179662	DQ179712	DQ179762	N/A	DQ179862	DQ179912	DQ179962	DQ180012	AY817625
T. nudicaudus	TTU77530	N/A	N/A	N/A	AF307839	N/A	N/A	N/A	N/A	N/A

To expand taxonomic coverage, skin clips were obtained from N. phenax. Small pieces of skin, approximately 2 mm × 2 mm in size, were obtained from the ventral suture of museum voucher specimens. Skin clips were rinsed, and DNA extracted following protocols outlined in Fulton et al. (2012) and Campos and Gilbert (2012). PCRs for skin clips used Phire II Hot Start DNA polymerase (Finnzymes Thermo Scientific, Rockford, Illinois), and the thermal profiles listed in Table 1. PCR products were purified, sequenced, aligned, and proofed as above. All DNA extractions and PCR methods were performed in a separate laboratory from the general PCR laboratory to minimize the risk of contamination; standard positive and negative controls were included at all stages to further reduce the possibility of contamination.

Phylogenetic and genetic divergence analyses

DNA sequence data were analyzed using four approaches. First, the 691 Cytb sequences were used to construct a robust phylogeny that could be used to (i) establish a global view of nucleotide diversity within Neotoma and its allies, (ii) denote genetically divergent taxa and delineate species boundaries, and (iii) based on the rapid genetic divergence associated with this gene, serve as the primary data set for identifying nodal support for more terminal taxa. Second, each of the nine individual genes (four mitochondrial loci—12S, 16S, CoII, and Cytb; five nuclear loci—AdhI2, BfibI7, En2, Mlr, and Myh6) were analyzed independently so that the contribution of each gene could be visualized relative to the overall phylogeny. Third, the DNA sequences from the nine genes were concatenated into a single data set and analyzed jointly, with each gene analyzed under its own priors. Fourth, genetic distances were estimated for DNA sequences obtained from the Cytb gene and used to examine the magnitude of genetic divergence between taxa.

In two of the analytical approaches (nine individual genes and concatenated), sequence data were analyzed using Bayesian inference (BI; MrBayes v3.2.6; Huelsenbeck and Ronquist 2001; Ronquist et al. 2012), maximum likelihood (ML; RAxML Version 8.2.12, Stamatakis 2014), and parsimony methods (PARS, PAUP* Version 4.0a167; Swofford 2002); only the BI and ML methods were used to analyze the Cytb data set. In all analyses, Ototylomys phyllotis and Tylomys nudicaudus were used as outgroups to set character polarity.

For the BI and ML analyses, 88 likelihood models were evaluated using jModelTest-2.1.10 (Guindon and Gascuel 2003; Darriba et al. 2012) and the Akaike information criterion with a correction for finite sample sizes (AICc; Hurvich and Tsai 1989; Burnham and Anderson 2004) was used to identify the most appropriate model of evolution for each data set. The “best-fit models” predicted for each gene are shown (Table 3); however, the GTR model of evolution (and +I +G when appropriate) was used because RAxML only uses the GTR model. The most appropriate nucleotide substitution model and the following parameters were used: two independent runs with four Markov chains (one cold and three heated; MCMCMC), 10⁶ generations, and sample frequency of every 1,000 generations from the last 9 million generated. A visual inspection of likelihood scores resulted in the first 2,000,000 trees being discarded (10% burn-in) and a consensus tree (50% majority rule) constructed from the remaining trees. Clade probability values (CPV) ≥ 0.95 were used to designate nodal support (Huelsenbeck et al. 2002) in BI analyses and bootstrap values (BS; ≥ 70) based on 1,000 iterations (Felsenstein 1985) were used to designate nodal support in the ML analyses.

Table 3.

Summary of best-fit evolutionary models as determined by the programs jModelTest-2.1.10 (Guindon and Gascuel 2003; Darriba et al. 2012) and the Akaike information criterion with a correction for finite sample sizes (AICc; Hurvich and Tsai 1989; Burnham and Anderson 2004). Columns depict: locus name for each gene examined; best-fit evolutionary model; and appropriate GTR model used in the program RAxML for maximum likelihood analysis.

Locus	Best-fit evolutionary model	RAxML model
12S	GTR+I+G	GTR+I+G
16S	TIM2+I+G	GTR+I+G
Adh	HKY+G	GTR+G
Bfib	TIM2+G	GTR+G
CoII	TIM2+I+G	GTR+I+G
Cytb	GTR+I+G	GTR+I+G
En2	HKY	GTR
Mlr	HKY+G	GTR+G
Myh6	TrN+I+G	GTR+I+G

In the parsimony analysis (PAUP* Version 4.0a167; Swofford 2002), characters were assigned equal weight and variable nucleotide positions were treated as unordered, discrete characters with four possible states; A, C, G, and T. Phylogenetically uninformative characters were removed from the analysis. The most parsimonious trees were estimated using the heuristic search and tree bisection–reconnection option. A strict consensus tree from the resulting pool of most parsimonious trees and bootstrap values (BS; ≥ 70) based on 1,000 iterations (Felsenstein 1985) were used to designate nodal support.

The Kimura 2-parameter model of evolution (Kimura 1980) was used to estimate genetic distances (Table 4). Comparisons were based on sequences generated herein, as well as sequences obtained from other studies (Edwards et al. 2001) and Edwards and Bradley (2002b), Ordóñez-Garza et al. (2014), and GenBank. These values were then used to assess levels of genetic divergence among species of Neotoma following the criteria for the Genetic Species Concept as outlined in Bradley and Baker (2001) and Baker and Bradley (2006).

Table 4.

Genetic distances, estimated from mitochondrial cytochrome-b sequences using the Kimura 2-parameter model (Kimura 1980), for sister taxa and closely related (supported) species. Within-species distance values are in parentheses.

Pairwise comparison	Genetic distance
N. albigula (2.57%) vs. N. melanura (2.19%)	7.51%
N. albigula (2.57%) vs. N. floridana (2.26%)	9.98%
N. albigula (2.57%) vs. N. magister (0.88%)	11.61%
N. albigula (2.57%) vs. N. goldmani (1.06%)	10.73%
N. floridana (2.26%) vs. N. magister (0.88%)	8.01%
N. floridana (2.26%) vs. N. melanura (2.19%)	9.61%
N. floridana (2.26%) vs. N. goldmani (1.06%)	11.06%
N. magister (0.88%) vs. N. melanura (2.19%)	10.38%
N. magister (0.88%) vs. N. goldmani (1.06%)	12.57%
N. melanura (0.88%) vs. N. goldmani (1.06%)	10.14%
N. leucodon (4.20%) vs. N. nelsoni (NA)	13.00%
N. leucodon (4.20%) vs. N. angustapalata (NA)	3.78%
N. leucodon (4.20%) vs. N. micropus (0.83%)	9.87%
N. nelsoni (NA) vs. N. angustapalata (NA)	12.70%
N. nelsoni (NA) vs. N. micropus (0.83%)	16.89%
N. angustapalata (NA) vs. N. micropus (0.83%)	9.42%
N. ferruginea (2.21%) vs. N. picta (1.04%)	7.84%
N. ferruginea (2.21%) vs. N. mexicana (2.98%)	9.52%
N. picta (1.04%) vs. N. mexicana (2.98%)	9.86%
N. bryanti (2.88%) vs. N. devia (2.11%)	8.34%
N. bryanti (2.88%) vs. N. lepida (2.71%)	8.70%
N. bryanti (2.88%) vs. N. insularis (NA)	6.28%
N. devia (2.11%) vs. N. lepida (2.71%)	4.68%
N. devia (2.11%) vs. N. insularis (NA)	8.31%
N. lepida (2.71%) vs. N. insularis (NA)	8.68%
N. fuscipes (6.82%) vs. N. macrotis (2.74%)	8.52%

