Introduction
Euphractus sexcinctus, the yellow or six-banded armadillo, is the largest of the five species of euphractine armadillos, also known as the hairy armadillos (Wetzel, 1985; Eisenberg and Redford, 1999). Yellow armadillos are distinguished by short ears and a flattened head that becomes triangular toward the snout, protected by large plates with patchy fur (Nowak, 1999). The upper body is light yellow, bronze or red-dish, with six to eight mobile bands on the carapace. An adult animal's head and body measure approximately 401–495 mm, and it weighs 3.2–6.5 kg. Its tail is short and cylindrical, with plates arranged in two to four separate bands at the base. All five toes on each paw have claws, the second of which is the longest (Nowak, 1999).
The geographic distribution of E. sexcinctus covers much of eastern South America, ranging from the southern mouth of the Amazon through all of southeastern Brazil, and extending into Uruguay, Paraguay and northeastern Argentina, as well as eastern and central Bolivia (Wetzel, 1985; Redford and Wetzel, 1985; Emmons and Feer, 1997; Eisenberg and Redford, 1999). E. sexcinctus also occurs in the savannas of Sipaliwini in Suriname, and of Paru in the Brazilian state of Pará. Together the records in these two savanna regions constitute what has been thought to be a disjunct population, separated by hundreds of kilometers from the main distribution to the southeast (Silva Júnior and Nunes, 2001).
Silva Júnior et al. (2001) recorded E. sexcinctus from 27 localities in the state of Maranhão, between the Rios Gurupí and Parnaíba, in the region known as “Pré-Amazônia Maranhense.” From northernmost Brazil, Silva Júnior and Nunes (2001) added four additional localities for this species in the state of Amapá, immediately to the north of the mouth of the Rio Amazonas. It is important to note that these records by themselves do not confirm a continuous distribution of E. sexcinctus in the Brazilian states of Amapá, Pará, Maranhão and Piauí. Here we record this species from the region between the Rios Tocantins and Gurupí in Pará, and suggest its continuous distribution in the Brazilian Amazon.
Methods
Data collection and morphological characterization
We identified E. sexcinctus from animals that had been hunted by local people in the municipalities of Bragança, Ourém, Augusto Corrêa, and Vizeu in the state of Pará, and in Bocaina in the state of Piauí (Table 1). Local residents donated the specimens during interviews carried out when we were conducting mammal surveys in this region. From the six animals donated to us, it was only possible to preserve the carapace and skull of a single specimen, now in the Zoological Collection of the Campus of Bragança in Bragança, Pará, under the field number 196PA. Only blood and muscle tissue could be collected from the other five specimens. We identified these individuals from the information provided in Emmons and Feer (1997), Eisenberg and Redford (1999), Nowak (1999), and Silva Júnior and Nunes (2001).
TABLE 1.
New locality records for Euphractus sexcinctus. Mesohabitat: 1 = clear-cut region, located between an urban area and mangrove; 2 = small fragments of secondary forest and salt marshes associated with mangroves; 3 = clear-cut region with small fragments of secondary forest; 4 = Caatinga region.
Molecular characterization
We used molecular markers to confirm the morphological identification of E. sexcinctus and examine intraspecies similarities of the sampled animals. We chose Dasypus novemcinctus and Cabassous unicinctus as outgroups, and obtained the sequences from a sample collected in the state of Pará and another sample supplied by GenBank, respectively. The GenBank accession numbers for C. unicinctus are AF232016 (cytochrome b) and Z48940 (16S rRNA).
For the molecular characterization, two mitochondrial genes with different rates of evolution were chosen: cytochrome b (protein coding gene) and 16S rRNA (ribosomal RNA). Genomic DNA was extracted from small quantities of blood or ear tissue of the collected specimens. DNA extraction was performed in accordance with the conventional phenol-chloroform extraction protocol modified from Sambrook and Russell (2001).
Fragments of nearly 600 base-pairs of both 16S and cytochrome b genes were isolated by PCR (Polymerase Chain Reaction), using the oligonucleotides described by Palumbi et al. (1991) and Smith and Patton (1993), respectively. DNA was sequenced using the ABI Prism™ Dye Terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems, USA), on an ABI 377 (Perkin Elmer) automated sequencer. PCR and sequencing procedures were conducted in accordance with those described by Barros et al. (2003).
