Range extension of the Atlantic Forest Hocicudo , Oxymycterus dasytrichus ( Schinz , 1821 ) , to the state of Santa Catarina , southern Brazil

Six individuals of Oxymycterus dasytrichus (Schinz, 1821) were found on the coast of Paraná and Santa Catarina (in the Atlantic Forest), expanding the known geographical distribution of the species ca. 280 km southward. The specimens represent the first record of the species for the state of Santa Catarina, and new localities to the region of southern Paraná. The identification of the species relied mainly on interspecies comparative assessment of genetic distance based on DNA sequences data from the mitochondrial cytochrome-b gene and geographic distribution of taxa across biomes, particularly in the Atlantic Forest. Our findings highlight the role of protected areas, particularly the Guaraqueçaba Environmental Protection Area, to preserve small mammals.

The Neotropical sigmodontine rodent genus Oxymycterus Waterhouse, 1837 is diverse, with 15 recognized species (Oliveira and Gonçalves 2015), and geographically widespread, from central Argentina and southern Uruguay to the Amazon basin (Hershkovitz 1994).The species boundaries, phylogenetic relationships, and distribution remain poorly understood (Jayat et al. 2008).Several recent studies have significantly advanced the understanding of the taxonomy and phylogeny of Oxymycterus (e.g., Hoffmann et al. 2002;Gonçalves and Oliveira 2004;D'Elía et al. 2008;Jayat et al. 2008;Oliveira and Gonçalves 2015).However, distinct aspects of the geographic distribution of the species are still in question, particularly in highly diverse biomes such as the Atlantic Forest.
Oxymycterus dasytrichus (Schinz 1821), commonly named the Atlantic Forest Hocicudo (Bonvicino et al. 2008), is widely distributed in Dense Ombrophilous Forest along the Brazilian coast, in the states of Pernambuco, Sergipe, Alagoas, Bahia, Espírito Santo, Minas Gerais, Goiás, Rio de Janeiro, and São Paulo, from the coastal lowlands to altitudes of around 2,000 m in Itatiaia, Rio de Janeiro (Oliveira 1998;Hoffmann et al. 2002;Gonçalves and Oliveira 2004;Oliveira and Gonçalves 2015).Oliveira and Gonçalves (2015) also recorded this species to the state of Paraná, although they do not list specific localities.The type locality was restricted to the Rio Mucuri in Bahia (Avila-Pires 1965).Gonçalves and Oliveira (2004) used integrative molecular and morphological techniques to study this species over most of its distribution in the Atlantic Forest, and their results suggested the existence of a morphological differentiation along a north-south axis.
The species is currently classified as of Least Concern (LC) according to the IUCN Red List of Threatened Species, due to its wide distribution and high population densities (Bonvicino et al. 2008).
The present study expanded the distributional range of O. dasytrichus (Figure 1) along the Atlantic Forest, to ca. 280 km south of the nearest previously recorded locality, at Cananéia, São Paulo.The four new localities are on the coast of Paraná (PR) and Santa Catarina (SC) (Figure 2; Table 1).The climate is humid subtropical (Köppen Cfa) (Kottek et al. 2006); the mean annual  rainfall on the coast of Paraná is 2,435 mm, and on the north coast of Santa Catarina is 1,690 mm, with mean annual temperatures of 21 and 20.2°C, respectively (Vanhoni and Mendonça 2008;Araujo et al. 2006) Serra da Graciosa, the Guaraqueçaba APA is part of the Biosphere Reserve.The area adjacent to this sample locality also covers the Guaraqueçaba Ecological Station, including Superagüi National Park, and the Salto Morato and Sebuí Private Natural Heritage Reserves.Two males (UFPR-P752, UFPR-P754) and one female (UFPR-P760) were collected.II) Guaraqueçaba-Massarapuã, PR (25°19ʹ S, 048°26ʹ W, 10 m a.s.l.), on a peninsula in the Guaraqueçaba APA.One male (UFPR-P757).III) Antonina, basin of the Rio Nunes, PR (25°25ʹ S, 048°42ʹ W, 5 m a.s.l.) situated at sea level adjacent to the Serra do Mar on the coast of Paraná.One female (UFPR-P177).Gonçalves and Oliveira (2004); 3, Oliveira and Gonçalves (2015).Geographical coordinates from 2 and 3 were estimated based on the municipality and/or location available.