Results

A total of 691 DNA sequences, obtained from the mitochondrial Cytb gene, were examined under ML and BI frameworks in order to ascertain the phylogenetic relationships, delineate species boundaries, and identify genetically diverse taxa (Fig. 1) between and within the 21 species of Neotoma included in this study. For all species, a monophyletic arrangement was recovered with the exceptions of a clade containing samples of N. leucodon, N. angustapalata, and N. nelsoni. In most cases, sister species and other closely related taxa exhibited substantial nodal support (CPV ≥ 0.95 and BS ≥ 70) at the terminal positions of the topology; however, little support was recovered for clades located at middle-level positions (Fig. 1) and no support was recovered for clades located at basal positions.

Fig. 1.

Phylogenetic tree generated using Bayesian methods (MrBayes; Huelsenbeck and Ronquist 2001), the GTR+I+G model of evolution, and 691 DNA sequences obtained from the mitochondrial Cytb gene for the 23 species of Neotoma and allies examined in this study. Clade probability values (≥ 0.95) are indicated by an asterisk and are depicted above branches and bootstrap support values (≥ 70) obtained from the maximum likelihood analysis (RaxML Version 8.2.12, Stamatakis 2014) of the same data set are depicted below the branches.

Results of the PARS, ML, and BI analyses of the nine individual genes (four mitochondrial loci: 12S, 16S, CoII, and Cytb; five nuclear loci: AdhI2, BfibI7, En2, Mlr, and Myh6) depicted a wide array of phylogenetic relationships among taxa. A synthesis of each analysis is provided in Fig. 2. Not only was substantial nodal support recovered in both the ML and PARS analyses at the mid-level and terminal position; several individual genes, particularly A, C–E, and I, provided support at the basal positions; however, phylogenetic resolution was limited in most genes.

Fig. 2.

Phylogenetic tree generated using Bayesian methods (MrBayes; Huelsenbeck and Ronquist 2001), the GTR+I+G model of evolution, and a single data set represented by concatenated DNA sequences obtained from nine individual genes (four mitochondrial loci: 12S, 16S, CoII, and Cytb - denoted by A, B, E, and I, respectively; five nuclear loci: AdhI2, BfibI7, En2, Mlr, and Myh6 - denoted by C, D, F, G, and H, respectively) for the 23 species of Neotoma examined in this study. Clade probability values (≥ 0.95) are placed above branches and depicted by an asterisk (*) to the left of the slash (/). Results of the maximum likelihood (ML; RaxML Version 8.2.12, Stamatakis 2014) and parsimony analysis (PARS, PAUP* Version 4.0a167; Swofford 2002) of individual genes (A–I) were superimposed on the Bayesian inference topology. Bootstrap support values (≥ 70) obtained from the ML analysis of individual genes (A–I) were placed above branches and to the right of the slash (/). Bootstrap support values (≥ 70) obtained from the PARS analysis of individual genes (A–I) were placed below branches.

The final BI analysis was based on concatenation of the nine genes into a single data set (Fig. 2). Several supported clades (CPV ≥ 0.95) were evident and are shown by an asterisk (*) above the node and to the left of the slash (/). Most of the supported clades corresponded to previously recognized species groups (floridana, micropus, mexicana, and lepida) and the subgenus Homodontomys. In general, CPV obtained from the concatenated analysis depicted support at the terminal and basal positions of the topology; however, little support was recovered for clades located at middle-level positions (i.e., species group relationships).

Genetic distances (Kimura 2-parameter; Kimura 1980) were estimated for DNA sequences obtained from the Cytb gene (Table 4) and used to examine the magnitude of genetic divergence between sister taxa and selected members of closely related clades. Values for within-species variation ranged from 0.88% (N. magister) to 6.82% (N. fuscipes). Between sister species values ranged from 4.68% (N. devia and N. lepida) to 12.70% (N. angustapalata and N. nelsoni); whereas members within closely related clades ranged from 8.70% (N. bryanti and N. lepida) to 12.57% (N. goldmani and N. magister).

Discussion

The first goal of this study was to use DNA sequence data from the Cytb gene (obtained from 691 individuals) to determine the magnitude of genetic variation within and between species and to develop a molecular phylogeny that could serve as a test of monophyly for taxonomic boundaries (Fig. 1). Within-species variation was relatively low compared to other species of rodents with most values < 3% and ranging from 0.88% for individuals of N. magister to 6.82% for individuals of N. fuscipes (Table 4). Given that sister species values obtained herein ranged from 4.68% (N. devia and N. lepida) to 12.70% (N. angustapalata and N. nelsoni) only N. magister seemed to approach those observed for other sister species of woodrats (see Hayes and Harrison 1992; Hayes and Richmond 1993; Edwards and Bradley 2001 for a synopsis of the relationship of N. magister to N. floridana). The intraspecific genetic variation identified within N. fuscipes (6.82%) primarily was due to the two genetically divergent groups (8.81% divergent at the Cytb locus) of N. fuscipes first noted by Matocq (2002). These groups exhibit divergence on the higher end of the between-species average for rodents (see Bradley and Baker 2001; Baker and Bradley 2006). Matocq (2002) referred to these groups as the “northern” and “west central” groups and estimated they may have diverged 1.8 mya. Recent genome-wide, single-nucleotide polymorphism data coupled with demographic modeling support this estimate of the timing and magnitude of divergence between the major lineages of N. fuscipes (Boria et al. 2021), and work is ongoing to reevaluate morphological variation across the group.

The second goal of this study was to determine if the phylogenetic relationships obtained from the expanded multigene data set added resolution to previous topologies generated for Neotoma and its allies. The topology obtained from the phylogenetic analysis of the 691 Cytb sequences (Fig. 1) was similar to that obtained in the concatenated analyses (Fig. 2, ML and BI), especially for terminal and mid-level clades. In no cases were clades based solely on Cytb sequences incongruent with supported clades (CPV ≥ 0.90; BS ≥ 70) obtained from the concatenated analyses, although the concatenated analyses provided superior support for basal and some mid-level arrangements. In addition, topologies and support values obtained from the ML, BI, and PARS were similar across most nodes.

