c Indian Academy of Sciences

RESEARCH ARTICLE

Genetic and morphological variability in South American (, Rodentia): evidence for a complex of species

C. C. ROSA1, T. FLORES2, J. C. PIECZARKA1, R.V.ROSSI3, M.I.C.SAMPAIO2, J. D. RISSINO1, P. J. S. AMARAL1 and C. Y. NAGAMACHI1∗

1Laboratório de Citogenética, Instituto de Ciências Biológicas, Universidade Federal do Pará, Belém 66.075-900, 2Laboratório de Genética e Biologia Molecular, Universidade Federal do Pará, Campus Universitário de Bragança, Pará 68.370-000, Brazil 3Departamento de Biologia e Zoologia, Instituto de Biociências, Universidade Federal do Mato Grosso 79.070-900, Brazil

Abstract The rodent genus Oecomys (Sigmodontinae) comprises ∼16 species that inhabit tropical and subtropical forests in Central America and South America. In this study specimens of Thomas, 1904 from Belém and Marajó island, northern Brazil, were investigated using cytogenetic, molecular and morphological analyses. Three karyotypes were found, two from Belém (2n = 68, fundamental number (FN) = 72 and 2n = 70, FN = 76) and a third from Marajó island (2n = 70, FN = 72). No molecular or morphological differences were found between the individuals with differing cytotypes from Belém, but differences were evident between the individuals from Belém and Marajó island. Specimens from Belém city region may represent two cryptic species because two different karyotypes are present in the absence of significant differences in morphology and molecular characteristics. The Marajó island and Belém populations may represent distinct species that have been separated for some time, and are in the process of morphological and molecular differentiation as a consequence of reproductive isolation at the geographic and chromosomal levels. Thus, the results suggest that O. paricola may be a complex of species.

[Rosa C. C., Flores T., Pieczarka J. C., Rossi R. V., Sampaio M. I. C., Rissino J. D., Amaral P. J. S. and Nagamachi C. Y. 2012 Genetic and morphological variability in South American rodent Oecomys (Sigmodontinae, Rodentia): evidence for a complex of species. J. Genet. 91, 265–277]

Introduction first described as a subgenus of (Thomas 1906), and later accepted as a full genus (Thomas 1909). In the are important members of most faunal communi- only taxonomic review of Oecomys, Hershkovitz (1960)fol- ties, as they are cosmopolitan and native to most terrestrial lowed Thomas (1906) in considering the group as a sub- areas. Wide variation in their morphology, and diversity of genus of Oryzomys, but others disagreed with this taxonomic their habitats, climatic tolerance and food sources make them arrangement. Only after the study of Musser and Carleton the most numerous and evolutionarily successful among the (1993) was the generic status of the genus Oecomys largely orders of (Emmons and Feer 1997). accepted. Recent phylogenetic studies including species of The subfamily Sigmodontinae, which includes most of Oecomys have confirmed the monophyly of this genus the species of South American rodents, occurs only in (Patton and Da Silva 1995; Smith and Patton 1993; Patton Americas. It includes 386 species in 81 genera and nine et al. 2000; Andrade and Bonvicino 2003; Weksler 2003, tribes: Abrotrichini, Akodontini, Ichthyomyini, , 2006), which now includes 16 valid species (Musser and Phyllotini, Reithrodontini, Sigmodontini, Thomasomyini Carleton 2005; Carleton et al. 2009). and Wiedomyini (Reig 1980; Smith and Patton 1999; Musser Genus Oecomys occurs in tropical and subtropical rain- and Carleton 2005; Weksler et al. 2006; D’Elía et al. forests in Central America and South America. In the 2007). The tribe Oryzomyini comprises 28 genera (Weksler Brazilian Amazon region there are nine species, of which et al. 2006) and includes the genus Oecomys,whichwas only five have had their karyotype described (Andrade and Bonvicino 2003; Langguth et al. 2005). Cytogenetic stud- ∗ For correspondence. E-mail: [email protected]. ies of Oecomys show that the diploid number ranges from Keywords. cytogenetics; chromosome; speciation; cryptic species; rodent; Oecomys paricola.

Journal of Genetics, Vol. 91, No. 3, December 2012 265 C. C. Rosa et al.

Table 1. Chromosomal characterization of the genus Oecomys. 58 to 86 (table 1), suggesting a high level of chromosomal rearrangement in this species (Gardner and Patton 1976; Species 2n FN Reference Patton et al. 2000). This karyotypic variability demonstrates the importance of chromosomal studies in taxonomic identi- Oecomys sp. 86 98 Gardner and Patton (1976) O. bicolor 80 134–136 Gardner and Patton (1976) fication of the species (Langguth et al. 2005). O. bicolor 80 140 Patton et al. (2000) Oecomys paricola Thomas, 1904 is an Amazonic species O. roberti 80 114 Patton et al. (2000) with an incompletely defined distribution south of the O. roberti 82 106 Langguth et al. (2005) Amazon river and east of the Xingu river (Voss et al. 2001; O. superans 80 108 Gardner and Patton (1976) Musser and Carleton 2005; Rocha et al. 2011). No karyotype Oecomys sp. 72 90 Andrade and Bonvicino (2003) O. paricola 70 76 Present work information has been reported for this species. O. paricola O. paricola 68 72 Present work has predominantly grey-based ventral fur, except for chin and O. paricola 70 72 Present work throat, which are usually selfcoloured (all pale), sometimes O. concolor 60 62 Patton et al. (2000) with a ventral self-coloured midline extending caudally to O. bahiensis 60 62 Langguth et al. (2005) the groin (Voss et al. 2001). The tail is uniformly dark, with O. trinitatis 58 96 Patton et al. (2000)

2n, diploid number; FN, fundamental number.

Table 2. Specimens analysed, with localities, specimen preparation type, collection and GenBank number.

Material Locality Skin Skull Fluid Cyt b Kar. Collection number GenBank number

Barcarena (01◦30S48◦40W) X X BAR006 X X X BAR013 JF759675 X X X BAR017 JF759676 X X X BAR023 JF759677 X X X BAR029 JF759678 Belém, Parque Ambiental de Belém X X MPEG2477 (Utinga) (01◦27S48◦29W) X X MPEG2584 X X MPEG2593 X X MPEG2605 X MPEG2610 X X MPEG2614 X X MPEG2615 X MPEG2632 X MPEG2635 X X MPEG2647 X X X MPEG38659 JF759679 X X X MPEG38664 JF759680 X X X X MPEG39699 JF759681 X X X X MPEG39701 JF759682 X X X MPEG39703 X X X X MPEG39705 JF759683 X X X MPEG39708 JF759684 X X MPEG39709 X MPEG40732 BR-010, km 87–97 (01◦40S47◦47W) X X MG8382 X X MG8438 Marajó island, Chaves, X X X MPEG40842 JF759668 Fazenda Tauarí (00◦39S50◦11W) X X X MPEG40843 JF759669 X X X MPEG40844 JF759671 X X X MPEG40845 JF759670 X X X MPEG40846 X X X MPEG40848 JF759672 X X X MPEG40849 JF759673 X X MPEG40850 X X X X MPEG40851 JF759674

Kar., karyotype; MPEG, Museu Paraense Emilio Goeldi; BAR, specimens from Universidade Federal do Pará (Barcarena project) X, material available.

