One male and one female were collected with pitfall traps in Santa Virgínia, Parque Estadual da Serra do Mar [45°03.00' to 45°11.00'W (DDM); 23°24.00' to 23°17.00'S (DDM)], state of São Paulo, Brazil. Pelage color and external measurements were taken during the fieldwork. Vouchers of both individuals are deposited in the Coleção de Mamíferos da Universidade Federal do Espírito Santo (UFES) under the catalog numbers UFES 2271 and UFES 2272.

Metaphases were obtained in vivo from spleen and bone marrow, according to Ford and Hamerton (1956) with modifications. Conventional Giemsa staining was used to determine the diploid (2n) and the number of autosome arms (FN). GTG and CBG-banding were performed according to Seabright (1971) and Sumner (1972), respectively, with modifications. Fluorescent in situ hybridization (FISH) with a FITC labeled (C₃TA₂) _npeptide nucleic acid (PNA) probe (DAKO) was carried out following the recommended protocol (Telomere PNA FISH Kit/FITC, Code No. K5325, DAKO). Mitotic plates were digitally captured with visible light or blue and green filters (emission at 461 and 517 nm, respectively) in an Axioskop 40 epifluorescence microscope (Carl Zeiss) equipped with an Axiocam camera and AxionVision software. Images were overlaid and contrast enhanced with Adobe Photoshop CS5.1.

DNA was extracted from liver with Chelex 5% (Bio-Rad) following Walsh et al. (1991). Amplification of an 820 bp fragment of Cyt b and a 782 bp of IRBP was performed with PCR using primers MVZ5 and MVZ16 (Irwin et al. 1991; Smith and Patton 1993), and A1 and F (Stanhope et al. 1992), respectively. Both extraction and PCR controls were used for each amplification. Each PCR mixture had 30 ng of DNA, 25 pmol of each primer, 0.2 mM of dNTP, and 2.52 µL of reaction buffer (50 mM KCl, 2.5 mM MgCl2, 10 mM Tris-HCl; pH 8.8), and 0.2 units of Taq DNA polymerase (Invitrogen) were added to complete 18 µL. Forty amplification cycles were performed in a thermal cycler (Eppendorf Mastercycler ep Gradient, Model 5341). Each cycle consisted of denaturation at 94°C for 30 s, annealing at 48°C for 45 s, and extension at 72°C for 45 s for Cyt b, and denaturation at 94°C for 30 s, annealing at 60°C for 60 s, and extension at 72°C for 180 s for IRBP. A final extension at 72°C for 5 min was performed for both Cyt b and IRBP amplifications. The PCR products were separated using 1% agarose gels in TAE buffer. Nucleotide sequencing was conducted using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems). Sequences of each animal were aligned with sequences from previously published data deposited on GenBank by Bonvicino and Moreira (2001), Weksler (2003), and Percequillo et al. (2011) using MAFFT ver. 5 (Katoh et al. 2005) under the iterative method of global pairwise alignment (G-INS-i). Our sequences were submitted to GenBank under accession numbers KF031014-KF031017.

We performed maximum likelihood (ML) and Bayesian analyses using concatenated Cyt b-IRBP data set. For both analyses we used gene-specific unlinked models. The best-fitting model of nucleotide substitution for each gene was selected using the Akaike information criterion in accordance with the procedure outlined by Posada and Buckley (2004), and implemented in jModelTest, version 0.1.1 (Posada 2008). The maximum-likelihood trees were calculated using RAxML (Stamatakis 2006). The statistical support for the nodes was estimated by the nonparametric bootstrap, with 1000 pseudoreplicates (Felsenstein 1985). Bayesian analysis was performed using MrBayes 3.04b (Ronquist and Huelsenbeck 2003). Markov chains were started from a random tree and run for 1.0 × 10⁷ generations, sampling every 1000th generation. The stationary phase was checked following Nylander et al. (2004). Sample points prior to the plateau phase were discarded as burn-in, and the remaining trees were combined to find the maximum a posteriori probability estimated of the phylogeny. Branch support was estimated by Bayesian posterior probabilities (BPP). Two simultaneous analyses were performed to ensure convergence on topologies.

The animals showed 2n=62, FN=62, and the autosome set composed of 29 acrocentric pairs decreasing in size, and one small metacentric pair (Fig. 1A). The X is a large submetacentric, and the Y is a large submetacentric slightly smaller than the X (Fig. 1A). CBG-banding revealed pericentromeric constitutive heterochromatic blocks in all autosomes and in the long arm of Y (Fig. 1B). GTG-banding allowed the identification of almost all autosomic pairs, the X chromosome exhibited two interstitial bands at the long arm while a conspicuous pattern in the Y was not found (Fig. 1C). FISH detected telomeric signals at the ends of all chromosomes and additional telomeric sequences were found in the pericentromeric region of both X and Y chromosomes (Fig. 1D).

The model selected for the phylogenetic analyses (ML and Bayesian) was GTR +I + Γ for each gene. The best ML tree had a -ln likelihood score of -22, 345.02. The Bayesian analysis recovered a consensus topology similar to the best ML tree and the results recovered the four well-supported clades A, B, C, and D (Fig. 2) previously reported by Weksler (2006) and Percequillo et al. (2011). In both phylogenetic analyses, Santa Virgínia specimens (UFES 2271 and UFES 2272) clustered with high statistical support to the recently described Drymoreomys albimaculatus (Fig. 2, grey area).

The specimens collected were medium sized (male body mass: 46.5 g, head and body length: 115 mm, and tail length: 142 mm; female body mass: 57 g, head and body length: 127 mm, and tail length: 170 mm). Tail was longer than head and body, and was a uniform color on both sides. Male hind footwas short (25 mm, 22% of head and body length) and ears were small (16 mm; 14% of head and body length). These external morphological measures overlapped with those of the Drymoreomys albimaculatus holotype (Percequillo et al. 2011). Dorsal pelage was reddish-brown; ventral pelage was predominantly grayish. Samples exhibited the pattern of short hind feet consistent with Oecomys. Fore and hind feet digits were covered by silvery-white hairs and the dorsal surface of hind feet were covered by brown hairs forming a patch, in a more conspicuous pattern than the observed for Rhipidomys. Thus, some external morphological traits were similar to those described for Rhipidomys and Oecomys as reported Percequillo et al. (2011). Nevertheless, our samples exhibited the characteristics of the Drymoreomys albimaculatus holotype that differentiate it from Rhipidomys, such as the shorter, thinner, and sparser mystacial vibrissae and presence of gular to pectoral patches of white hair. Additionally, we detected that, contrary to what is found in Rhipidomys, a tuft of hairs on the tail’s end is absent in our samples. On the other hand, several anatomical traits that distinguish the Drymoreomys albimaculatus holotype and species of Oecomys wereobserved in our samples, such as the plantar surface of pes covered with squamae; dorsal surface of pes with dark patches of brown hairs and the ventral pelage with gular and thoracic white patches (Percequillo et al. 2011).

Percequillo et al. (2011) reported that most of the Drymoreomys specimens were collected in pitfall traps; in the present work, the animals were also collected in the same way. These reiterate the importance of further fieldwork effort, with different collecting methods in order to increase the spectrum of small mammals collected. Consequently, our knowledge of small mammal biodiversity will be improved as a whole, which will allow improvements in relevant laws and policies for biodiversity protection.