Dactylomys, I. Geoffroy Saint-Hilaire, 1838

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 : 171

Dactylomys I. Geoffroy Saint­Hilaire, 1838

Bamboo rats

The bamboo rats are among the largest arboreal rodents in Amazonia, equivalent in size to the tree squirrels Sciurus igniventris and S. spadiceus but considerably smaller than porcupines of the genus Coendou . They are also remarkable for their loud vocalizations that can be heard for considerable distances. Despite their conspicuousness, however, they are very poorly known in all aspects of their biology and evolutionary relationships. They are folivores that do not readily enter traps, either because the baits used are not attractive or for other reasons, and their weak eye­shine and methodical movements make them difficult to see at night. Indeed, individuals are more readily detected by their distinctive vocalization or by the strong, musky odor that permeates their living areas rather than seen (Emmons, 1981; Emmons and Feer, 1997). The common name stems from their characteristic association with patches of bamboo, apparently a major food item, but they also occur in riverine vegetation, especially canebrakes, and in upland forest on the rich soils of western Amazonia (Emmons, 1981).

The genus is readily recognizable by both external and cranial characteristics. It is the largest echimyid within Amazonia, with body lengths in excess of 280 mm; the tail is furred at the base but otherwise naked and distinctly scaly, and exceeds the head and body length; the ears are small; the dorsal coloration is grizzled yellow­olivaceous streaked with black, with thighs and sides varyingly yellow to rusty orange, and with either a pale beige or blackish stripe that extends from between the eyes to the nape of the neck; the muzzle is distinctly square in lateral profile; and the forefeet have four long digits (the middle ones separated by a gap) with nails, not claws. The skull (fig. 113) is large and robust, flat in lateral profile, with a short rostrum, broadly overhanging supraorbital ledges with triangular postorbital processes, deep and massive zygomatic arches with dorsal postorbital processes comprised equally of squamosal and jugal, and enlarged cheekteeth comprised of complete, or nearly complete transverse laminae (fig. 114).

Three species of bamboo rats were listed by Woods (1993): D. dactylinus (Desmarest) occurring throughout most of lowland Amazonia; D. boliviensis Anthony , apparently restricted to lowland Bolivia and southeastern Peru´; and D. peruanus Allen , known only from the cloud forests of southeastern Peru above 1000 m. Da Silva and Patton (1993) suggested that two species are present in lowland Amazonia, while Emmons and Feer (1997) listed only the single species D. dactylinus . These latter authors noted, however that both external color pattern and vocal call structure (number and timing of pulses) varies geographically, and acknowledged that more than one species might be present

Da Silva and Patton (1993) based their conclusion that two species of bamboo rats occurred in western Amazonia on mtDNA sequences from the few specimens obtained from the Rio Juruá and elsewhere, supplemented by notes on vocalizations. The limited geographic sampling of that study has been supplemented only minimally by more recent sampling in central Brazil and central Perú (fig. 115, top; table 47). Sequences separate into two well­differentiated groups, one restricted to the headwaters of the Rio Juruá and adjacent Perú and the other found throughout central and eastern Amazonia of Brazil and northern Bolivia (fig. 115, bottom). These differ by an average of 8.74% sequence divergence, and the two haplotypes obtained from the Rio Juruá (one from the Headwaters region [locality a, the other from the Mouth region [locality 14]) differ by 9.53%. The seven haplotypes from the broad area across central and eastern Amazonia differ among themselves by only 2.88%, on average.

We follow da Silva and Patton (1993) in recognizing two species in western Amazonia in general, and along the Rio Juruá specifically. However, we are unsure of the proper application of the available names to either species. Da Silva and Patton used the name D. boliviensis Anthony to refer to the mitochondrial clade in the Rio Juruá headwaters, and D. dactylinus (Desmarest) to that in central Amazonia. The type locality of boliviensis is in the Departamento de Cocha­ bamba, Bolivia, but the Bolivian specimens of bamboo rats for which mtDNA data are available (from the Departamento de Pando) belong to the central Amazonian clade we allocate to D. dactylinus (fig. 115). We have not examined the specimens from the Departamento de Pando for which we have mtDNA data, but, based on comparisons with the holotype of D. boliviensis (AMNH 38709), we apply this name to our specimens from the Headwaters clade. The holotype of D. boliviensis has the same overall color pattern (including color of middorsal hairs and dark head stripe) and cranial features described below for these specimens. The type locality of D. dactylinus is unknown, although Thomas (1911) restricted it to the upper Amazon, and proposed the name canescens for individuals from central Amazonia with the rusty underfur characteristic of the specimens we allocate to dactylinus . However, Lönnberg (1921) correctly noted that the holotype of dactylinus in the Paris Museum has rusty underfur, and thus probably came from somewhere downriver from Iquitos in northwestern Perú. The genus is under current review by Louise H. Emmons. Clearly there is much remaining to understand about species boundaries as well as the most basic aspects of the natural history of these animals.