266 Journal of Genetics, Vol. 91, No. 3, December 2012 Genetic and morphological variability in Oecomys a terminal tuft of hairs 6–10 mm long. Cranially, O. pari- region and Marajó island, in Pará, Brazil. This is the first cola has a primitive pattern of carotid arterial supply, sep- description of the karyotype of the species, and provides new arated accessory oval and buccinator–masticatory foramina; information on molecular and morphologic variation in the the subsquamosal fenestra are absent. species. The identification of rodent species using morphological traits is very complex (Patton and Gardner 1972; Gardner and Emmons 1984), leading to uncertainties in the of many genera of this group. Studies combining morpho- Material and methods logical, molecular and karyotype analysis have the potential Karyotype analysis to demonstrate occurrence of a greater diversity of Oecomys species, resulting in identification of new species, and reval- We karyotyped six specimens of O. paricola (table 2), of idation of previously described ones (Patton and Da Silva which four were collected in the Environmental Park in 1995; Smith and Patton 1999; Andrade and Bonvicino 2003; Belém (01◦27S, 48◦29W) and two (males) were collected at Weksler 2006). Molecular studies using the cytochrome b Marajó island (01◦00S, 49◦30W), in the region of the Ama- gene as a marker in phylogenetic studies of Sigmodonti- zon river mouth (figure 1). The specimens were collected nae rodents are particularly relevant (Smith and Patton 1991, using pitfall traps (buried 60 L buckets) connected by plas- 1993, 1999; Patton et al. 2000; Andrade and Bonvicino tic tapes, and with conventional traps including the Sherman 2003; D’Elía 2003; Miranda et al. 2007; D’Elía et al. 2008; trap and cages on the ground containing baits of peanuts, Catzeflis and Tilak 2009). This gene has been reported to flour and canned fish. Samples for metaphase chromosome have arrangements consistent with the species boundaries analysis were taken in the field using the bone marrow tech- based on classic taxonomic studies (Johns and Avise 1998; nique (Ford and Hamerton 1956). We also made laboratory Avise and Walker 1999), making it highly appropriate for culture of fibroblasts. biodiversity studies (Bradley and Baker 2001). The chromosome analysis involved conventional stain- In this study we undertook chromosomal, molecular and ing, G-banding and C-banding (Seabright 1971; Sumner morphological analysis of O. paricola from the Belém city 1972), and Ag-NOR staining (Howell and Black 1980). The

Figure 1. Map with the collect locations.

Journal of Genetics, Vol. 91, No. 3, December 2012 267 C. C. Rosa et al. karyotypes were organized by size and morphology of the bisection-reconnection (TBR) algorithm. Clade support was chromosome pairs. performed for both analysis using bootstrap with 1000 repli- cates. Estimates of evolutionary divergence of sequence pairs between and within Oecomys species and populations were Molecular analysis conducted using the Kimura-2 parameter method in MEGA4 (Kimura 1980;Tamuraet al. 2007). All codon positions were We extracted 731-bp sequences of the cytochrome b gene included. from 17 ethanol-preserved samples of muscle tissue of O. paricola, including four of the six specimens we used for karyotype analysis (table 2). We also used nine Oecomys Morphological and morphometric analyses sequences available in GenBank (several species, table 3), two sequences of O. auyantepui Tate, 1939 and O. rutilus We examined the morphological characteristics of 35 spec- Anthony, 1921 (JF759666 and JF759665, respectively) imens of O. paricola from Pará State, held by the Museu obtained in our laboratory, and one GenBank sequence of Paraense Emílio Goeldi (MPEG); these included skulls, dried megacephalus Fischer, 1814 (AY275124), as the skins and preserved fluids (table 2). Specimens temporar- outgroup in the phylogenetic analysis. ily housed in the Universidade Federal do Pará (UFPA), DNA extraction involved the phenol–chloroform and pro- identified in this study by the abbreviation BAR (Barcarena teinase K–RNAase protocol (Sambrook et al. 1989). Frag- project), will be deposited in MPEG. All six specimens for ments of cyt b were isolated and amplified by polymerase which we obtained karyotypes were included in the morpho- chain reaction (PCR) using primers MVZ05 and MVZ16 logical and morphometric analyses (table 2). For compari- (Smith and Patton 1993). The amplification protocol con- son, we also examined the following 15 MPEG specimens of sisted of initial denaturation at 94◦C for 3 min, 35 denatura- O. auyantepui: 7169, 13132, 39793, 39794, 39798, 39804, tion cycles of 30 s at 94◦C, 1 min of annealing at 45◦C, 2 min 39816, 39817, 39825, 39831, 39836, 40076, 40078, 40084 of extension at 72◦C, and a final extension at 72◦Cfor7min. and 40085. Sequences were edited and aligned in BioEdit 7.0.5.2 To enable more accurate morphological and morphomet- (Hall 1999) using ClustalW (Larkin et al. 2007), following ric comparisons, the specimens were classified into three age the parameters proposed by Schneider (2006). We verified classes based on the eruption pattern and the differential wear sequence saturation using DAMBE 5.2.34 software (Xia and of the occlusal surface of the superior molars. These classes, Xie 2001), and undertook phylogenetic analysis. Using Mod- modified from Voss (1991) and Brandt and Pessôa (1994), eltest 3.6 (Posada and Crandall 1998) in PAUP* 4.0, we found were: (i) age class 1 (young)—M3 incompletely erupted or the best evolutionary TrN+G model fit to our sequences unworn; (ii) age class 2 (adult)—occlusal surface exhibiting with a substitution rate of 6, a gamma distribution param- slight to moderate wear, but still tubercular; mesoflexus and eter of 1.6368, and an invariable sites proportion of 0.0. paraflexus of M1 and M2 sometimes as enamel islands; all Maximum parsimony (MP) was determined using PAUP* M3 flexus, except paraflexus, obliterated and sometimes as 4.0 with a heuristic search; the starting tree was obtained enamel islands; (iii) age class 3 (old adult)—occlusal sur- by stepwise addition and branch-swapping using a tree- face flat or concave; only paraflexus, metaflexus, protoflexus

Table 3. GenBank sequences used in our phylogenetic analysis.