Although the three molecular-based studies by Edwards and Bradley (2002b), Longhofer and Bradley (2006), and Matocq et al. (2007) supported similar interpretations of the phylogenetic relationships of Neotoma and its allies; Matocq et al. (2007) was the most inclusive in terms of taxonomic and character coverage. Consequently, the topology presented in Matocq et al. (2007) served as the primary comparison and point of discussion for the topology obtained from the concatenated analyses (Fig. 2). Given that the current study contained additional taxa not examined by Matocq et al. (2007), there were instances where the topologies could not be compared directly. To simplify the comparison of topologies (current study and Matocq et al. 2007) and to initiate a discussion of phylogenetic relationships that was used as the basis of developing a classification (Table 5), clades were described in relation to species-group assignment (Fig. 2) following that presented by Edwards and Bradley (2002b), Matocq et al. (2007), and modified herein.

Table 5.

Classification of Neotoma and allies (following Edwards and Bradley 2002; Musser and Carleton 2005; Matocq et al 2007). Neotoma anthonyi, N. bunkeri, and N. martinensis are included as either subspecies of, or synonyms of, N. bryanti following Patton et al. (2007). Neotoma palatina and N. chrysomelas were not examined in this study; however, they were placed into the classification based on the most current synopsis of the literature. Letter representations are as follows: X = no sequences available for examination, did not evaluate, or taxon was not recognized at time of study; A = agreement of previous taxonomy with that proposed herein; D = disagreement of previous taxonomy with proposed herein; Y = agreement of previous classifications (Goldman 1910; Hall 1981; Edwards and Bradley 2002; Musser and Carleton 2005) with that proposed herein; N = disagreement of previous classification with proposed herein; and NSG = no species-group assignment. Explanations relative to disagreements in taxonomy or classification between previous and current studies are placed in parentheses.

Proposed classification	Goldman (1910)	Hall (1981)	Edwards and Bradley (2002)	Musser and Carleton (2005)
Genus Hodomys
H. alleni Merriam, 1894	A, Y	D (N. alleni), N (subgenus Hodomys)	A, Y	A, Y
Genus Xenomys
X. nelsoni Merriam, 1892	A, Y	A, Y	A, Y	A, Y
Genus Neotoma	A, N (subgenus Neotoma)	A, N (subgenus Neotoma)	A, N (subgenus Neotoma)	A, N (subgenus Neotoma)
Species group floridana
N. albigula Hartley, 1894	A, N (albigula group)	A, NSG	A, Y	A, Y
N. floridana (Ord, 1818)	A, Y	A, NSG	A, Y	A, Y
N. magister Baird, 1858	A, N (pennsylvanica group)	D (N. fl. magister), NSG	A, Y	A, Y
N. melanura Merriam, 1894	D (N. a. melanura), N	D (N. a. melanura), NSG	X	D (N. a. melanura), Y (floridana group)
N. goldmani Merriam, 1903	A, N (desertorum group)	A, NSG	A, Y	A, Y
Species group lepida
N. bryanti Merriam, 1887	A, N (intermedia group)	A, NSG	X, X	A, Y
N. devia Goldman, 1927	X, X	D (N. l. devia), NSG	A, Y	A, Y
N. insularis Townsend, 1912	X, X	D (N. l. insularis), NSG	X, X	D (N. l. insularis), Y
N. lepida Thomas, 1893	A, N (desertorum group)	A, NSG	A, Y	A, Y
Species group mexicana
N. chrysomelas J. A. Allen, 1908	A, Y	A, NSG	X, X	A, Y
N. ferruginea Tomes, 1862	D (N. fe. ferruginea), Y	D (N. fe. ferruginea), NSG	X, X	D (N. me. ferruginea), Y
N. mexicana Baird, 1855	A, Y	A, NSG	A, Y	A, Y
N. picta Goldman, 1904	D (N. fe. picta), Y	D (N. fe. picta), NSG	A, Y	D (N. me. picta), Y
Species group micropus
N. angustapalata Baker, 1951	X, X	A, NSG	X, X	A, N (mexicana group)
N. leucodon Merriam, 1894	D (N. a. leucodon), N (albigula group)	D (N. a. leucodon), NSG	A, Y	A, Y
N. micropus Baird, 1855	A, Y	A, NSG	A, Y	A, Y
N. nelsoni Goldman, 1905	A, N (albigula group)	A, NSG	X, X	A, Y
N. palatina Goldman, 1905	A, N (albigula group)	A, NSG	X, X	A, Y
Species group stephensi
N. stephensi Goldman, 1905	D (N. l. stephensi), N (desertorum group)	A, NSG	A, N (lepida group)	A, N (lepida group)
Species group fuscipes
N. fuscipes Baird, 1858	A, N (Homodontomys)	A, NSG	A, N (lepida group)	A, N (lepida group)
N. macrotis Thomas, 1893	D (N. fu. macrotis), N (fuscipes group)	D, (N. fu. macrotis), NSG	X, X	A, N (lepida group)
Species group phenax
N. phenax (Merriam, 1903)	D (T. phenax), N (genus Teanopus)	A, N (subgenus Teanopus)	X, X	A, N (subgenus Teanopus)
Species group Teonoma
N. cinerea (Ord, 1815)	A, N (subgenus Teanoma)	A, N (subgenus Teanoma)	A, NSG	A, N (subgenus Teanoma)

A sister relationship between Xenomys and Hodomys was recovered (to the exclusion of a monophyletic Neotoma—remaining 21 species examined); this relationship agreed with that discussed by Matocq et al. (2007). This finding supports: (i) the interpretation that Hodomys should be recognized as a separate genus (Goldman 1910; Carleton 1980; Edwards and Bradley 2002b; Musser and Carleton 2005) and not as a subgenus within Neotoma (Burt and Barkalow 1942; Hall 1981) and (ii) Neotoma is monophyletic (Edwards and Bradley 2002b; Musser and Carleton 2005).

Four well-supported clades (Fig. 2) were recovered that corresponded to the floridana (albigula, floridana, goldmani, magister, and melanura), lepida (bryanti, devia, insularis, and lepida), mexicana (ferrunginae, mexicana, and picta), and micropus (angustapalata, leucodon, micropus, and nelsoni) species groups as described in Edwards and Bradley (2002b) and Matocq et al. (2007). These species groups were identified based on support as defined by BS (> 70) and CPV (> 0.95) and are indicated by letters A–I (corresponding to the nine mitochondrial and nuclear markers, respectively) and superimposed on Fig. 2, where appropriate.

