Isothrix Wagner, 1845

PATTON, JAMES L., DA SILVA, MARIA NAZARETH F. & MALCOLM, JAY R., 2000, Mammals Of The Rio Juruá And The Evolutionary And Ecological Diversification Of Amazonia, Bulletin of the American Museum of Natural History 2000 (244), pp. 1-306 : 178-180

Isothrix Wagner, 1845

Bushy­tailed tree rats

The Amazonian taxa of brush­tailed tree rats of the genus Isothrix were reviewed by Patton and Emmons (1985). They recognized two species, which replace one another geographically from west to east within the basin. Isothrix bistriata is widely distributed in western Amazonia, from southeastern Colombia south along the base of the Andes to northern and eastern Bolivia, thence east on both sides of the Rio Solimões to at least the Rio Madeira and Rio Negro (fig. 117) Throughout this area, the species is most commonly encountered in seasonally flooded forest, either várzea or igapó. On the east banks of the Rio Negro and Rio Madeira, I bistriata is replaced by I. pagurus Wagner the second species recognized by Patton and Emmons (1985), which apparently prefers terra firme forest (e.g., Malcolm, 1990). A third species of brush­tailed rat has recently been described from French Guiana, I. sinnamariensis (Vié et al., 1997) . This species is similar to I. pagurus , its geographically closest relative, in both morphological and mtDNA sequences, but differs extensively in chromosome complement.

Isothrix is unusual among echimyid rodents of Amazonia because it has soft fur that lacks any stiffened hairs or spines characteristic of other genera, and its tail is wellhaired, bushy along its entire length instead of naked or with a long terminal pencil. By virtue of the bushy tail, the Aguaruna Jívaro of northern Perú grouped Isothrix with the tree squirrels rather than with its biological relatives (Patton et al., 1982). The skull has a short, broad rostrum and somewhat flat lateral profile, broad interorbital region with parallel, overhanging supraorbital ledges extending posteriorly as ridges onto the parietals (fig. 118). The maxillary cheek­toothrows are parallel, slightly bowed posteriorly and separated by a narrow palate with the mesopterygoid fossa penetrating to the level of M2. The teeth (fig. 119A) are nearly laminate, each with three labial flexi extending two thirds across the tooth surface and meeting a short median flexus in the unworn condition. All flexi are notably broad, being wid­

er than the lophs they separate. The median flexus is particularly wide and confluent with the first lateral flexus in early stages of wear on all four teeth. The first and third lateral flexi close with minimal wear, becoming islands surrounded by oval lophs. The combination of flat frontal region with strongly developed but parallel supraorbital ledges, short, broad rostrum, and distinctive cheekteeth readily differentiate the skull of I. bistriata from any other sympatric, or nearly sympatric echimyid taxa.

A limited view of molecular phylogeographic variation in Isothrix was provided by da Silva and Patton (1993), who identified two clades within I. bistriata that were nearly equidistant between each other and I. pagurus . Vié et al. (1997) provided sequence data for I. sinnamariensis from French Guiana, and compared it with both I. bistriata and I. pagurus . We have added slightly to this database recently, and summarize all available mtDNA cytochrome­b data in fig. 120, a strict consensus tree based on 798 base pairs of sequence for 20 haplotypes from the three known species. Available geographic localities are mapped in fig. 117, as are regional monophyletic clades based on the topology presented in fig. 120. Provenance and vouch­ er specimen catalog numbers are given in table 49. Average Kimura two­parameter distances among regional clades and species are provided in table 50.

Consistent with earlier findings (da Silva and Patton, 1993), I. bistriata comprises two geographically differentiated forms, both of which are found within the Rio Juruá basin but replace one another geographically. One of these, an upriver clade, matches specimens from northern and southern Perú and northern Bolivia. The other, a downriver clade also is known from both sides of the lower Rio Negro north of Manaus and on the opposite side of the Rio Solimões (figs. 117 and 120). Also consistent with our previous results, these two clades do not form a sisterpair relationship according to available data Rather, they differ by 12.2% from each other and together average only 12.4% different from the species pair of I. pagurus and I. sinnamariensis (table 50). Substantial additional variation exists between haplotypes within each of the two clades of I. bistriata with up to 7% divergence between some lo­

calities in the upriver clade and nearly this level within the downriver clade (da Silva and Patton, 1993: fig. 120). Additional geographic sampling and analyses are required to determine if the mtDNA clades of I. bistriata represent separate species, as implied by their deep divergence and lack of sister relationship. Meanwhile, we continue to treat both clades as members of a single polytypic species.