Species GenBank entry Locality

Oecomys bicolor AF108699 Peru∗ (Smith and Patton 1999) O. bicolor OBU58382 Brazil, Acre, Sobral, left bank Rio Juruá. 08◦22S72◦49W (Patton and Da Silva 1995; Patton et al. 2000) O. roberti ORU58384 Brazil, Amazonas, Penedo, right bank Juruá. 06◦50S70◦45W (Patton and Da Silva 1995; Patton et al. 2000) Oecomys sp. AY072772 Brazil, Mato Grosso do Sul, Corumbá. 19◦00S57◦36W (Andrade and Bonvicino 2003) Oecomys sp. OSU58388 Brazil, Amazonas, Lago Vai-Quem-Quer, right bank Rio Juruá. 03◦19S66◦01W (Patton and Da Silva 1995; Patton et al. 2000) O. superans OSU58385 Brazil, Amazonas, Penedo, right bank Juruá. 06◦50S70◦45W (Patton and Da Silva 1995; Patton et al. 2000) O. trinitatis OTU58390 Brazil, Acre, Opposite Igarapé Porongaba, left bank Rio Juruá. 08◦40S72◦47W (Patton and Da Silva 1995; Patton et al. 2000) AY275124 Not specified (Smith and Patton 1991)

∗There are two localities in Peru associated with two specimens of (Patton et al. 2000), but it is not clear from which one is the GenBank sequence.

268 Journal of Genetics, Vol. 91, No. 3, December 2012 Genetic and morphological variability in Oecomys and hypoflexus are present in M1 and M2; others, when The following external measurements were obtained present, are just enamel islands; paraflexus of M3 always as directly from the specimen labels, and were used for descrip- an enamel island, or totally absent. tive statistics only: head and body length (HBL), tail length We evaluated morphological characteristics of skin and (TL), foot length (FL), ear length (EL) and weight (W). We skull of two populations of O. paricola (one from the Belém also obtained 10 craniodental measurements (to the near- region and the other from Marajó island) for use in sepa- est 0.01 mm) from each of the 32 skull specimens using rating groups of individuals. Comparisons among specimens digital calipers during skull examination with a stereomi- were made with respect to sex and age classes. To avoid mis- croscope. These included: (i) condylo-incisive length (CIL), interpreting ontogenetic variation as taxonomic variation we from the greater curvature of one upper incisor to the artic- did not restrict comparisons to the specimens for which we ular surface of the condyle on the same side; (ii) length of have karyotypes, as they belonged to different age classes. diastema (LD), from the crown of the first cheek tooth to The anatomical nomenclature used followed Pocock (1914), the lesser curvature of the incisor on the same side; (iii) Hershkovitz (1962, 1977), Carleton and Musser (1984, length of molars (LM), the crown length from M1 to M3; (iv) 1989)andVoss(1988) for external morphology, and breadth of M1 (BM1), the greatest crown breadth of the first McDowell (1958), Hershkovitz (1962), Wahlert (1974), upper molar (M1); (v) length of incisive foramen (LIF), the Carleton and Musser (1984, 1989), Voss (1988), Steppan greatest anterior–posterior dimension of one incisive fora- (1995) and Weksler (2006) for cranial morphology. men; (vi) breadth of incisive foramen (BIF), the greatest

Figure 2. The karyotype with 2n = 68 from Belém. (A) G-banding. (B) C- banding. Boxed chromosomes: pair XY from a male karyotype. Bars = 10 μm.

Journal of Genetics, Vol. 91, No. 3, December 2012 269 C. C. Rosa et al.

and posterior edges of the ; (ix) least interor- bital breadth (LIB), the least distance across the frontal bones between the orbital fossa; and (x) zygomatic breadth (ZB), the greatest transverse dimension across the squamosal zygomatic processes. We evaluated the presence of sexual dimorphism in our largest sample, from the Belém region (six females and seven males), using Hotelling’s T-square test (Hotelling 1931). We also used this test to evaluate morphometric differ- ences between specimens from the Belém region and Marajó island. The morphometric analyses were conducted on spec- imens of age class 2 only, using PAST 2.02 (Hammer et al. 2001), as this class had the greatest number of samples, and because skull size appeared to vary greatly among different age classes. Figure 3. Ag-NOR staining (arrows) on the karyotype with 2n = 68 from Belém. Bars = 10 μm. Results

Karyotype analysis transverse dimension across both incisive foramina; (vii) breadth of palatal bridge (BPB), measured between the pro- Environment Park, Belém: The specimens MPEG 39703 tocones of the right and left M1; (viii) breadth of zygo- (male) and MPEG 39699 (female) had the karyotype 2n = 68, matic plate (BZP), the least distance between the anterior fundamental number (FN) = 72, comprising 30 one-armed

Figure 4. The karyotype with 2n = 70 from Belém. (A) G-banding. (B) C-banding. Bars = 10 μm.

270 Journal of Genetics, Vol. 91, No. 3, December 2012 Genetic and morphological variability in Oecomys

this cytotype. Constitutive heterochromatin was found at the centromeric region of all chromosomes. The short arm of the X chromosome was almost all heterochromatic, and almost all the Y chromosomes, were heterochromatic (figure 2B). The nucleolus organizer region (NOR) was found on the small short arm of three one-armed pairs (one mid-sized and two small; figure 3). Specimens MPEG 39701 (male) and MPEG 39705 (fe- male) had the karyotype 2n = 70, FN = 76, comprising 30 one-armed chromosome pairs and four bi-armed pairs. The X chromosome was large and bi-armed and the Y chromosome was small and submetacentric (figure 4A). Constitutive hete- rochromatin was found in the centromeric region of all chro- mosomes. The short arm of the X chromosome was mostly heterochromatic, and almost the entire Y chromosome was heterochromatic (figure 4B). The NOR was found on the small short arm of two one-armed pairs (one mid-sized and Figure 5. Ag-NOR staining (arrows) on the karyotype with 2n = 70 from Belém. Bars = 10 μm. one small; figure 5).

Marajó island: The MPEG specimens 40846 and 40851 pairs and three bi-armed pairs. The X chromosome was large (both males) had a karyotype of 2n = 70, FN = 72, com- and bi-armed, and the Y chromosome was small and sub- prising 32 one-armed chromosome pairs and two bi-armed metacentric. Figure 2A shows the G-banded karyotype of pairs. The X chromosome was large and bi-armed, and

Figure 6. The karyotype with 2n = 70 from Marajó island. (A) G-banding. (B) C-banding. Bars = 10 μm.