Two species (phenax and stephensi) were not associated within any species group because of lack of nodal support values (Fig. 2). However, both species were included within the traditional subgenus Neotoma (Fig. 2) as defined by data from AdhI2 and CoII markers. Three additional species (macrotis, fuscipes, and cinerea) formed a clade based on data from the AdhI2 gene. Within this clade, macrotis and fuscipes were sister taxa as defined by the 12S, AdhI2, BfibI7, CoII, Mlr, and Cytb genes; the taxonomic ramifications of this clade are discussed in the classification section.

The four mitochondrial genes (12S, 16S, CoII, and Cytb) provided support at 7, 11, 10, and 23 nodes, respectively, whereas the five nuclear markers (AdhI2, BfibI7, En2, Mlr, and Mlr) provided support at 18, 10, 5, 3, and 2 nodes, respectively. The distribution and contribution of nodal support for each marker is provided (Fig. 2).

The final goal of this study was to produce a classification for Neotoma and its allies based on DNA sequences obtained from single-gene analyses and from analyses of nine concatenated mitochondrial and nuclear genes. To accomplish this task, monophyletic groups obtained from analyses were evaluated for: (i) levels of support (BS > 70 and CPV > 0.95) and (ii) congruence to previous classifications and taxonomic arrangements. For this exercise, we used the arrangements presented in Goldman (1910), Hall (1981), Edwards and Bradley (2002b), and Musser and Carleton (2005) knowing that these studies differed in the types of data sets used and number of taxa recognized. In several cases, we drew upon information presented in more narrowly focused taxonomic studies (Patton et al. 2007; Rogers et al. 2011; Fernández 2014; Ordóñez-Garza et al. 2014; Bradley and Mauldin 2016). The proposed classification (Table 5) and below we provide a synopsis highlighting the differences and similarities between the five classification schemes.

Recognition of genera

Essentially all analyses (single or concatenated) depicted a sister relationship between Xenomys and Hodomys and all remaining ingroup taxa contained within a monophyletic Neotoma. The recognition of Xenomys as a separate genus is supported by all other classifications discussed herein. Support for the recognition of Hodomys as a separate genus is supported in all other classifications except that of Hall (1981) where Hodomys was treated as a subgenus of Neotoma and, by extension H. alleni was assigned to N. alleni. Recognition of a traditional Neotoma assemblage (Musser and Carleton 2005) was supported by the concatenated data set and five single-gene data sets (Fig. 2). Recognition of Teanopus (T. phenax, Goldman 1910) was not supported by the sequence data analyzed herein because N. phenax was included in a clade containing 13 other species of Neotoma (although this support was weak; Fig. 2). To recognize Teanopus as a genus would require a major reorganization of the genus Neotoma and given the limited data we were able to assemble for N. phenax, any restructuring at that level would be premature.

Recognition of subgenera

Historically, five subgenera (Hodomys, Homodontomys, Neotoma, Teanopus, and Teonoma) have been recognized (see Edwards and Bradley 2002b; Matocq et al. 2007; Table 5 for a discussion of the historical perspectives). In short, the subgenus Neotoma contained all recognized species except: alleni which previously was assigned to subgenus Hodomys; macrotis and fuscipes were assigned to subgenus Homodontomys; cinerea assigned to subgenus Teonoma; and phenax was assigned to subgenus Teanopus. Recognition of subgenera historically was based on sound morphological interpretations; and one could argue for the use of subgenera based on those data; however, given the lack of statistical support provided by the phylogenetic analysis of the DNA sequence data presented herein, there appears to be no justification for the continued recognition of subgenera. First, although Hodomys represents a monophyletic assemblage, it was statistically supported as sister to Xenomys and outside of Neotoma; consequently, Hodomys was so different from other species of Neotoma that we follow Goldman (1910), Edwards and Bradley (2002b), Musser and Carleton (2005), and Matocq et al. (2007) in considering it distinct at the generic level.

Second, recognition of the remaining four subgenera (Homodontomys, Neotoma, Teanopus, and Teonoma) would require several major reorganizations of the remaining 22 species. For example: (A) the subgenus Neotoma is not monophyletic because bryanti, devia, insularis, and lepida are sister to a clade containing cinerea (subgenus Teonoma), fuscipes and macrotis (subgenus Homodontomys) resulting in a need to reorganize the subgenus Neotoma by splitting it into at least two clades (one poorly supported clade representing 14 species formerly comprising the subgenus Neotoma and the second being a strongly supported but unnamed subgenus containing (bryanti, devia, insularis, and lepida)); (B) Teonoma and Homodontomys are each monophyletic and represent distinct clades that could be recognized at the subgeneric level, but doing so would require a reevaluation of many species groups within the former subgenus Neotoma which appears to have evolved prior to the divergence of the Teonoma and Homodontomys clades; (C) alternatively, given that Teonoma and Homodontomys are sister groups they could be combined into a single subgenus, with Teonoma having taxonomic priority; and (D) similarly, to recognize Teanopus as a subgenus would require a major reorganization of the subgenus Neotoma and as discussed above, is premature. Based on the data presented herein, it seems prudent to abandon the use of subgenera.

Recognition of species groups

Edwards and Bradley (2002b), Musser and Carleton (2005), Matocq et al. (2007), and Patton et al. (2007) placed most Neotoma species into four basic species groups (Table 5; floridana, lepida, mexicana, and micropus). Herein, we use information obtained from the inclusion of additional sequence data and increased taxon sampling to expand on these earlier premises, by further refining the species groups; as well as suggesting the formation of four new species groups. This refinement necessitates five taxonomic interpretations: (i) the lepida species group should be restricted to bryanti, devia, insularis, and lepida (recently extinct taxa from the Baja California region could be added following the synopsis provided by Patton et al. 2007); (ii) stephensi should be removed from the lepida species group as tentatively suggested by Edwards and Bradley (2002b) and recognized as its own species group (stephensi species group); (iii) given its level of genetic and morphologic distinction (summarized by Matocq et al. 2007) cinerea should be recognized as its own species group (cinerea species group) unless it is combined into a larger species group with fuscipes and macrotis with cinerea species group having priority; (iv) fuscipes and macrotis should be recognized in a separate species group (fuscipes species group) unless it is combined into a larger species group with cinerea as mentioned above; and (v) although little DNA sequence data were available for analyses, it appears that phenax should be recognized as its own species group (phenax species group) unless it is combined into a larger species group with members of the mexicana species group with the mexicana species group having priority. Conservatively, this interpretation results in the recognition of eight species groups (cinerea, floridana, fuscipes, lepida, mexicana, micropus, phenax, and stephensi).

At this juncture, the use of additional species groups to partition and define monophyletic groupings over the historical implementation of subgenera and a limited number of species groups seems to be a reasonable and superior approach given that: (i) six of the eight species (as defined herein) are monophyletic and are defined by statistical support; only phenax and stephensi received no support as species groups but given their exclusion from any species group predicates their independent evolutionary trajectory compared with other species groups that were defined by statistical support, and (ii) this simplifies the controversies in trying to invoke the subgenera concepts onto the phylogenetic results.

