2.1.Species and investigating area

In this work, 44 specimen of O. nigripes and 2 of O. flavescens coming from the rivers Alpestre (27° 04′49″S-53° 01′51″W), Rio dos Índios (27° 18′ 00″ S-52° 50′ 27″ W), Erechim (Horto Florestal, 27° 38′ 03″ S-52° 16′ 26″ W) and Mato Castelhano (FLONA Passo Fundo, 28° 16′ 42″ S-52° 11′ 30″ W), in Rio Grande do Sul state, and from Chapecó (border of Uruguay river, 27° 05′ 04″ S-52° 37′ 06″ W), in Santa Catarina state was used (Figure 1). One Euryoryzomys russatus individual was included in the analysis for external comparison, collected in Erechim (Horto Florestal, 27° 38′03″ S-52° 16′ 26″ W).

Figure 1. Geographical location of the collecting process (Passo Fundo, Erechim, Rio dos Índios and Alpestre at RS and Chapecó at SC). 

The specimens submitted to karyotyping are deposited in the Regional Museum of Alto Uruguay (MuRAU – ZOO 113 and 114 to O. nigripes, ZOO 115 and 116 to O. flavescens), Universidade Regional Integrada, Erechim (RS). The specimens collected in Chapecó (SC) and Alpestre (RS), were kindly supplied by the Collection of Mammals of the Department of Ecology and Zoology of Universidade Federal de Santa Catarina, Florianópolis (SC), and Scientific Collection of Universidade Regional de Blumenau (FURB), Blumenau (SC), respectively.

Collecting at Chapecó (SC) city was performed at the border of Uruguay River, and in the cities of Rio do Índios and Alpestre, it was carried out at the opposite border side, in Rio Grande do Sul state. Besides such data, samples were taken from specimens belonging to Erechim and Mato Castelhano, with these populations much further from each other compared to those next to the Uruguay River (75 and 140 Km, respectively). These captures and collections are in accordance with license number 15224-1 - IBAMA/ICMBio.

2.2.Karyotype cytogenetic analysis

Two specimens of O. nigripes and two from O. flavescens were submitted to karyotyping. The metaphases were obtained from direct preparation of osseous marrow, according to ^{Ford and Hamerton (1956)} and ^{Guerra and Souza (2002)}. Metaphase preparations were obtained from bone marrow, after in vivo injection of a 0.1% colchicine solution (1mL/100g of body weight). Cells were suspended in a 0.075M KCl solution for 30 min at 37 °C, fixed in 3:1 methanol: acetic acid, spread onto clean slides and air-dried. The resulting material was colored in buffer phosphate-Giemsa solution, at the ratio of 3:1, during 10 min, washed with distilled water, and left dry at ambient temperature. At least 20 metaphases per individual were analysed to define the diploid and fundamental numbers. The karyotype were determined on the base of the works of ^{Sbalqueiro et al. (1991)}, ^{Andrades-Miranda et al. (2001)}, ^{Paresque et al. (2004)} and ^{Pereira and Geise (2007)}.

2.3.DNA extraction and molecular analysis

Tissue samples were obtained from an ear fragment of the individuals and the genetic material was extracted following the protocol described by ^{Medrano et al. (1990)}, replacing NaCl by chloroform:isoamyl alcohol 24:1. Amplification of RAPD was carried out as described by ^{Levi (1993)}, with some modifications: reaction buffer (50 mM Tris-HCl pH 9.0; 50 mM KCl; 0.5% Triton-X 100), dNTPs (200mM of each), 0.2 µM of primer, 3mM of MgCl₂, 80ng of DNA and 1.5 U of Taq DNA polymerase. Amplification was accomplished in a thermocycler (MJ Research Inc.). The amplification process was as follows: 3 min at 92 °C, 40 cycles of 1 min at 92 °C, 1 min at 36 °C and 2 min at 72 °C. Then, after 3 min at 72 °C cooling at 4 °C up to samples taking. The electrophoretic separation was performed in agarose gel at 1.4% TBE (trisma, boric acid and EDTA) buffer in a horizontal electrophoresis bath. As a molar mass marker, Lambda DNA cleaved with HindIII and EcoRI restriction enzymes was adopted. Fragment visualisation was carried out using ethidium bromide, observation performed in ultraviolet light and gels were digitally photographed. The primer belonging to seven kits from Operon Technologies was selected, with duplicate analysis, observing the intensity and reproducibility of the bands.

