Oecomys bicolor (Tomes, 1860)

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 : 122-128

Oecomys bicolor (Tomes, 1860)

TYPE LOCALITY: ‘‘Gualaquiza,’’ Río Gualaquiza, 885 m, Provincia de Morona­Santiago, Ecuador.

DESCRIPTION: One of the two small­bodied species in the river basin, with average total length of 220 mm (table 29 and fig. 82) and a tail 112% of head and body length. The hind feet are short (22.7 mm) and broad, with tendency for a slightly darkened patch over the metatarsals. The tail is long and uniformly dark in color above and below, typically with 20 scale rows per cm, and clothed in short (about two scale rows long) hairs and terminating in a short, but distinct pencil The dorsal pelage is short and rather bright rufescent tawny in color; the ventral pelage is comprised of pure white hairs; a thin band of gray­based and white­tipped hairs may be present at the transitional boundary between the dorsal and ventral coloration, otherwise the dorsal and ventral coloration are sharply demarcated. The skull (fig. 85) is small (CIL averages 27.85 mm) with small toothrows (MTRL 4.03 mm), with very shallow zygomatic notches, more so than other species well­developed and strongly diverging supraorbital ledges extending as ridges onto the parietals; upper incisors slightly opisthodont bullae small and uninflated; incisive foramen

short (4.7 mm) and teardrop in shape; posterior palatal pits small and undivided; parapterygoid fossae relatively deeply excavated; alisphenoid strut present in 19 of 22 individuals, at least on one side; hamular process of squamosal very short and broad, totally occluding subsquamosal foramen in most individuals; tegmen tympani not in contact, or only barely so, with squamosal. The combination of small size and pure white venter distinguishes O. bicolor from all other species within the Rio Jurua´.

SELECTED MEASUREMENTS: See table 29.

NONGEOGRAPHIC VARIATION: The number of specimens of this species is limited, but when pooled across localities, there are substantial differences in nearly all individual measurements of the skin and skull relative to an individual’s toothwear age class (table 32). In every case of significance, the relationship is strongly positive (Pearson product moment correlation coefficients at p <0.05 range from 0.420 –0.495; at p <0.01, 0.560 – 0.627; and at p <0.001, 0.648–0.753). No variables exhibit significant sexual dimorphism, however. Consequently, in studies of geographic variation in this species, although the sexes might be effectively pooled, care must be taken to account for age differences among samples. The apparent continual individual growth throughout life is a characteristic of many sigmodontine rodents, such as Oligoryzomys (Myers and Carleton, 1981) , Akodon (Myers, 1989; Myers et al., 1990), and Zygodontomys (Voss, 1991) , a phenomenon substantiated by laboratory growth studies on the latter genus (Voss et al., 1990)

GEOGRAPHIC VARIATION: Substantial differentiation among some localities in mtDNA cytochrome­b sequence exists for this species, with specimens from localities within the Mouth Region averaging over 5% different in relation to those from all upriver sites (fig. 86). Unfortunately, we lack sufficient materials to determine if a similar degree of differentiation exists between these geographic areas in morphological characters

only a single adult is available from the Mouth localities. In general color, size, or qualitative features of the skull, however, there are no notable differences between this specimen and those representing the upriver clade. Additional specimens should be sought to examine this question further.

DISTRIBUTION AND HABITAT: We found this species throughout the Rio Jurua´, with specimens taken at localities within each of the four regional sample sites and on both sides of the river (fig. 81). The majority were taken in undisturbed várzea forest (36 of 43, or 84%); three were captured on the terra firme plots (two at Barro Vermelho [locality 12] and one at Sobral [locality 4]); and four were taken in second­growth forest at Igarapé Porongaba (locality 1). Most individuals were obtained in canopy platform traps (32 of 43, or 74%); only four were taken in traps placed on the ground or on logs laying on the ground. The remainder (seven) were trapped or shot in low vine tangles approximately 2 m above ground.