Recognition of species

Description of woodrat species has taken place over two major time intervals, loosely defined here as the morphology/natural history era (circa pre-1970) and the molecular data era (circa post-1970). Our intent is not to justify one method over the other but simply to reflect the efforts of alpha- and beta-taxonomists as different research methods were employed; however, it is interesting that although several subspecific names have been elevated to species rank, no novel species name has been proposed since N. angustapalata (Baker 1951) and no new subspecies name has been proposed since N. albigula subsolana (Alvarez 1962; now N. leucodon subsolana - Edwards et al. 2001). A detailed list of species names and the authorities is provided in Table 5.

For species described prior to the utilization of molecular data in taxonomic decision-making, data presented herein unequivocally support the initial species recognition of Hartley (albigula, 1894), Merriam (bryanti, 1887), Ord (cinerea, 1818), J. A. Allen (chrysomelas, 1908), Ord (floridana,1815), Baird (fuscipes, 1858), Merriam (goldmani, 1903), Thomas (lepida, 1893), Baird (mexicana and micropus, 1855), Goldman (palatina, 1905), Merriam (phenax, 1903), and Goldman (stephensi, 1905). Recognition of two additional species (angustapalata - Baker 1951 and nelsoniGoldman 1905), described during the morphology/natural history era, is less certain; as these taxa were either embedded within the N. leucodon clade (Cytb data set) or were poorly supported as distinct taxa (concatenated data set). The only Cytb sequence available to this study (GenBank accession number KC758854) differs from other Neotoma sequences near the 3′ end. Elimination of the last ~230 bp of the sequence reduces the genetic divergence estimate between angustapalata and nelsoni from 12.7% to 7.6% and between angustapalata and leucodon from 12.7% to 8.8%, respectively; indicating a closer genetic relationship than originally implied. However, until additional samples of these two taxa can be examined, we take a conservative approach and retain both as species—agreeing with the findings of Rogers et al. (2011) but disagreeing with Fernández (2014). In defense of Fernández (2014), whose position was based on a single specimen, the leucodon clade is more complex than depicted herein and additional studies are needed to determine the taxonomic status of populations in Mexico relative to those in the United States. During the molecular data era, two taxa originally described as subspecies were elevated to species status (devia - Goldman 1927, see Patton et al. 2007 and melanura - Merriam 1894, see Bradley and Mauldin 2016) and six taxa originally described as species but eventually relegated to subspecies, were reelevated to species status (ferruginea - Tomes 1862, see Ordóñez-Garza et al. 2014; insularis - Townsend 1912, see Patton et al. 2007; leucodon - Merriam 1894, see Edwards et al. 2002; macrotis - Thomas 1893, see Matocq 2002; magister - Baird 1858, see Edwards et al. 2001; and picta - Goldman 1904, see Edwards and Bradley 2002b). The elevation of these eight species was strongly supported by the data presented herein.

Epilogue

Clearly, a more complete sequence data set and multiple samples would enhance the analyses and ultimately the interpretation of this taxonomic group. However, the combination of data (sequence and morphometric) and concepts discussed in Edwards and Bradley (2002b), Musser and Carleton (2005), Matocq et al. (2007), Patton et al. (2007), and this study act as a hypothesis for now. Samples of N. chrysomelas and N. palatina are needed to complete the complement of species. In addition, our understanding of phylogenetic relationships within Neotoma will be enhanced by examination of whole genomes; for example, the genomes of N. lepida, N. bryanti, N. fuscipes, and N. macrotis currently are being completed (M.D. Matocq and M.D. Dearing, pers. comm.). Genome-wide data coupled with morphological analyses and field collections targeted to represent the full scope of within- and between-species variation will be needed to further elucidate the evolutionary history of Neotoma and its allies.

Appendix I

Specimens for which cytochrome-b sequences were either generated herein or obtained from GenBank. For each specimen, museum catalog numbers (abbreviations for museum acronyms follow Hafner et al. 1997) are provided to the left of the slash (/) and GenBank accession numbers are provided to the right of the slash. Abbreviations are as follows: Centro de Investigaciones Biológicas del Noroeste (CIB), Cornell University Museum of Vertebrates (CUMV), Denver Museum of Nature and Science (DMNS), Louisiana State University, Museum of Natural Science (LSUMZ), Midwestern State University (MWSU), Monte L. Bean Life Science Museum (BYU), New Mexico Museum of Natural History (NMMNH); Recent Collection of Mammals, Museum of Texas Tech University (TTU), Royal Alberta Museum (RAM), Royal British Columbia Museum (RBCM), Royal Ontario Museum (ROM), United States National Museum of Natural History (USNM), Universidad Autónoma de Morelos (CMC), University of Alaska Museum (UAM), University of California, Berkeley, Museum of Vertebrate Zoology (MVZ), University of California Los Angeles, Dickey Collection (UCLA), University of Colorado, Museum of Natural History (UCM), University of Kansas, Natural History Museum and Biodiversity Research Center (KU), University of New Mexico, Museum of Southwestern Biology (MSB), University of Utah, Natural History Museum of Utah (UMNH), and University of Washington, Thomas Burke Memorial Washington State Museum (UWBM). If museum catalog numbers were unavailable, specimens were referenced with a corresponding collector or other special number (CR, Camp Roberts collection by Marjorie D. Matocq; JLP, James L. Patton; JO, James G. Owen, collector number; MDM, Majorie D. Matocq collector number; MDMNCN, Majorie D. Matocq no collector number; OK, Scott J. Steppan, collector number; OSR, Clinton Epps, collector number; SP, special number of the Carnegie Museum of Natural History; and TK, number special number of the Museum of Texas Tech University).

Ototylomys phyllotis.—TTU84371/AY009789.

Tylomys nudicaudus.—TTU62082/AF307839.

Xenomys nelsoni.—TTU37790/AF307838; TTU28546/KY754179.

Hodomys alleni.—TK45042/AF186801; TK45043/AF186802.