Specimens collected on the PR coast were caught in a Tomahawk trap, baited with corn and pineapple, in two field trips, one in June 1986 (UFPR-P177) and the other in April 1993 (UFPR-P752, UFPR-P754, UFPR-P757 and UFPR-P760).The single specimen captured on the northern coast of SC (FURB 18669; Figure 3) was caught in a Sherman trap baited with corn, in September 2011.These specimens are deposited in scientific collections at the Universidade Federal do Paraná (UFPR) and the Universidade Regional de Blumenau (FURB).Individuals were measured, tissues (muscle, kidney, liver and heart) were preserved in 96% ethanol at -20°C, and the skull and skin were prepared.
Total genomic DNA was purified from tissue samples of the six O. dasytrichus specimens using a PureLink DNA Extraction Kit (Invitrogen) following the manufacturer's instructions.A partial sequence (801 base pairs, bp) of the cytochrome-b (cyt-b) gene was amplified by polymerase chain reaction (PCR) using ca.50 ng of DNA per sample, with primers and conditions described by Smith and Patton (1993).The PCR products were purified using the enzymatic method (Exonuclease I and Thermosensitive Alkaline Phosphatase, Thermo Scientific) and sequenced on an ABI 3700 sequencer (Applied Biosystems).Sequence data obtained were deposited in GenBank under accession numbers KU161271-KU161276 (Table 2).
In order to identify the genealogical position of the newly sequenced haplotypes of O. dasytrichus, we incorporated 11 sequences from southern Brazil (considered here as a reference for the species) provided by Hoffmann et al. (2002) and Gonçalves and Oliveira (2004) (Table 2).Ten other representative species of the genus were incorporated in the analysis (obtained from GenBank; Table 2) to discuss relationships of the species within the genus, using taxa with occurrence in Brazilian biomes, particularly the Atlantic forms, based on Gonçalves and Oliveira (2004) phylogenetic hypothesis.All data were aligned in the program MUSCLE (http://www.ebi.ac.uk/tools/msa/muscle) and manually adjusted in the software MEGA 6 (Tamura et al. 2013).The best-fit model of nucleotide substitution was determined based on the Akaike Information Criterion.Bayesian inference was used to reconstruct the phylogenetic relationships using BEAST 2.02 (Drummond et al. 2012) under the HKY model with a proportion of invariant sites and a gamma distribution.All parameters were estimated from the data.The Markov chains were run for 10 million generations and repeated four times to test for Markov chain-Monte Carlo chain convergence, and priors exceeded 200 to ensure effective sample sizes (ESS).Burn-in was determined in Tracer v1.5 (Drummond and Rambaut, 2007) based on parameter trajectories, and 10% of the first trees were then removed in TreeAnnotator.The consensus tree was visualized and edited in FigTree 1.3.1 (Rambaut 2009).Nodes with posterior probability (PP) ≥ 95% were considered to be supported (Alfaro et al. 2003).In order to characterize the genetic divergence between O. dasytrichus collected in PR and SC and those used as a reference (from southeast Brazil), pairwise genetic distances were calculated using the Kimura 2-parameters (K2P) model with 1000 replications boot straps.
Taxonomy within Oxymycterus has been controversial because of difficulties in detecting discrete morphological and cytogenetic variation (Musser and Carleton 1993;Hershkovitz 1994;Oliveira 1998; Gonçalves and Oliveira 2004).Therefore, the identification of the six vouchers was based on some external morphology characters (such as coat color and skull) coupled with comparative assessment of DNA sequences (the mitochondrial gene cytochrome b) and distributional ranges.
Oxymycterus dasytrichus presents a dark-brown to paler brownish red without strong lining of back in the dorsum and dark-gray to paler, cinnamon or pinkish buff in the ventral (Oliveira and Gonçalves, 2015).Its skull is robust, proportionally larger in width dimensions of braincase, interorbital region and zygomatic plates, with longer molar tooth rows (mean 5.6 mm) comparably to other species of Atlantic Coast (Figure 4).Its variation in cranial measurements was reported by Gonçalves and Oliveira (2004).On the other hand, some species predominantly occurs in certain biomes; for example, the Atlantic Forest in the case of O. dasytrichus.