2.4.Statistical analysis of experimental data

Genetic variability was evaluated taking two data matrices, one of them built with 13 individuals identified as HO-15, O. flavescens, O. flavescens, O. nigripes, O. nigripes, IN-41, IN-42, IN-43, IN-39, CH-8, CH-11, FL-28 and HO-24 (HO - Horto Florestal, Erechim-RS; IN - Rio dos Índios-RS; CH - Chapecó-SC; FL - FLONA Mato Castelhano-RS), including 2 O. nigripes and 2 O. flavescenskaryotyped. The second evaluation was carried out considering 45 individuals with 44 O. nigripes (CH - Chapecó-SC; HO - Horto Florestal, Erechim-RS; FL - FLONA Mato Castelhano-RS; AL - Alpestre-RS; IN - Rio dos Índios-RS) and 1 Euryoryzomys russatus(Horto Florestal, Erechim-RS).

Experimental data were obtained through determining presence or absence of 205 bands in the first case and 211 bands for the second, which comprised two matrices analysed by the computational code MVSP 3.1. The similarity dendrograms were built by the UPGMA algorithm (Unweighted Pair Group Method Using Arithmetic Averages), making use of the Jaccard similarity coefficient.

Genetic results were evaluated taking into account genetic diferences among individuals analysed, comparing, in the first case the karyotyped specimen with others and among them. In the second case, genetic differences were considered among populations, mainly for the Chapecó individuals with others and among them, in order to assess the genetic variability and to check whether the Uruguay River acts as a geographical boundary.

3.1.Karyotype analysis

The two karyotyped specimen of Oligoryzomys nigripespresented 2n = 62; NA = 82, due the variation in autosome fundamental number due to pericentric inversions (^{Yonenaga et al., 1976}; ^{Myers and Carleton, 1981}). The Oligoryzomys flavescensshowed 2n = 64; NA = 66, because the variation in diploid and autosome fundamental number which is related to the acrocentric or metacentric B chromosomes (^{Sbalqueiro et al., 1991}). For O. nigripes, the largest chromosomes showed submetacentric arms, while for O. flavescens, the largest chromosomes are acrocentric, as already described by ^{Sbalqueiro et al. (1991)} and ^{Almeida and Yonenaga-Yassuda (1991)}.

^{Paresque et al. (2007)}, in studying the karyotype variability of 85 specimens of O. nigripes, found 2n = 62 for 84 individuals and only one specimen, captured in the Espírito Santo state, presented 2n = 61, due to a monosomy in the X chromosome. ^{Andrades-Miranda et al. (2001)}, found similar results for most specimen studied, with 2n = 62 and only one with 2n = 61, also due to a monosomy in the X sexual chromosome. The same authors found a variation of 2n = 64 to 67 for O. flavescens, a fact attributed to the presence of 0-3 accessories chromosome, described as B- chromosome.

3.2.Genetic analysis among karyotyped and non-karyotyped individuals

The primers selection revealed that OPA 19, OPB 8, OPF 3, OPF 11, OPF 12, OPF 14, OPW 1, OPW 9, OPW 17, OPW 19, OPY 6, OPY 16, OPY 19 and OPY 20 provided the best amplification results among samples analysed, then used for the amplification reactions of 13 individuals studied in the first analysis (HO-15, O. flavescens, O. flavescens, O. nigripes, O. nigripes, IN-41, IN-42, IN-43, IN-39, CH-8, CH-11, FL-28 and HO-24 (HO - Horto Florestal Erechim-RS; IN – Rio dos Índios-RS; CH – Chapecó-SC; FL – FLONA Mato Castelhano-RS). Considering the 13 individuals analysed, a total of 205 loci was identified, with 180 (87.8%) being polymorphic. ^{Trott et al. (2007)} evaluated 75 polymorphic loci in species of the genus Oligoryzomys. The Jaccard similarity coefficient varied from 0.36 to 0,74 for the specimens analysed (Figure 2).

Figure 2. Dendrogram of 13 individuals of Oligoryzomysanalysed by RAPD, using the Jaccard coefficient and UPGMA grouping technique (IN - Rio dos Índios-RS; CH – Chapecó-SC; HO - Horto Florestal Erechim-RS; FL - Flona Mato Castelhano -RS). 

The UPGMA grouping analysis and Main Coordinate Analysis technique (MCA) allowed separating two distinct groups, as can be seen in Figure 3. The first group corresponds to the two karyotyped specimens identified as Oligoryzomys flavescens, and the second one represents the other individuals, including those karyotyped defined as Oligoryzomys nigripes, a clear grouping between the members of this group not occurring with regard to the collecting area. This result indicates the capacity of the RAPD molecular markers in separating these two species, a result also obtained by ^{Trott et al. (2007)}, allowing the identification of the individuals analysed in the second part of this work. Besides, a clear separation was noted for the individual Euryoryzomys russatus, collected at Horto Florestal Erechim-RS, and included in the analysis as external standard.

