Isothrix bistriata 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 : 180-186

Isothrix bistriata Wagner, 1845

TYPE LOCALITY: Rio Guapore´, Estado do Mato Grosso, Brazil.

DESCRIPTION: This is the largest of the three known species in the genus. External craniodental, and genetic characteristics by which I. bistriata differs from I. pagurus and the closely similar I. sinnamariensis , are detailed in Patton and Emmons (1985) and Vié et al. (1997). As this is the only species of Isothrix known to occur west of the rios Negro­Madeira axis, it differs from other cooccurring echimyids in the Rio Juruá and elsewhere in western Amazonia by the same set of characters detailed above for the genus The most obvious external features of I. bistriata are its grizzled yellow­brown to olive dorsal coloration mixed with black hairs, its well­defined black or dark brown supraorbital stripes extending over the forehead to the nape and bordering a median pale creamy patch on the crown, its pale yellow to buff venter, and its bushy tail typically rust or golden­colored over the basal third to half with the terminal portion black. The skull is illustrated in figure 118.

SELECTED MEASUREMENTS: We present the mean, range, and standard error of the four external and 22 cranial variables in table 51 We separate these data for each of the two mitochondrial DNA clades identified in figures 117 and 120.

GEOGRAPHIC VARIATION: The number of specimens available to us is insufficient to examine either geographic or nongeographic patterns of morphometric characters within the Rio Juruá basin. However, because of the extensive degree of sequence divergence between individuals from the central and headwaters region, and from those at river’s mouth (12.2%, table 50), we compare them morphometrically using discriminant function analysis. We include samples of I. bistriata from southern Venezuela and northern and eastern Perú because Patton and Emmons (1985) demonstrated that samples from Venezuela, for example, were nonoverlapping in discriminant space relative to those from Perú and Bolivia. Here, therefore, we wished to determine not only the degree to which the two mtDNA clades from the Rio Juruá are distinct, but also to examine their morphometric relationship to specimens from elsewhere in the species’ range.

The downriver sample is larger in most measurements compared to individuals of the upriver mtDNA clade (table 51), significantly so in 13 of the 20 cranial variables despite its small size (n = 3) (one­way ANOVA, p <0.05 — RD, OL, BUL, CD; p <0.01 — CIL, ZB, RL, RW­2 ; p <0.001 — MB, IOC RW­1 , and D). The two are not significantly different in any external dimension, although the downriver clade still tends to be larger Bivariate plots of the four geographic samples on the first three discriminant axes are illustrated in figure 117, with the standardized discriminant functions given in table 52 The two Juruá clades do not differ significantly (one­way ANOVA, Fisher’s PLSD mean difference = —0.566, p> 0.359) in their placement on the first axis, but do on the second and third (one­way ANOVA Fisher’s PLSD mean difference = 3.334 and 32.059, respectively, p <0.001 in both cases). The sample from northern and eastern Perú is indistinguishable from that of the upriver Juruá clade on all three axes (one­way ANOVA, Fisher’s PLSD mean difference = 0.512, 0.900, and —0.555 respectively, p> 0.05). This parallels the mtDNA data in which haplotypes from northern Perú fall well within that clade (figs. 117 and 120) Interestingly , the sample from the mouth of the Rio Juruá is not intermediate in discriminant space between the upriver and Peruvian group of samples and that from southern Venezuela, as might be expected by its somewhat intermediate geographic position. Rath­ er, the downriver clade and specimens from Venezuela are the most divergent in mensural characters among the set of geographic samples examined, differing significantly along all three axes (one­way ANOVA, Fisher’s PLSD mean difference = 4.413, —2.130, and —2.860 respectively, p <0.001). On the oth­ er hand, the upriver and Peruvian samples differ from those from Venezuela only on the first axis (one­way ANOVA, Fisher’s PLSD mean difference = 3.848 and —3.328, respectively, p <0.001), and not on the other two .