Journal of Genetics, Vol. 91, No. 3, December 2012 271 C. C. Rosa et al.

Figure 7. Ag-NOR staining (arrows) on the karyotype with 2n = 70 from Marajó island. Bars = 10 μm. the Y chromosome was mid-sized and submetacentric (figure 6A). Constitutive heterochromatin was found at the centromeric region of all chromosomes. The X chromosome had a heterochromatic block at the distal region of the short arm, and the Y chromosome was almost entirely heterochro- matic (figure 6B). The NOR was found on the proximal long arm of two pairs of mid-sized acrocentric autosomes (figure 7).

Figure 8. Maximum parsimony analysis of Oecomys paricola Molecular analysis from Belém region and Marajó island. The data set of 731 bp of cyt b had 91 variable sites and 134 parsimony informative characters. In our maximum parsimony (MP) analysis based on the cyt Cladogram shows the best tree found, with 482 steps long, consis- b gene (figure 8) the specimens from Marajó island grouped tency index (CI) of 0.577, homoplasy index (HI) of 0.423, reten- tion index (RI) of 0.693, and rescaled consistency index (RC) of together in a clade that was distinct from those collected in 0.400. Bootstrap support ≥70% is shown above branches. Legend: Belém, and there was high nodal support (93% for the Belém 1Oecomys paricola from Belém region; 2Oecomys paricola from clade; 99% for the Marajó island clade). Nodal support for Marajó island.

Table 4. Estimates of evolutionary divergences over sequence pairs within and among Oecomys species based on 731 bp of cyt b gene.

1234567891011

Oecomys paricola (Belém region) 0.4 O. paricola (Marajó island) 3.9 0.6 O. auyantepui 9.9 10.4 n/c O. rutilus 13.3 14.1 12.1 n/c O. bicolor 11.0 11.5 11.7 12.3 6.1 O. roberti 10.4 11.7 11.4 14.6 9.0 n/c Oecomys sp. (Corumbá.MT) 11.8 11.9 11.4 12.9 8.8 10.7 n/c Oecomys sp. (Juruá river) 11.5 11.6 11.6 12.0 8.7 10.5 8.0 n/c O. superans 9.5 10.5 12.3 12.0 9.0 7.7 9.7 9.8 n/c O. trinitatis 13.3 13.7 13.2 15.5 12.1 11.8 10.9 11.6 11.3 n/c Hylaeamys megacephalus 18.8 17.8 16.9 18.5 16.9 17.4 17.0 16.9 16.7 17.4 n/c

For Oecomys paricola, divergences within and among populations are shown. Bold numbers represent the evolutionary divergences within species or populations calculated using the Kimura-2 parameter method in MEGA4 (Geise et al. 1998, 2001). The rate of variation among sites was modelled with a gamma distribution (shape parameter = 1.6368). All estimates are shown in percentage. n/c, cases in which it was not possible to estimate evolutionary distances.

272 Journal of Genetics, Vol. 91, No. 3, December 2012 Genetic and morphological variability in Oecomys 5 72 = MPEG 40851 4 70, NF = n pes discussed in text are shown MPEG 40846 6 3.7 (97–106) 43.0 (21–28) 4 103 23 125 22 1.4 (13–16) 4 14 16 0.5 (13.8–15.0) 4 14.26 15.63 5.9 (110–124) 4 120 132 0.36 (24.4–25.3) 40.21 (7.10–7.56) 40.18 (3.98–4.37) 4 25.02 7.10 3.99 26.81 7.90 4.29 0.03 (1.14–1.20) 40.23 (4.44–4.95) 40.06 (2.16–2.28) 40.23 (2.47–2.93) 4 1.200.05 (2.37–2.46) 4 4.440.12 (4.68–4.95) 4 2.17 2.47 2.37 4.68 1.19 5.03 2.48 2.80 2.81 5.23 ± ± ± ± ± ± ± ± ± ± ± ± ± ± Descriptive statistcs 5 76 2 = MPEG 39705 4 70, NF = MPEG 39701 Based on the follow specimens of age class 2: MPEG 40844, 40846, 40848, 408. 6 3 from Belém region and Marajó island. MPEG 39703 Female, age class 3. 5 72 2n 2 Belém region Marajó island = Oecomys paricola 68, NF = MPEG 39699 2n Male, age class 2. 4 Based on the following specimens of age class 2: MPEG 2477, 2605, 2614, 2615, 8438, 38659, 39701, 39708, 40732; BAR 006, 013, 023, 029. 1 1 Male, age class 1. 0.8 (22–24) 9 22 22 23 22 21 2.1 (12–19) 9 13 14 15 14 15 12.4 (90–129) 9 78 71 99 104 105 13.8 (80–125) 9 98 89 97 124 119 3 0.11 (2.0–2.3) 12 1.83 1.95 2.11 2.18 2.30 0.53 (5.8–7.6) 130.10 (3.8–4.2) 13 5.82 - 5.43 - 6.66 3.95 7.20 7.52 3.89 4.18 0.33 (3.9–5.0) 130.14 (2.4–2.9) 120.25 (1.0–2.6) 130.17 (4.6–5.2) 12 4.14 2.23 2.02 4.50 3.63 2.21 1.73 4.71 4.08 2.44 2.07 4.77 4.51 2.66 2.62 4.89 4.88 2.75 2.52 4.97 1.11 (21.5–25.0) 10 20.77 19.91 23.03 24.68 25.6 0.61 (12.1–14.5) 11 12.1 12.11 12.87 13.64 14.83 0.06 (1.01–1.18) 13 1.14 1.12 1.12 1.12 1.19 ± ± ± ± ± ± ± ± ± ± ± ± ± ± Selected external and cranial dimensions for Female, age class 1. Mean, standard deviation and range (in parenthesis) of measurements are given in mm, followed by sample size. Measurements of specimens with karyoty in different columns for comparisons. Table 5. VariableHBL Descriptive statistcs 103 2 TL 108 FL 23 EL 15 CIL 23.30 LDLM 6.77 4.02 BM1 1.12 LIFBIFBPBBZP 4.36 LIB 2.15 ZB 2.57 2.35 4.84 13.58