Figure 3. Main Coordinate Analysis of 13 individuals of Oligoryzomys (IN - Rio dos Índios-RS; CH - Chapecó-SC; HO - Horto Florestal Erechim-RS; FL - Flona Mato Castelhano -RS). 

^{Trott et al. (2007)}, in investigating RAPD markers of the genus Oligoryzomysdistributed in different biomes of Brazil, used a total of 75 polymorphic products simultaneously amplified in 151 specimens and observed that there is a great genetic variation among the taxa of Oligoryzomysinvestigated. However, they also noted that among the species of this genus, O. nigripes and O. flavescens presented the smallest genetic distance.

Although there is not an a priori hypothesis on how the species of a genus may arrange themselves, taxa investigated grouped, apparently, by their karyotype and to a certain extent by the geographical area occupied. There was a separation among karyotyped individuals, proven to be O. flavescens and O. nigripes, which corroborates the fact that the molecular markers based on RAPD are capable of genetically separating these two species. Such results are also in agreement with those found by ^{Trott et al. (2007)}, because these authors were able to distinguish 6 species and 13 populations using RAPD markers, thus showing that the taxa of Oligoryzomys are genetically different.

3.3.Population genetics analysis

Of the 45 individuals analysed, 12 were captured at Horto de Erechim-RS, 5 from Rio dos Índios-RS, 14 at Chapecó-SC, 12 at Flona in Mato Castelhano-RS and 2 form Alpestre-RS. From the 45 individuals investigated, a total of 211 loci were identified, 181 (85.78%) being polymorphic (Table 1 and 2). The population that presented the greatest number of loci was Chapecó (192), but the percentage of polymorphic bands was relatively different and smaller only for the Alpestre population (142 bands, 29.58% polymorphic), which may be explained in terms of the small sampling for this place.

The Jaccard similarity coefficient varied in the range of 0.45 to 0.87 for the populations analysed (see Figure 4). In this sense, a trend of individuals grouping was observed at each place, mainly regarding the population of Rio dos Índios, which, although formed by few specimens, most of them were very close to each other.

Figure 4. Dendrogram of the individuals analysed by RAPD, using the Jaccard coefficient and UPGMA grouping technique (CH - Chapecó-SC; HO - Horto Florestal Erechim-RS; FL - Flona Mato Castelhano-RS; AL - Alpestre-RS; IN - Rio dos Índios-RS). 

This trend can be clearly verified from the dendrogram produced by the Jaccard similarity coefficient. One group corresponds to almost the entire population of Rio dos Índios (80% of individuals), lacking only one specimen. The other groups showed a smaller trend to population grouping, with individuals distributed throughout the populations, though 58.3% of specimen for Horto, 66.6% at Flona and 57.1% for Chapecó were found grouped. The individuals of Euryoryzomys russatus, included in the analysis as an external standard, kept a clear separation, even with a greater number of individuals analysed.

It is worth noticing the lack of genetic proximity between the populations of Chapecó (Santa Catarina state) and Alpestre and Rio dos Índios-RS, due to the geographical immediacy, despite being separated by the Uruguay River. Note also that the individuals of Alpestre are not genetically similar to those of Rio dos Índios, in spite of the geographical nearness of these two places.

The RAPD analysis performed in this work revealed a significant genetic heterogeneity among the taxa helping discriminate among the species of Oligoryzomys, a genus considered until now, to be composed of several little different species. Nevertheless, genetic relationships among populations of Oligoryzomys coming from such data seem not to be dependent on geographical distances nor influenced by rivers acting as reproductive barriers.

^{Trott et al. (2007)}demonstrated that several diversity estimations presented considerable differences among the species and populations, hence pointing to a great genetic variety among taxa of Oligoryzomys investigated. The grouping analysis using the genetic distance method by Nei, however, did not correlate the genetic heterogeneity of species and populations with the geographical areas, corroborating the results found in this work. Thus, it is possible that the relationships presented may be a result of a specific biology history of each species, like the population dimension in the past or the different strategies developed to explore and occupy the actual places (^{Trott et al., 2007}). Furthermore, it was observed that rivers, like the Uruguay River, might not be a barrier for individuals migration and accordingly for genetic flux among populations. Thus, one should expect that the genetic similarity among individuals would be greater, with unclear separation among populations according to their occurrence regions.