Neotoma albigula.—MSB77708/AF186811; MSB77331/AF186810; NMMNH3795/KM488335; NMMNH4891/KM488336; MSB82999/EU141962; TTU99846/EU141960; TTU99895/EU141964; MSB60812/AF186804; MSB60818/KF267873; TTU78448/AF186816; TTU106657/EU141959; TTU89870/KM488337; TTU97573/KF733991; TTU106915/KM488340; TTU99958/KM488341; TTU97577/KM488339; TTU97812/KF733995; TTU97811/KF733994; TTU97692/KF733990; TTU97776/KF733993; TTU106768/KM488338; TTU106619/KM488347; TTU97139/KF733989; TTU116582/KC250463; TTU76476/AF376472; TTU97148/EU141961; TTU88387/EU141963; TTU97657/KM488342; TTU88153/MZ064559; TTU99878/MZ064560; TTU97156/KF733996; TTU76474/AF186803; MSB77977/AF186807; MSB77371/AF186808; MSB83932/KM488343; MSB41715/AF186814; MSB83827/KM488346; MSB42599/KM488345; NMMNH5325/KM488344; TTU78447/AF186817; TTU78450/AF376476; TTU78451/AF376477; CIB14472/HQ328520; CIB1572/HQ328519; CIB15474/HQ328518; CIB4551/HQ328516; CIB14458/HQ328515; CIB14445/HQ328514; CIB4546/HQ328513; CIB14444/HQ328512; CIB14443/HQ328511; CIB4542/HQ328510; CIB4539/HQ328509; CIB14432/HQ328508; CIB4558/HQ328507; CIB4536/HQ328506; CIB14467/HQ328517.

Neotoma angustapalata.—KU37062/HM989965; BYU27733/HM989966.

Neotoma bryanti.—MSB42979/AF307832; TTU41898/AF376467; TTU83326/KF267875; TTU79131/AF307835; TTU119510/MZ064560; TTU119509/MZ064561; TTU119508/KF267874; MVZ186296/DQ781160; MVZ197380/DQ781135; MVZ195972/KY754056; MVZ186295/DQ781159; MVZ195981/DQ781158; MVZ198589/DQ781157; MVZ198588/DQ781156; MVZ198585/DQ781155; MVZ198584/DQ781154; MVZ195968/DQ781153; MVZ195967/DQ781152; MVZ195963/DQ781151; MVZ195962/DQ781150; MVZ196769/DQ781149; MVZ196098/DQ781148; MVZ196097/DQ781147; MVZ195977/DQ781146; MVZ195976/DQ781145; MVZ198580/DQ781144; MVZ198583/DQ781143; MVZ198582/DQ781142; MVZ196756/DQ781141; MVZ196755/DQ781140; USNM137201/DQ781138; USNM137173/DQ781137; MVZ197381/DQ781136; MVZ197376/DQ781134; MVZ197375/DQ781133; MVZ197175/DQ781132; MVZ197174/DQ781131; MVZ195244/DQ781130; MVZ195243/DQ781129; MVZ196146/DQ781128; MVZ196145/DQ781127; MVZ196144/DQ781126; MVZ196140/DQ781125; MVZ196138/DQ781124; MVZ202544/DQ781123; MVZ202543/DQ781122; MVZ198658/DQ781121; MVZ198657/DQ781120; MVZ196103/DQ781119; MVZ196101/DQ781118; MVZ202539/DQ781117; MVZ199807/DQ781116; MVZ199806/DQ781115; MVZ195325/DQ781114; MVZ195323/DQ781113; MVZ198577/DQ781112; MVZ196120/DQ781111; MVZ196119/DQ781110; MVZ196133/DQ781109; MVZ196132/DQ781108; MVZ198678/DQ781107; CIB7577/DQ781106; CIB7576/DQ781105; CIB7580/DQ781104; CIB7579/DQ781103; CIB7584/DQ781102; CIB7583/DQ781101; CIB2781/DQ781100; CIB2785/DQ781099; UCLA19720/DQ781098; CIB11575/DQ781097; CIB11574/DQ781096; CIB11572/DQ781095; CIB11579/DQ781094; CIB11595/DQ781093; CIB9840/DQ781092; CIB9844/DQ781091; CIB9842/DQ781090; CIB7590/DQ781089; CIB9835/DQ781088; CIB9246/DQ781087; CIB10908/DQ781086; CIB10907/DQ781085; CIB10909/DQ781084; CIB8659/DQ781083; CIB8657/DQ781082; CIB865/DQ781081; CIB8654/DQ781080; CIB7594/DQ781079; CIB2798/DQ781078; CIB7707/DQ781077; CIB7706/DQ781076; CIB7595/DQ781075; CIB5158/DQ781074; CIB7591/DQ781073; CIB7589/DQ781072; CIB8651/DQ781071; CIB7586/DQ781070; CIB8650/DQ781069; CIB7585/DQ781068; CIB3396/DQ781067; CIB2788/DQ781066; CIB7709/DQ781065; CIB7708/DQ781064; USNM139030/DQ781139.