Journal of Genetics, Vol. 91, No. 3, December 2012 273 C. C. Rosa et al. the monophyletic clade that corresponded to O. paricola was bilaterally present in four others. The mandible also differed 100% for the MP analysis. Relationships among Oecomys between the two populations with respect to the capsular pro- species were not supported by bootstrap statistics. cess of the lower incisor alveoli, which varied from slightly The intrapopulation genetic divergences were 0.4% for the to moderately curved in the specimens from Marajó island, Belém clade and 0.6% for the Marajó Island clade, while but was slightly curved in the specimens from Belém. the divergence between the populations was 3.9%. The intra- Hotelling’s T-square test for sexual dimorphism showed generic divergences ranged from 7.7% to 15.5%, with diver- that there was no statistically significant difference between gences between O. paricola and other congeneric species the sexes (P = 0.4684) in the Belém population. The varying from 9.7% to 14.1% (table 4). same test was applied to evaluate morphometric differ- ences between the Belém and Marajó island specimens. This showed that there were no statistically significant differences Morphological and morphometric analyses between the specimens from the two locations (P = 0.2825), We found no morphological characters that could be used to though the mean values for almost all external measure- discriminate the two Belém cytotypes, other than the smaller ments and skull dimensions of the Marajó island specimens size of the specimens of karyotype 2n = 68, NF = 72 rel- examined were greater (table 5). ative to those of karyotype 2n = 70, NF = 76 (table 5). However, the size difference may be explained by ontoge- netic variation, as the former cytotype was based on young specimens from age class 1, while the latter was based on Discussion adult and older specimens belonging to age classes 2 and 3. In contrast, the specimens from Belém had several charac- Here we have provided the first description of the karyotype ters distinguishing them from the Marajó island specimens. of O. paricola. We found three cytotypes, two in the spec- With respect to external features, the specimens from Marajó imens from the Belém city region (2n = 68, FN = 72 island were very similar to those from Belém. However, in and 2n = 70, FN = 76) and a third from Marajo island the former the colour of the ventral pelage was grey-based (2n = 70, NF = 72). The karyotypes with 2n = 68 and with a broad midline of cream or white hairs, but latter it 2n = 70 are new for the genus Oecomys. The differences was completely grey-based. In both populations, the chin in the diploid number and chromosome morphology of the and throat were always selfcoloured (cream or white). Foot three karyotypes of O. paricola can be explained by rear- colouration and the plantar surface were also different in rangements, including fusion/fission and pericentric inver- the two populations. In the Marajó island specimens the feet sions. Because of the number of chromosomes and differ- were dark brown with a darker spot on the dorsum, while in ences in their condensation, comparative analysis of the G- the Belém specimens the feet were cream or light brown with banding patterns was not possible, and consequently we were a brown spot on the dorsum. The plantar surface was smooth not able to precisely identify the chromosomes involved in or had several squamae near the plantar pads in specimens the karyotype differentiation, or the chromosome rearrange- from Marajó island, but was always smooth in Belém spec- ments that occurred. No hybrid karyotypes were found in the imens. We also observed that the tail length was ∼116% of Belém sample, indicating that there has been no gene flow head and body length in Marajó island specimens, whereas among the different karyotypes bearers. it was only ∼106% of head and body length in the Belém Chromosome rearrangement can be the main cause of spe- specimens (table 5). ciation when it causes a negative heterosis effect, by reducing The skulls were also notably different between the two the fertility of heterozygous bearers (King 1987, 1993). This populations. In the Marajó island population, the nasal is a postmating reproductive isolation (rearrangements that bone was short and did not surpass the lachrymal–frontal– segregate badly in meiosis in structural hybrids, producing maxillary suture, whereas in specimens from Belém this unbalanced gametes and consequent reduction in fertility). bone was long, and aligned with or surpassing the lachrymal– There is some evidence of correlation between population frontal–maxillary suture. In the specimens from Marajó structure and karyotypic variability, with greater variability island the supraorbital ridges were well developed, project- occurring in species that have small inbreeding populations, ing dorsally from the border of the frontal bone at the orbital such as occurs in rodents (Bush et al. 1977; Bengtsson 1980; fossa and extending slightly onto the parietal bone, whereas Maruyama and Imai 1981), relative to species with outbreed- in the specimens from Belém they were poorly developed ing populations, such as cats and whales (Bush et al. 1977). and restricted to the frontal bone. The frequency of sub- Population structure has a crucial role in fixing rearrange- squamosal fenestra, which was used previously to distin- ments, as this only happens if the rearranged chromosome guish O. auyantepui from O. paricola (Voss et al. 2001), was can rapidly become homozygous. In small inbreeding pop- highly variable in the specimens from Marajó island. Among ulations the probability of this homozygous condition arising the 16 specimens from Belém it was bilaterally present in 15, is greater, making the rearranged chromosome more common. and unilaterally present in one, whereas in specimens from Bradley and Baker (2006) assessed whether the degree Marajó island it was bilaterally absent in four specimens and of cyt b sequence divergence in mammals can be used for