Despite the limited sample sizes, there appear to be at least three lineages of I. bistriata within Amazonia. The first of these occurs throughout western Amazonia, ranging from northern Perú south to Bolivia and east as far as the middle Rio Juruá in western Brazil. This form exhibits limited geographic variation in mtDNA sequences (table 50) and (with available specimens) minimal morphometric variation as well. On distributional grounds, it is likely that specimens from Estado do Mato Grosso, Brazil, where the type locality of bistriata Wagner is located, belong to this form, and thus Wagner’s name would apply to it. If this proves not to be the case, the earliest name available for the western Amazonian taxon would be villosa Deville, 1852 (see Patton and Emmons, 1985). The second lineage would contain samples of the second mtDNA clade, namely those from the mouth of the Rio Jurua´, the upper Rio Urucu southeast of Tefe´, and from both sides of the lower Rio Negro northwest of Manaus (Rio Jaú and Ilha das Onças, table 49). This form is likewise distinct in mtDNA sequence (fig. 120, table 50) and apparently morphometrically as well, although larger samples are needed for verification. The earliest available name for this taxon is most likely negrensis Thomas, 1920, since the type locality of Acajutuba, on the lower Rio Negro above Manaus, is within this range Finally, the third taxon is that from the upper Río Orinoco region in east­central Colombia and the Casiquiare region of southern Venezuela. Although mtDNA sequences are lacking, this form remains nearly nonoverlapping in cranial morphometric pattern relative to the other two, especially on the first discrim­

inant axis (fig. 121). As documented by Patton and Emmons (1985), the name orinoci Thomas, 1899, with its type locality at Maipures on the upper Río Orinoco in Intendencia Vichada, Colombia, most likely applies to this form. Additional specimens are required for both molecular and morphological analysis before it is possible to determine if each of these three groups warrants separate status at the species level.

DISTRIBUTION AND HABITAT: Isothrix bistriata is a denizen of seasonally inundated forest, either igapó or várzea. All specimens (n = 32) were obtained either in canopy platform traps placed within these forest types (or along their edge, such as on line ‘‘P’’ at Vai­Quem­Quer, locality 15; see fig. 23), were shot in trees, or were taken from tree holes. In the latter case, we tied Tomahawk traps over the mouth of a tree hole and beat on the trunk forcing the animal into the trap. We commonly saw individuals with their heads protruding from tree holes during daylight hours. In such cases, those trees were climbed and Tomahawk live traps nailed over the nest entrance, usually with successful capture during the night.

REPRODUCTION: We caught this species in all months we trapped, and from localities within each of the four geographic regions along the river (fig. 1). None of the females were pregnant, although all adults (those with all cheekteeth erupted and exhibiting some wear) were parous and had enlarged nipples. One female taken at Penedo (locality 7) in late August was lactating, and another taken at Lago Três Unidos (locality o) was captured together in the same trap with a small (49 g) young individual, presumably her offspring. Placental scars were noted in several individuals but were never more than two, suggesting a small litter size. All fully adult males had developed midventral chest glands, approximately 25 mm in length. The testes in these individuals were large, averaging 24 Χ 12 mm (length Χ width), but were always abdominal in position rather than descended into the obvious scrotum These same individuals also had enlarged vesicular glands and enlarged seminiferous tubules in the cauda epididymides, suggesting reproductive competency. Pairs of adult individuals, presumably a male and female were observed on three occasions, one time each at Sacado (locality 5), Nova Empresa (locality 8), and opposite Altamira (locality 10).

KARYOTYPE: The karyotype of I. bistriata was described by Leal­Mesquita (1991), who illustrated both non­differentially stained and C­band complements for a female specimen from the hydroelectric dam at Samuel, Estado do Rondônia, near Porto Velho on the Rio Madeira. Chromosomal data from 16 specimens from the Rio Juruá (10 males and 6 females) do not differ. The diploid number is 60, the fundamental number 116. The autosomal complement consists of one pair of large subtelocentric elements and 28 pairs of metacentrics and submetacentrics grading in size from large to small. The X­chromosome is a large acrocentric and the Y a very small metacentric. In contrast, the diploid and fun­ damental numbers of I. pagurus are 22 and 38, respectively (Patton and Emmons, 1985); those of I. sinnamariensis are 28 and 42 (Vié et al., 1997).

SPECIMENS EXAMINED (n = 41): (1) 1f — MNFS 1411; (2) 1m, 1f — MNFS 1188, 1273; (3) 1m — JUR 240; (5) 2m, 1f, 1 unknown — JLP 15568, 15625–15627; (7) 1m, 1f — MNFS 471, 500; (8) 1m, 1f — MNFS 431, 506; (10) 6m, 5f — JUR 199–200, MNFS 893, 900–901, 914, 920, 934, 941– 942, 957; (11) 3m, 1f — MNFS 797, 809– 810, 833; (14) 1m — MNFS 1725; (o) 3f — JUR 544–546; (15) 1f — JUR 296. Specimens from Royal Natural History Museum, Stockholm (see Patterson, 1992): João Pessoa [= Eirunepe´] (5m, 3f — 2118, 2138, 2140, 2161, 2183, 2217, 2235, 2475); Lago Grande (1m — 2488).