Neotoma cinerea.—MDM148/AF337751; MSB121427/AF186799; MSB121427/AF186799; MSB74610/AF186800; BYU17790/DQ179859; MSB149469/JN593138; RAM04121/JQ241228; UWBM78604/JQ241243; RBCM0207/JN593210; UAM35116/JN593213; MVZ223394/JN593165; MVZ201915/JN593152; MVZ223440/JN593189; MVZ223425/JN593181; MVZ223445/JN593193; MVZ223443/JN593192; MVZ223435/JN593187; MSB140843/JN593136; MVZ207659/KY754057; UWBM76808/JQ241242; UWBM72137/JQ241241; UWBM72108/JQ241240; UWBM72106/JQ241239; UMNH32654/JQ241238; UMNH32653/JQ241237; UMNH32187/JQ241236; UMNH31874/JQ241235; UMNH3187/JQ241234; UAM50132/JQ241233; UAM35118/JQ241232; UAM35117/JQ241231; UAM24567/JQ241230; UAM2456/JQ241229; RAM0183/JQ241227; MVZ223462/JQ241226; MVZ223461/JQ241225; MVZ223460/JQ241224; MVZ223459/JQ241223; MVZ223458/JQ241222; MVZ223455/JQ241221; MVZ223448/JQ241220; MVZ223447/JQ241219; MVZ223446/JQ241218; MVZ223444/JQ241217; MVZ223439/JQ241216; MVZ223437/JQ241215; MVZ223436/JQ241214; MVZ223432/JQ241213; MVZ223429/JQ241212; MVZ223427/JQ241211; MVZ223426/JQ241210; MVZ223421/JQ241209; MVZ223420/JQ241208; MVZ223419/JQ241207; MVZ223418/JQ241206; MVZ223417/JQ241205; MVZ223415/JQ241204; MVZ223412/JQ241203; MVZ223411/JQ241202; MVZ223410/JQ241201; MVZ223409/JQ241200; MVZ223408/JQ241199; MVZ223402/JQ241198; MVZ223400/JQ241197; MVZ223399/JQ241196; MVZ223396/JQ241195; MVZ220747/JQ241194; MVZ218379/JQ241193; MVZ201916/JQ241192; MSB92136/JQ241191; MSB76967/JQ241190; MSB76527/JQ241189; MSB76526/JQ241188; MSB74646/JQ241187; MSB74611/JQ241186; MSB74609/JQ241185; MSB152693/JQ241184; MSB152633/JQ241183; DMNS11517/JQ241182; UWBM79658/JN593237; UWBM79495/JN593236; UWBM78852/JN593235; UWBM78606/JN593234; UWBM78139/JN593233; UWBM78001/JN593232; UWBM76813/JN593231; UWBM76811/JN593230; UWBM76799/JN593229; UWBM76697/JN593228; UWBM75217/JN593227; UWBM73810/JN593226; UWBM49066/JN593225; UMNH32655/JN593224; UMNH32468/JN593223; UMNH32202/JN593222; UMNH31875/JN593221; UMNH31172/JN593220; UMNH29794/JN593219; UCM18902/JN593218; UAM71646/JN593217; UAM54423/JN593216; UAM49980/JN593215; UAM35134/JN593214; UAM35063/JN593212; UAM35061/JN593211; RBCM20000/JN593209; RAM953034/JN593208; RAM03132/JN593207; RAM03131/JN593206; OSU/JN593205; NMMNH614/JN593204; NMMNH1006/JN593203; MVZ223463/JN593202; MVZ223457/JN593201; MVZ223456/JN593200; MVZ223454/JN593199; MVZ223453/JN593198; MVZ223452/JN593197; MVZ223451/JN593196; MVZ223450/JN593195; MVZ223449/JN593194; MVZ223442/JN593191; MVZ223441/JN593190; MVZ223438/JN593188; MVZ223434/JN593186; MVZ223433/JN593185; MVZ223431/JN593184; MVZ223430/JN593183; MVZ223428/JN593182; MVZ223424/JN593180; MVZ223423/JN593179; MVZ223422/JN593178; MVZ223416/JN593177; MVZ223414/JN593176; MVZ223413/JN593175; MVZ223407/JN593174; MVZ223406/JN593173; MVZ223405/JN593172; MVZ223404/JN593171; MVZ223403/JN593170; MVZ223401/JN593169; MVZ223398/JN593168; MVZ223397/JN593167; MVZ223395/JN593166; MVZ222572/JN593164; MVZ220748/JN593163; MVZ220746/JN593162; MVZ220623/JN593161; MVZ220570/JN593160; MVZ220134/JN593159; MVZ219951/JN593158; MVZ219950/JN593157; MVZ218378/JN593156; MVZ216409/JN593155; MVZ215473/JN593154; MVZ206431/JN593153; MVZ199348/JN593151; MVZ197092/JN593150; MSB99135/JN593149; MSB92141/JN593148; MSB92140/JN593147; MSB92137/JN593146; MSB90783/JN593145; MSB86004/JN593144; MSB76968/JN593143; MSB74612/JN593142; MSB73684/JN593141; MSB70026/JN593140; MSB152695/JN593139; MSB141113/JN593137; MDM148/JN593135; LSUMZ452/JN593134; DMNS11518/JN593133; DMNS11383/JN593132; CUMV20269/JN593131; BYU18998/JN593130; BYU18997/JN593129; BYU16618/JN593128; BYU16617/JN593127; BYU14868/JN593126; BYU13888/JN593125; BYU13887/JN593124; BYU13886/JN593123; MDMNCN 182/JN593122; MDMNCN 137/JN593121; DMNS11178/JN593120.

Neotoma devia.—MSB41675/AF307829; MSB41678/AF307830; MVZ200714/DQ781302; MVZ195239/DQ781288; MVZ197123/DQ781283; BYU18948/DQ781290; MVZ197117/KY754058; MVZ200713/DQ781301; MVZ199819/DQ781300; MVZ200712/DQ781299; MVZ200711/DQ781298; MVZ200710/DQ781297; MVZ200709/DQ781296; MVZ202447/DQ781295; MVZ200708/DQ781294; MVZ200707/DQ781293; MVZ200706/DQ781292; MVZ200705/DQ781291; MVZ195240/DQ781289; MVZ195238/DQ781287; MVZ195237/DQ781286; MVZ195236/DQ781285; MVZ199818/DQ781284; MVZ197122/DQ781282; MVZ197121/DQ781281; MVZ197120/DQ781280; MVZ197119/DQ781279; MVZ197118/DQ781278; MVZ197117/DQ781277; MVZ197116/DQ781276; MVZ197115/DQ781275; MVZ197097/DQ781274; MVZ197096/DQ781273; MVZ197095/DQ781272; MVZ197093/DQ781271; BYU18947/DQ781270; MVZ199820/DQ781269; MVZ199384/DQ781268; MVZ199381/DQ781267; MVZ199380/DQ781266; MVZ199378/DQ781265; MVZ197112/DQ781264; MVZ197107/DQ781263;l MVZ197106/DQ781262; MVZ197105/DQ781261; MVZ197104/DQ781260; MVZ197103/DQ781259; MVZ197114/DQ781258; MVZ197113/DQ781257.

Neotoma ferrunginae.—TTU104897/MZ064574; JO9027/KF772873; TTU36179/AF298840; TTU82666/AF305567; TTU82665/AF329079; USNM569657/KF772874; USNM569672/KF772875; USNM569553/KF772876.

Neotoma floridana.—TTU75413/AF186819; TTU71587/AF294343; TTU75402/AF294344; MSB74955/AF294335; TK28244/AF186818; TK27751/AF294341; TTU54731/AF294340; MWSU15766/AF294342; MSB84770/AF294333; TTU54717/AF294339; MSB81532/AF294334; OK107/KY754059.

Neotoma fuscipes.—MVZ195212/DQ781303; MVZ196356/DQ179826; MVZ196405/DQ179822; CRS697/KP129298; CR3356/KP129297; CRS514/KP129296; CRS492/KP129295; CRS374/KP129294; CR758/KP129293; CR3696/KP129292; CR3600/KP129291; CR3560/KP129290; CRS129/KP129289; CRS358/KP129288; CRS573/KP129287; CRS521/KP129286; MDM662/AF337767; MDM495/AF337766; MDM656/AF337765; MDM314/AF337764; MDM764/AF337763; MDM433/AF337762; MDM299/AF337761; MDM356/AF337760; MDM273/AF337759; MDM1824/AF337758; MDM415/AF337757; MDM1912/AF337756; MDM259/AF337755; MDM351/AF337754; JLP17478/AF337753; MDM313/AF337752; MVZ196371/DQ179825; MVZ196394/DQ179824; MVZ196386/DQ179823.

Neotoma goldmani.—TTU45227/AF186829; TTU45228/AF186830.

Neotoma insularis.—UCLA19911/DQ781161.