274 Journal of Genetics, Vol. 91, No. 3, December 2012 Genetic and morphological variability in Oecomys species-level differentiation within the genetic species con- but there has been insufficient time for the fixation of molec- cept. They observed that sister species recognized on the ular and morphological differences. This shows that chromo- basis of morphology often have cyt b distance values >5%. some rearrangements can have an important role as a post- With respect to rodents they found that intrapopulation diver- mating reproductive isolation mechanism. A similar situation gence values ranged from 0.0 to 1.4%, intraspecific diver- was reported previously in a morphological and cytogenetic gence values ranged from 0.0 to 4.7%, and intrageneric study of rodents from the Lake Chad region, Africa (Granjon divergence values ranged from 1.3 to 16.9%. In molecular and Dobigny 2003). They demonstrated that a significant studies of sigmodontine rodents, not considered by Bradley portion of the species are cryptic (morphologically indistin- and Baker (2006), values have been reported that range from guishable), and can only be distinguished by karyotype anal- 0 to 3.87% for intrapopulation divergence, 0 to 11.37% for ysis. This situation occurs in many vertebrates groups in the intraspecific divergence, and 1.23 to 21% for intrageneric Neotropical region. For instance, Milhomem et al. (2008) divergence (Smith and Patton 1991, 1993, 1999; Patton et al. used classical cytogenetics to describe two cryptic species 2000; D’Elía 2003; Miranda et al. 2007; D’Elía et al. 2008; in the species complex of Gymnotus carapo Linnaeus, 1758 Catzeflis and Tilak 2009). For Oecomys, no intrapopulation (a Neotropical electric fish); the species are morphologically divergence values have been reported, but intraspecific diver- identical, and have diverged only in the karyotype (2n = 40 gence ranges from 0 to 10.3%, and intrageneric divergence and 2n = 42), apparently by a fusion and some inver- ranges from 6.5 to 12% (Smith and Patton 1999; Patton et al. sions. Nagamachi et al. (2010) used chromosome painting to 2000; Andrade and Bonvicino 2003; Rocha et al. 2011). demonstrate that the level of chromosome reorganization is Specifically for O. paricola, Rocha et al. (2011) reported greater than previously thought, confirming that Gymnotus intraspecific divergence of 3.6% based on cyt b gene among carapo with different karyotypes are truly different species. populations from eastern and western mid-Araguaia river. The chromosome differentiation between the Marajó The intrapopulation divergences of O. paricola from island and Belém cytotypes must have occurred some time Marajó island (0.6%) and Belém region (0.4%) are in agree- ago, because our results indicate differences at the morpho- ment with those reported by Bradley and Baker (2001)for logic level, which must have occurred since the interrup- rodents in general, and for sigmodontine species (citations tion of gene flow. A similar situation has been reported in above). Similarly, the intraspecific evolutionary divergence the genus Akodon (Geise et al. 2005), where A. montensis we found (3.9% in O. paricola) was close to the 3.6% Thomas, 1913, A. cursor Winge, 1887 and A.affcursor form found by Rocha et al. (2011). The intrageneric divergences a species complex with other undescribed species (Silva (7.7 to 15.5%) we found are greater than those previously and Yonenaga-Yassuda 1998). These species are so similar reported for Oecomys, but are comparable to those of other that they cannot be distinguished at the morphological level sigmodontine rodents. (Christoff et al. 2000). However, mitochondrial (Geise et al. Direct morphological comparisons among the specimens 2001) and karyological data (Silva and Yonenaga-Yassuda for which we had karyotype data were not possible, as one of 1998;Geiseet al. 1998) show that they are distinct species. the Belém cytotypes was represented by young individuals. The occurrence of morphologically identical taxa with dif- However, our morphological analysis, which included sev- ferent diploid and fundamental numbers is common among eral specimens from the Belém region showed no significant rodents. For instance, in Oecomys, the karyotype of O. differences consistent with the presence of more than one bicolor Tomes, 1860 has 2n = 80 and FN = 134–136 taxonomic unit. Similarly, our molecular analysis showed (Gardner and Patton 1976). Another karyotype was subse- low genetic divergence (0.4%) and no genetic structure quently found in O. bicolor (Patton et al. 2000); this had within this population, even among specimens representing the same diploid number, but had three bi-armed chromo- two of the three cytotypes described (figure 8). somes instead of the three one-armed chromosomes in the In contrast, we found consistent morphological differences initial karyotype, (2n = 80, FN = 140; table 1). The phylo- between the populations from the Belém region and Marajó genies described by Smith and Patton (1999), and Andrade island. These differences, despite (i) the genetic divergence and Bonvicino (2003) suggest that O. bicolor is a species among the populations (3.9%), (ii) the genetic structure indi- complex. Two karyotypes have also been described for O. cated by the molecular analysis (figure 8), and (iii) the roberti Thomas, 1904, one (2n = 80, FN = 114) with 21 unique karyotype of the specimens from Marajó island (fig- one-armed chromosome pairs and 18 bi-armed pairs (Patton ures 6–7), suggest that these two populations might be dis- et al. 2000), and the other (2n = 82, FN = 106) with 27 one- tinct species. Morphologically, the specimens from Marajó armed chromosome pairs and 13 bi-armed pairs (Langguth island are similar to specimens of O. auyantepui, with which et al. 2005). they share the same pattern of ventral pelage colour, length of the nasal bones, and development and extension of the supraorbital ridges, despite the genetic divergence of 10.4% Conclusions among these species. It is possible that the two cytotypes from Belém represent cryptic species, whereby recent specia- In sum, in the present study, different analytic methods pro- tion mediated by chromosome rearrangements has occurred, duced different results with respect to the species status of the

Journal of Genetics, Vol. 91, No. 3, December 2012 275 C. C. Rosa et al.