In the absence of discrete characters in O. dasytrichus, the identification of specimens relied mostly on lesser genetic distance associated to geographically distinct groups, using the nucleotide BLAST tool from the  NCBI database.Species presenting less than 5% of divergence collected in the same biome and assigned to O. dasytrichus by Gonçalves and Oliveira (2004), were used to attribute our samples to this taxon.Although some taxonomic mistakes are known to occur in the Genbank records, the consistency of close relationship to a large number of specimens collected along the Atlantic Forest previously identified as O. dasytrichus by experts in this group (using comparative morphology as well; Gonçalves and Oliveira 2004) led us to consider reliable such result.
The results from Bayesian inference analysis (Figure 4) indicated that the samples collected along the PR and SC coast belong to the O. dasytrichus group, because the haplotypes corresponding to the newly collected specimens from these sites are nested in the major clade formed by those previously referred as this species.Thus, our identification of the newly reported specimens relies on lesser divergence comparative to other species of the genus and diagnose made by Gonçalves and Oliveira (2004) of the specimens they reported from the Atlantic Forest, based on morphological and molecular data.Because no morphological diagnostic characters to identify species in this genus are known, in order to identify specimens we had to trust on comparisons based on genetic distances and morphology with previously identified specimens.Two internal haplogroups were recovered for O. dasytrichus: clade 1 formed by eight specimens/haplotypes (six from this study, and two from São Paulo and clade 2 formed by nine specimens/ haplotypes from Rio de Janeiro, Espírito Santo, Minas Gerais and São Paulo.Whereas clade 1 received strong support, clade 2 was not supported.The K2P mean genetic distance between clades 1 and 2 was 2.1%, whereas that between species ranged from 6 to 12%. Our data indicate that discoveries on general aspects of the biology (e.g., distributional range) of the genus Oxymycterus can still be made, even in a well-studied biome as the Atlantic Forest.Hoffmann et al. (2002) reported an intraspecific distance of 1.3% for the species across its northern and southern Atlantic Forest distribution.In our study we found similar divergence (1.4%), suggesting a shallow structure among haplotypes of clades 1 and 2. Such divergence does not seem to be determined strongly by a geographic cline since both clades include specimens from São Paulo.Thus, we did not find evidence to suggest that the populations represented by Clade 1 and Clade 2 are not conspecific.On the other hand, O. dasytrichus also occurs in the Caatinga and Cerrado (Costa and Leite 2012); therefore, such populations deserve particular effort in terms of comparative morphology and genetics since there is a greater distance to the Atlantic Forest populations and higher fixation of differences, and/or local adaptation to these distinct biomes, might exist.
The Atlantic Forest (lato sensu) coverage was reduced to 11-16% of the original area, in which 9% is protected areas, and 1% represents the original forest (Ribeiro et al. 2009).However, it is estimated that 1-8% of all the biodiversity in the planet is within this area (Silva and Casteleti 2003).According to Kauano et al (2012) the northern coast of Paraná is likely one of the least fragmented areas of the Atlantic, and is inserted in the "Serra do Mar", an important corridor of fauna and flora.Three of our four records were sampled in this region, which reinforces the significant role of protected areas, particularly the APA of Guaraqueçaba, not only to save small mammals, but as hotspot of endemic biodiversity refugee of the Atlantic Forest.

CheckFigure 2 .
Figure 2. Oxymycterus dasytrichus distribution, primarily along the Atlantic Forest biome in Brazil.Stars indicate the locations of four new records from the Paraná and Santa Catarina coast.Numbers from I to IV are indicated in Materials and Methods.
. The new recording localities are: I) Guaraqueçaba-Índios, PR (25°21ʹ S, 048°26ʹ W, 10 m above sea level [a.s.l.]), on the northern coast of Paraná, flanked by Paranaguá and Laranjeiras bays and the Serra do Mar coastal mountain range.The Guaraqueçaba Environmental Protection Area (APA) includes the largest continuous remnant of the Atlantic Forest on the Brazilian coast.Together with Vale do Ribeira and Peçanha et al. | Range extension of Oxymycterus dasytrichus

Table 1 .
Locality records for Oxymycterus dasytrichus, including four new sites presented in this study.