The two species of two-toed sloths, Choloepus didactylus and C. hoffmanni, are the only extant representatives of the Megalonychidae (Adam, 1999), occurring in partial sympatry in the Andean regions and western Amazonia (Le Pont and Desjeux, 1992; Emmons and Feer, 1997; Adam, 1999; Eisenberg and Redford, 1999). Although their distributions are reasonably well-understood on a broad scale, the precise boundaries of their ranges are still unresolved. According to Wetzel and Ávila-Pires (1980), Wetzel (1985), Adam (1999) and Eisenberg and Redford (1999), C. didactylus occurs across all of northern Amazonia, from the eastern Andes to northeastern Surinam, reaching the northern coast of Brazil in the states of Amapá, Pará and Maranhão. The maps presented by Wetzel (1985), Emmons and Feer (1997) and Eisenberg and Redford (1999) all suggest that in Western Amazonia, this distribution extends southward to 10°S latitude. In central and eastern Amazonia, however, their maps show Choloepus didactylus as being restricted to a narrow belt along the southern edge of the Amazon River — although there is no ecological reason why the species might not occur further to the south. Wetzel and Ávila-Pires (1980) reported the only records for this region available at the time, referring to specimens from Santarém, Taperinha and Rio Barcarena in Pará, as well as Humberto de Campos in Maranhão. Specimens in the mammal collection of the Museu Paraense Emílio Goeldi (MPEG) confirm this distribution, including those from the following localities in eastern Pará: Marituba (MPEG 22070), Paragominas (MPEG 30677), Fazenda Cauaxi, municipality of Paragominas (MPEG 26315, 26316, 26317), and Rodovia Belém-Brasília, Km-75 (MPEG 2385, 2391, 2394). With the exception of the last three individuals, all these specimens were collected after 1980 and were thus not available to Wetzel and Ávila-Pires.
Mascarenhas and Puorto (1988) reported C. didactylus from the middle-lower Rio Tocantins, south of the range indicated by other authors. The specimens which Mascarenhas and Puorto collected are now in the MPEG collection (Tucuruí, Base 5: MPEG 12597; Jacundá: MPEG 11884), and these localities were included in the more recent range maps of Adam (1999) and Eisenberg and Redford (1999). Soon there-after, Toledo et al. (1999) extended the distribution of C. didactylus to the region of Carajás in southern Pará, approximately 06°S, 50°W. Here we demonstrate that the eastern part of the distribution of this species extends further south than previously known.
While one of us was conducting fieldwork in a settlement in northern Mato Grosso (Trinca, 2004), a local man found a skull of C. didactylus in the nearby plantation of Fazenda do Tenente, Japuranã district (10°03′56.6″S, 58°00′02.3″W), municipality of Nova Bandeirantes (Fig. 1). This area is located in the Juruena interfluve between the Rios Juruena and Teles Pires.
The specimen, an adult skull without the mandible, is now in the MPEG mammal collection (MPEG 36871). We identified the specimen based on the diagnostic characters provided by Wetzel (1985) and by direct comparison with specimens of Choloepus in the MPEG collection. The skull fits the description of C. didactylus, possessing a large anterior and small posterior inter-pterygoid space, large pterygoid expansions, absence of a posterior pair of foramina in the inter-pterygoid space, pterygoid sinuses broadly inflated and broad contact of maxilla with frontal bone, not interrupted by lacrimals.
The original vegetation of the Juruena basin was mainly dense rainforest, interspersed with open rainforest and transitional areas (Dinerstein et al., 1995; Miranda and Amorim, 2000). Within the past 30 years, these forests have been destroyed by the expansion of agriculture and pastures, and the increase in cattle-ranching and soybean farming may have extirpated C. didactylus from many areas in east-central Mato Grosso.
This new record suggests that the presumed absence of this species in south-central and eastern Amazonia may be an artifact of undersampling, which has long been an impediment to understanding the diversity and biogeography of Neotropical fauna (Vivo, 1996; Silva Júnior, 1998; Silva et al., 2001). Surveys covering this entire region would most likely indicate that C. didactylus has a wider geographic distribution within the forests of Amazonia than now understood.
We are grateful to Mr. Marcos, who found the skull at Fazenda do Tenente and gave it to C. T. Trinca. We also thank Mr. Leonar Dallagnol, Mr. Antonio Geraldo Conjiu and Mr. Arley Brumati for their logistical support during the first author's field research in the Juruena interfluve. Special thanks are due to Luís Cláudio Barbosa for drawing the map and to Marinus Hoogmoed, Teresa Cristina Sauer de Ávila-Pires, Ana Lima and John Aguiar for their reviews of the manuscript.
 Cristiano Trapé Trinca and Francesca Belem Lopes Palmeira, Faculdade de Ciências Biológicas, Centro de Ciências Médicas e Biológicas, Pontifícia Universidade Católica de São Paulo, Praça José Ermírio de Moraes, 290, Vergueiro, Sorocaba 18030-230, São Paulo, Brasil, and Departamento de Pesquisa, Reserva Brasil, Av. Dr. Silva Melo, 520, complemento 606, Jardim Taquaral, São Paulo 04675-010, São Paulo, Brazil, e-mail: <firstname.lastname@example.org> and <email@example.com>
 José de Sousa e Silva Júnior, Coordenação de Zoologia, Museu Paraense Emílio Goeldi (MPEG), Av. Perimetral, 1901/1907, Terra Firme, Belém 66077-530, PA, Brazil, e-mail: <firstname.lastname@example.org>.