O. paricola populations. We conclude that O. paricola from Carleton M. D., Emmons L. H. and Musser G. G. 2009 A new Belém may represent two cryptic species because two differ- species of the rodent genus Oecomys (: Sigmodonti- ent karyotypes are present in the population in the absence nae, Oryzomyini) from eastern , with emended definitions of O. concolor (Wagner) and O. mamorae (Thomas). Am. Mus. of significant differences in morphology and molecular Novit. 3661, 1–32. characteristics. No hybrid karyotypes were found, which Catzeflis F. and Tilak M. 2009 Molecular systematic of Neotropi- indicates absence of gene flow among the cryptic species. cal spiny mice (: Sigmodontinae, Rodentia) from the Additionally, the Marajó island and Belém populations rep- Guianan Region. Mammalia 73, 239–247. resent distinct species that have been separated for some time Christoff A. U., Fagundes V., Sbalqueiro I. J., Mattevi M. S. and Yonenaga-Yassuda Y.2000 Description of new species of Akodon and are undergoing morphological and molecular differentia- (Rodentia: Sigomodontinae) from southern Brazil. J. . tion following reproductive isolation at the geographical and 81, 838–851. chromosomal levels. This is possibly an example of specia- D’Elía G. 2003 Phylogenetics of Sigmodontinae (Rodentia, tion where the morphological and molecular differences are Muroidea, Cricetidae), with special reference to the akodont accumulating following reproductive isolation. group, and with additional comments on historical biogeography. Cladistics 19, 307–323. D’Elía G., Pardiñas U. F. J., Teta P. and Patton J. L. 2007 Definition and diagnosis of a new tribe of sigmodontine rodents (Cricetidae: Acknowledgements Sigmodontinae), and a revised classification of the subfamily. Gayana 71, 187–194. C. C. Rosa is a recipient of FAPESPA Mastership Scholarship D’Elía G., Pardiñas U. F. J., Jayat J. P. and Salazar-Bravo J. in Neurosciences and Cellular Biology; J. C. Pieczarka, M. I. C. 2008 Systematics of Necromys (Rodentia, Cricetidae, Sigmod- Sampaio, C. Y. Nagamachi are CNPq Researchers; P. J. S. ontinae): species limits and groups, with comments on historical Amaral is a recipient of FAPESPA Doctorship Scholarship in biogeography. J. Mammal. 89, 778–790. Genetic and Molecular Biology at Laboratório de Citogenética, Emmons L. H. and Feer F. 1997 Neotropical rainforest mammals: Instituto de Ciências Biológicas, Universidade Federal do Pará, a field guide. 2nd edition, University of Chicago, Chicago. Belém, Brazil. T. Flores is a recipient of CAPES Mastership Scho- Ford C. E. and Hamerton J. L. 1956 A colchicine, hypotonic-citrate, larship in Zoology at Laboratório de Genética e Biologia Molecular, squash sequence for mammalian chromosomes. Stain. Technol. Universidade Federal do Pará, Campus Universitário de Bragança, 31, 247–251. Pará, Brazil. Gardner A. L. and Emmons L. 1984 Species group in Proechimys The authors thank the Conselho Nacional de Desenvolvimento (Rodentia, Echimyidae) as indicated by karyology and bullar Científico e Tecnológico (CNPq) and the Fundação de Amparo à morphology. J. Mammal. 65, 10–25. Pesquisa do Estado do Pará (FAPESPA) for financial support, and Gardner A. L. and Patton J. L. 1976 Karyotypic variation in ory- the Instituto Brasileiro de Meio Ambiente e dos Recursos Naturais zomyine rodents (Cricetinae) with comments on chromosomal Renováveis (IBAMA) for authorizing sample collection. evolution in the neotropical cricetine complex. Occ. Pap. Mus. Zool. Lousiana St. Univ. 49, 1–48. Geise L., Canavez F. C. and Seuánez H. N. 1998 Comparative kary- ology in Akodon (Rodentia, Sigmodontinae) from Southeastern References Brazil. J. Hered. 89, 158–163. Geise L., Smith M. F. and Patton J. L. 2001 Diversification in the Andrade A. F. B. and Bonvicino C. R. 2003 A new karyologi- genus Akodon (Rodentia, Sigmodontinae) in Southeastern South cal variant of Oecomys (Rodentia: Sigmodontinae) and its phy- America: mitochondrial DNA sequence analysis. J. Mammal. 82, logenetic relationship based on molecular data. Genome 46, 92–101. 195–203. Geise L., De Moraes D. A. and Da Silva H. S. 2005 Morphometric Avise J. C. and Walker D. 1999 Species realities and numbers in differentiation and distributional notes of three species of Akodon sexual vertebrates: perspectives from an asexually transmitted (Muridae, Sigmodontinae, Akodontini) in the Atlantic Coastal genome. Proc. Natl. Acad. Sci. USA 96, 992–995. area of Brazil. Arq. Mus. Nac., Rio de Janeiro 63, 63–74. Bengtsson B. O. 1980 Rates of karyotype evolution in placental Granjon L. and Dobigny G. 2003 The importance of cyto- mammals. Hereditas 92, 37–47. taxonomy in understanding the biogeography of African Bradley R. D. and Baker R. J. 2001 A test of the genetic species rodents: Lake Chad murids as an example. Mammal Rev. 33, concept: cytochrome-b sequences and mammals. J. Mammal. 82, 77–91. 960–973. Hall T. A. 1999 BioEdit: a user-friendly biological sequence Bradley R. D. and Baker R. J. 2006 Speciation in mammals and the alignment editor and analysis program for Windows 95/98/NT. genetic species concept. J. Mammal. 87, 643–662. Nucleic Acids Symp. Ser. 41, 95–98. Brandt R. S. and Pessôa L. M. 1994 Intrapopulational variabil- Hammer O., Harper D. A. T. and Ryan P. D. 2001 PAST: Pale- ity in cranial characters of Oryzomys subflavus (Wagner, 1842) ontological Statistics software package for education and data (Rodentia: Cricetidae), in northeastern Brazil. Zool. Anz. 233, analysis. Palaeontol. Electron. 4, 1–9. 45–55. Hershkovitz P. 1960 Mammals of northern Colombia, preliminary Bush G. L., Case S. M., Wilson A. C. and Patton J. L. 1977 Rapid report no. 8: arboreal rice rats, a systematic revision of the subgenus speciation and chromosomal evolution in mammals. Proc. Natl. Oecomys, genus Oryzomys. Proc.U.S.Natl.Mus.110, 513–568. Acad. Sci. USA 74, 3942–3946. Hershkovitz P. 1962 Evolution of Neotropical cricetine rodents Carleton M. D. and Musser G. G. 1984 Muroid rodents. In Orders (Muridae) with special reference to the phyllotine group. Fiel- and families of recent mammals of the World (ed. S. Anderson diana, Zool. 46, 1–524. and J. K. Jones Jr), pp. 289–379. John Wiley, New York, USA. Hershkovitz P. 1977 Living New World monkeys (Platyrrhini),1st Carleton M. D. and Musser G. G. 1989 Systematic studies of volume. The University of Chicago Press, Chicago. Oryzomyine rodents (Muridae, Sigmodontinae): a synopsis of Hotelling H. 1931 The generalization of Student’s ratio. Ann. Math. . Bull. Am. Mus. Nat. Hist. 191, 1–83. Stat. 2, 360–378.