Phylogenetic analysis
Sequences were aligned and edited with BIOEDIT (Hall, 1999) and ClustalX (Thompson et al., 1997). The phylogenetic analyses were performed with a series of programs, including DAMBE (Xia and Xie, 2001), Modeltest (Posada and Crandall, 1998), MEGA2 (Kumar et al., 2001) and PAUP* (Swofford, 1998). Cluster significance was estimated by Boot-strap analysis (Felsenstein, 1985).
Results
The nucleotide sequences of the DNA fragments were approximately 600 base pairs for each of the two genes we examined (16S and cytochrome b). All sequences were submitted to GenBank (Accession Numbers DQ243709–DQ243724). The nucleotide divergence matrix, built according to the Kimura two-parameter model for the two gene segments examined, showed a divergence of less than 3%. Genetic distances among E. sexcinctus specimens captured in different habitats ranged from 0.2 to 1.3% for the cytochrome b gene and from 0.6 to 2.8% for the 16S rRNA (Tables 2 and 3). The distance approach among species using cytochrome b resulted in divergences slightly above 22% for the E. sexcinctus and D. novemcinctus specimens, whereas the comparative analysis with C. unicinctus produced values ranging from 21–24.7% (Table 2). The same approach using 16S rRNA resulted in lower values, which averaged 15.6% between E. sexcinctus and D. novemcinctus, while C. unicinctus averaged less than 11.1% (Table 3).
TABLE 2.
Distance method analysis of mitochondrial cytochrome b gene for E. sexcinctus, D. novemcinctus and C. unicinctus.
TABLE 3.
Distance method analysis of the mitochondrial 16S rRNA gene for E. sexcinctus, D. novemcinctus and C. unicinctus.
The four methods for performing the phylogenetic analysis (Maximum Parsimony, Neighbor-Joining, Minimum Evolution and Maximum Likelihood) generated trees of identical topology. Specimens of E. sexcinctus — regardless of origin, gene segment analyzed, or different methods of analysis — were always included in the same clade (Fig. 2) with high statistical support (bootstrap = 100%), matching the low divergence values found in individuals of this genus. Similarly, a high genetic divergence was found among the three different genera, a fact that was also confirmed by the relationships within the resulting phylogenetic trees.
Discussion
The degree of divergence among the five individuals of Euphractus sexcinctus was much less than that among the three armadillo genera. The strong similarity among these five individuals, acquired from different localities and ecosystems, suggests that the Rio Gurupí is not a barrier to gene flow in the Euphractus populations of this region.
These new locality records support the suggestion of Silva Júnior and Nunes (2001) that the disjunct range of E. sexcinctus may be an artifact of undersampling, rather than a genuine division. The records we present here extend its known distribution to the interfluve between the Rios Tocantins and Gurupí. Taken together with the localities presented by Silva Júnior and Nunes (2001) and Silva Júnior et al. (2001), it seems likely that E. sexcinctus is continuously distributed at least to the southern margin of the mouth of the Rio Amazonas. Additional surveys between the Rios Tocantins and Xingu may provide evidence of a much broader total range than had been previously assumed.
Acknowledgments
We are grateful to the Instituto do Milênio for financial support. We also want to thank both the Laboratory of Mangrove Ecology and the Laboratory of Genetic and Molecular Biology of the Federal University of Pará (UFPA) for logistical support. Frederic Delsuc, Liliana Cortés-Ortiz and Paula Lara-Ruiz kindly reviewed the manuscript and offered valuable comments. The first author was supported by a fellowship from the Brazilian Scientific and Technological Council (CNPq: Process N°. 390007/2004-8).
References
Notes
[1] Fernanda Atanaena Gonçalves de Andrade, Marcus Emanuel Barroncas Fernandes*, Universidade Federal do Pará, Campus de Bragança, Colegiado de Biologia, Laboratório de Ecologia de Manguezal, Alameda Leandro Ribeiro s/n, Aldeia, Bragança 68600-000, Pará, Brazil, e-mail: <atanaena@yahoo.com.br> and <mebf@ufpa.br>
[2] Maria Claudene Barros, Universidade Estadual do Maranhão, Centro de Estudos Superiores de Caxias, Departamento de Química e Biologia, Praça Duque de Caxias s/n, Caxias 65604-370, Maranhão, Brazil, e-mail: <claudene@sesc.uema.br>
[3] Horácio Schneider, Universidade Federal do Pará, Campus de Bragança, Colegiado de Biologia, Laboratório de Genética e Biologia Molecular, Alameda Leandro Ribeiro s/n, Aldeia, Bragança 68600-000, Pará, Brazil, e-mail: <hschneider@uol.com.br>.