Neotoma lepida.—JLP16824/AF337749; BYU18153/DQ781256; BYU18300/DQ781254; BYU15035/DQ781234; MSB77775/AF307833; MSB56401/AF307831; TTU137769/KC250464; BYU18154/DQ781255; MVZ199397/DQ781253; MVZ199396/DQ781252; MVZ199395/DQ781251; MVZ199393/DQ781250; MVZ199392/DQ781249; MVZ199391/DQ781248; MVZ199404/DQ781247; MVZ199400/DQ781246; MVZ199399/DQ781245; MVZ199389/DQ781244; MVZ199373/DQ781243; MVZ199370/DQ781242; MVZ199369/DQ781241; MVZ197154/DQ781240; MVZ199376/DQ781239; MVZ197152/DQ781238; MVZ197151/DQ781237; MVZ197148/DQ781236; BYU15034/DQ781233; MVZ197158/DQ781232; MVZ197157/DQ781231; MVZ197156/DQ781230; MVZ197155/DQ781229; MVZ199359/DQ781228; MVZ199358/DQ781227; MVZ199356/DQ781226; MVZ199355/DQ781225; MVZ199354/DQ781224; MVZ199353/DQ781223; MVZ199352/DQ781222; MVZ202485/DQ781221; MVZ197167/DQ781220; MVZ197130/DQ781219; MVZ197165/DQ781218; MVZ197126/DQ781217; MVZ199366/DQ781216; MVZ199364/DQ781215; MVZ199362/DQ781214; MVZ199361/DQ781213; MVZ195261/DQ781212; MVZ195311/DQ781211; MVZ198670/DQ781210; MVZ198914/DQ781209; MVZ195912/DQ781208; MVZ198668/DQ781207; MVZ198662/DQ781206; MVZ195923/DQ781205; MVZ202461/DQ781204; MVZ202459/DQ781203; MVZ195282/DQ781202; MVZ195277/DQ781201; MVZ195266/DQ781200; MVZ197161/DQ781199; MVZ197159/DQ781198; MVZ195934/DQ781197; MVZ198578/DQ781196; MVZ195931/DQ781195; MVZ195930/DQ781194; MVZ195919/DQ781193; MVZ195917/DQ781192; MVZ199803/DQ781191; MVZ199817/DQ781190; MVZ199772/DQ781189; MVZ199816/DQ781188; MVZ199349/DQ781187; MVZ199776/DQ781186; MVZ202511/DQ781185; MVZ195926/DQ781184; MVZ202497/DQ781183; MVZ202495/DQ781182; MVZ192240/DQ781181; MVZ199787/DQ781180; MVZ199786/DQ781179; MVZ199798/DQ781178; MVZ202486/DQ781177; MVZ202540/DQ781176; MVZ198665/DQ781175; MVZ202527/DQ781174; MVZ195319/DQ781173; MVZ202524/DQ781172; MVZ199809/DQ781171; MVZ199808/DQ781170; MVZ199805/DQ781169; MVZ199804/DQ781168; MVZ202502/DQ781167; MVZ199800/DQ781166; MVZ199814/DQ781165; MVZ199813/DQ781164; MVZ199812/DQ781163; MVZ195324/DQ781162; TTU119266/KY754060; MVZ197094/DQ179836; MVZ159790/DQ179835; MVZ195223/DQ179834; MVZ197379/DQ179833; MVZ202447/DQ179832; MVZ197142/DQ179831; TTU79131/DQ179830.

Neotoma leucodon.—TTU71198/AF186806’ TTU77530/AF186828; TTU75440/AF186809; MSB48164/AF186812; TTU44923/AF186805; TTU35381/AF186813; TTU43294/AF186815; TTU111780/KM488349; TTU111790/KM488350; TTU111791/MZ099558; TTU109269/GU220381; TTU83378/KF733992; MSB58295/KM488348; TTU119729/MK253559; TTU138867/MK253561; TTU135171/MK253560; TTU120435/MK253564; TTU138862/MK253565.

Neotoma macrotis.—TTU81391/AF376479; MDM800/DQ179841; TTU79134/AF307836; TTU79132/AF307837; TTU79133/AF376475; TTU83037/FJ744107; MVZ198597/DQ781304; TTU83010/KC250457; TTU83358/KC250456; TTU83629/KC250459; TTU83033/KC250458; TTU83026/KC250460; TTU107307/KC250461; TTU115841/KC250462; TTU83017/KF267876; TTU107272/KF860897; TTU83019/KF860898; MDM1824/AF337758; CRS367/KP129305; CRS67/KP129304; CRS266/KP129302; CRS439/KP129301; CRS344/KP129300; CRS477/KP129299; MVZ196550/KY754061; MVZ196585/DQ179842.

Neotoma magister.—MSB74952/AF294336; SP799/AF294338; SP798/AF294337.

Neotoma melanura.—MVZ147661/AF337750; MVZ147667/AF108704; MSB55524/KM488355; MSB140889/KM488352; NMMNH2750/KM488360; NMMNH5132/KM488354; TTU110072/KM488358; TTU110074/KM488359; TTU110070/KM488357; NMMNH4947/KM488351; NMMNH4934/KM488361; MSB53951/KM488362; MSB55037/KM488353.

Neotoma mexicana.—TTU104970/KF801364; TTU104969/KF801365; MSB121363/AF298841; TTU101643/AF294346; TK45631/AF298842; TTU100791/FJ716222; DMNS8577AF186821; TTU79129/AF294345; TTU79128/AF298849; MSB74280/AF298848; MSB82309/AF298846; MSB74277/AF298847; TTU81714/AF376478; TTU107426/FJ716223; TTU110066/KF772877; TK47774/AF298843; TTU110064/KM488363; TTU122944/KY754062; CMC1204/MK803421; TTU137443/MZ064563.

Neotoma micropus.—TK31643/AF186822; TTU79086/AF376473; TTU79096/AF376474; TTU79078/KC153474; TTU79095/KC153473; TTU78977/KC153475; TTU79069/KC153472; TTU77529/AF298844; TTU136394/AF298845; TTU81033AF186825; TTU80855/AF186826; TTU80856/AF186827; TTU75113/AF376469; TTU81029/FJ716221; TTU80957/KC153488; TTU80915/FJ716220; TTU100185/KC153485; TTU115428/KC153482; TTU115413/KC153483; TTU115412/KC153484; TTU115427/KC153481; TTU115424/KC153480; TTU115410/KC153486; TTU115411/KC153487; TTU119505/KC153477; TTU109272/KC153476; TTU130662/KC153479; TTU115759/KC153478; TTU43296/FJ716217; TTU43297/KC812730; TTU35383/AF186824; TTU70894/DQ179818; ROM114902/EF989952; ROM114903/EF989953; TTU116487/MK253562; TTU116475/MK253563; TTU138864/MK253566; TTU42833/EU286808; TTU116316/KY754063.

Neotoma nelsoni.—LSUMZ36663/KC758854.

Neotoma phenax.—TTU6947/MK202784.

Neotoma picta.—TTU82667/AF305568; TK93390/AF305569; 1118.27/AB618728.

Neotoma stephensi.—MSB72986/AF307834; TTU78505/AF308867; MVZ197170/DQ781305; MVZ197173/KY754064.