276 Journal of Genetics, Vol. 91, No. 3, December 2012 Genetic and morphological variability in Oecomys

Howell W. M. and Black D. A. 1980 Controlled silver-straining of RochaR.G.,FerreiraE.,CostaB.M.A.,MartinsI.C.M.,Leite nuclear organizer regions with protective colloidal developer: a Y. L. R., Costa L. P. and Fonseca C. 2011 Small mammals of 1-step method. Experientia 36, 1014–1015. the mid-Araguaia river in central Brazil, with the description of a Johns G. C. and Avise J. C. 1998 A comparative summary of genetic new species of climbing rat. Zootaxa 2789, 1–34. distances in the vertebrates from the mitocondrial Cytochrome-b Sambrook J., Fritsch E. F. and Maniatis T. 1989 Molecular cloning: gene. Mol. Biol. Evol. 15, 1481–1490. a laboratory manual. Cold Spring Harbor Laboratory Press, Kimura M. 1980 A simple method for estimating evolutionary rates New York, USA. of base substitutions through comparative studies of nucleotide Schneider H. 2006 Métodos de análise filogenética: um guia sequences. J. Mol. Evol. 16, 111–120. prático. 3rd edition, Holos Editora, Sociedade Brasileira de King M. 1987 Chromosomal rearrangements, speciation and the Genética, Ribeirão Preto, Brasil. theoretical approach. Heredity 59, 1–6. Seabright M. 1971 The use of proteolitic enzymes for the mapping King M. 1993 Species evolution. The role of chromosome change. of structural rearrangements of man. Chromosoma 36, 204–210. Cambridge University Press, Cambridge. Smith M. F. and Patton J. L. 1991 Variation in mitochondrial Langguth A., Maia V. and Mattevi M. 2005 Karyology of large Cytochrome-b sequence in natural populations of South Amer- size Brazilian species of the genus Oecomys Thomas, 1906 ican akodontine rodents (Muridae: Sigmodontinae). Mol. Biol. (Rodentia, Muridae, Sigmodontinae). Arq. Mus. Nac., Rio de Evol. 8, 85–103. Janeiro 63, 183–190. Smith M. F. and Patton J. L. 1993 The diversification of South Larkin M. A., Blackshields G., Brown N. P., Chenna R., American murid rodents: evidence from mitochondrial DNA se- McGettigan P. A., McWilliam H. et al. 2007 ClustalW and quence data for akodontine tribe. Biol. J. Linn. Soc. 50, 149–177. ClustalX version 2. Bioinformatics 23, 2947–2948. Smith M. F. and Patton J. L. 1999 Phylogenetic relationships and Maruyama T. and Imai H. T. 1981 Evolutionary rate of the mam- radiation of Sigmodontinae rodents in South America: evidence malian karyotype. J. Theor. Biol. 90, 111–121. from Cytochrome b. J. Mamm. Evol. 6, 89–128. McDowell Jr S. B. 1958 The greater antillean insectivores. Bull. Steppan S. J. 1995 Revision of the tribe Phyllotini (Rodentia: Sig- Am. Mus. Nat. Hist. 115, 113–214. modontinae), with a phylogenetic hypothesis for the Sigmodon- Milhomem S. S. R., Pieczarka J. C., Crampton W. G. R., Silva D. S., tinae. Fieldiana, Zool. 80, 1–112. de Souza A. C. P., Carvalho Jr J. R. and Nagamachi C. Y. 2008 Sumner A. T. 1972 A simple technique for demonstrating cen- Chromosomal evidence for a putative cryptic species in the Gym- tromeric heterochromatin. Exp. Cell Res. 75, 304–306. notus carapo species-complex (Gymnotiformes, Gymnotidae). Silva M. J. J. and Yonenaga-Yassuda Y. 1998 Karyotype and chro- BMC Genet. 9, 75 (doi:10.1186/1471-2156-9-75). mosomal polymorphism of an undescribed Akodon from Central Miranda G. B., Andrades-Miranda J., Oliveira L. F. B., Langguth A. Brazil, a species with the lowest diploid chromosome number in and Mattevi M. S. 2007 Geographic patterns of genetic variation rodents. Cytogenet. Cell Genet. 81, 46–50. and conservation consequences in three South American rodents. Tamura K., Dudley J., Nei M. and Kumar S. 2007 MEGA4: Molec- Biochem. Genet. 45, 839–856. ular evolutionary genetics analysis (MEGA) software version Musser G. G. and Carleton M. D. 1993 Family Muridae. In Mam- 4.0. Mol. Biol. Evol. 24, 1596–1599. mals species of the world (ed. D. E. Wilson and D. M. Reeder) Thomas O. 1906 Notes on South American rodents. II. On the allo- 2nd edition, pp. 501–753. Smithsonian Institute, Washington, USA. cation of certain species hitherto referred respectively to Ory- Musser G. G. and Carleton M. D. 2005 Superfamily Muroidea. In zomys, Thomasomys,andRhipidomys. Ann. Mag. Nat. Hist. 118, Mammal species of the world a taxonomic and geographic ref- 442–448. erence (ed. D. E. Wilson and D. M. Reeder) 3rd edition, pp. Thomas O. 1909 New species of Oecomys and Marmosa of Ama- 894–1531. Johns Hopkins University Press, Baltimore. zonia. Ann. Mag. Nat. Hist. 3, 378–380. Nagamachi C. Y., Pieczarka J. C., Milhomem S. S. R., O’Brien Voss R. S. 1988 Systematics and ecology of ichthyomyine rodents P. C. M., de Souza A. C. P. and Ferguson-Smith M. A. 2010 (Muroidea): patterns of morphological evolution in a small adap- Multiple rearrangements in cryptic species of electric knifefish, tative radiation. Bull. Am. Mus. Nat. Hist. 188, 259–493. Gymnotus carapo (Gymnotidae, Gymnotiformes) revealed by Voss R. S. 1991 An introduction to the neotropical muroid rodent chromosome painting. BMC Genet. 11,28 genus . Bull. Am. Mus. Nat. Hist. 210, 1–113. Patton J. L. and Da Silva M. N. F. 1995 A review of the spiny mouse Voss R. S., Lunde D. P. and Simmons N. B. 2001 The mammal of genus (Rodentia: Muridae: Sigmodontinae) with the Paracou, : a Neotropical lowland rainforest fauna description of a new species from the western Amazon of Brazil. part 2. Nonvolant species. Bull. Am. Mus. Nat. Hist. 263, 1–236. Proc. Biol. Soc. Wash. 108, 319–337. Wahlert J. H. 1974 The cranial foramina of Protrogomorphus Patton J. L. and Gardner A. L. 1972 Notes on the systematics rodents: an anatomical and phylogenetic study. Bull. Mus. Comp. of Proechimys (Rodentia, Echimyidade), with emphasis on Peru- Zool. 146, 363–410. vian forms. Occ. Pap. Mus. Zool. Louisiana St. Univ. 44, 1–30. Weksler M. 2003 Phylogeny of neotropical oryzomyine rodents Patton J. L., Da Silva M. N. F. and Malcom J. R. 2000 Mammals (Muridae: Sigmodontinae) based on the nuclear IRBP exon. Mol. of the Rio Juruá and the evolutionary and the ecological Phylogenet. Evol. 29, 331–349. diversification of Amazonia. Bull. Am. Mus. Nat. Hist. 244, Weksler M. 2006 Phylogenetic relationships of oryzomine rodents 1–306. (Muroidea: Sigmodontinae): separate and combined analyses of Pocock R. I. 1914 On the facial vibrissae of Mammalia. Proc. Zool. morphological and molecular data. Bull. Am. Mus. Nat. Hist. 296, Soc. London 84, 889–912. 1–149. Posada D. and Crandall K. A. 1998 Modeltest: testing the model of Weksler M., Percequillo A. R. and Voss R. S. 2006 Ten new genera DNA substitution. Bioinformatics 14, 817–818. of Oryzomyine rodents (Cricetidae, Sigmodontinae). Am. Mus. Reig O. A. 1980 A new fossil genus of South American cricetid Novit. 3537, 1–29. rodents allied to Wiedomys, with an assessment of the Sigmodon- Xia X. and Xie Z. 2001 DAMBE: Software package for data analy- tinae. J. Zool. 192, 257–281. sis in molecular biology and evolution. J. Hered. 92, 371–373. Received 20 December 2011, in revised form 16 March 2012; accepted 1 June 2012 Published on the Web: 23 October 2012

Journal of Genetics, Vol. 91, No. 3, December 2012 277