REPRODUCTION: Pregnant females were taken from at least the months of August through February, a period spanning the dry season and the beginning of the rainy season The modal litter size was 2 (mean 2.5, range 1–4; n = 11). Adult males with scrotal testes were taken in at sites from the months of September through March. The few data suggest that both sexes breed for a prolonged period of the year, perhaps continuously throughout. No females were taken, however that were simultaneously pregnant and lactating.

KARYOTYPE: 2N = 80, FN = 140 (fig. 87) We karyotyped 17 specimens from seven separate localities, at least one in each of the four geographic sample regions as follows locality 1 (n = 1), locality 2 (n = 5), locality 5 (n = 5), locality 8 (n = 3), locality 12 (n = 1), locality 13 (n = 1), and locality 14 (n = 1). All specimens exhibited the same karyotype with 19 pairs of medium to small metacentrics and submetacentric autosomes 12 pairs of subtelocentric autosomes, and 8 pairs of acrocentric autosomes. The X­chromosome is a large metacentric, the largest in the complement; the Y­chromosome is a small chromosome that appears acrocentric This karyotype differs slightly from that reported by Gardner and Patton (1976) for specimens of O. bicolor from eastern Perú (Balta, Río Curanja, Depto. Ucayali), which was reported with three additional pairs of uniarmed and three fewer biarmed autosomal elements. However, distinctions between the morphological categories can be quite difficult with such small chromosomes, and not too much emphasis should be placed on the presumptive differences between these two karyotypes. Certainly, there are no apparent differences between karyotypes belonging to individuals of the two geographic mtDNA clades illustrated in figure 86.

SPECIMENS EXAMINED (n = 43): (1) 1m, 3 — MNFS 1210, 1260–1261, 1320; (2) 2m 4f — MNFS 1187, 1238, 1252, 1332–1333 1396; (b) 1f — MNFS 1002; (3) 2m — MNFS 1514, 1679; (4) 1m — MNFS 1499

(5) 7m, 7f — JLP 15744, MNFS 581, 613, 625, 639–640, 651–653, 671–675; (8) 4m, 4f — JLP 15403, 15414, 15433, 15449, MNFS 466, 505, 507, 513; (h) 1m — JLP 15214; (j) 1f — MNFS 439; (11) 1f — MNFS 834; (12) 2f — JLP 15777, MNFS 749; (13) 1m — JUR 304; (14) 1f — 566.

Oecomys species

DESCRIPTION: We caught a second smallbodied species at three localities (Condor, locality 6; Vira­Volta, locality 14; and Vai­ Quem­Quer, locality 15). It was sympatric with O. roberti at the first of these, and with three other species, including O. bicolor , at the second, but was the only species found at the last locality (fig. 80). Only three individuals were trapped; one old adult and two young animals. This taxon is quite different from all others, including O. bicolor in molecular mtDNA sequence (figs. 82 and 86; table 28), and is also demonstrably distinct in morphology from O. bicolor , the only species with which it might be confused. G. G. Musser (personal commun.) provisionally considers it a member of the ‘‘ O. bicolor group,’’ largely by virtue of its small size.

In comparison to O. bicolor , this mouse is smaller in all external and cranial measurements, significantly so in most despite the very limited sample (table 29; fig. 83). Not surprisingly, therefore, it is completely separable from O. bicolor in multivariate space based on a discriminant analysis (fig. 84, and above), with 100% of individuals of both species correctly assigned with posterior probabilities> 0.9999. It is darker dorsally dull reddish brown as opposed to lighter, orangish brown. The venter is pure white as in O. bicolor , but there is a narrow band of gray­based hairs separating ventral and lateral coloration. The hind feet are slightly shorter than those of O. bicolor , and the dorsal surface is paler. The tail is uniformly dark brown both above and below and with a short pencil, as in O. bicolor . The dorsal fur is short and thick, but longer than that of O bicolor . Cranially, this species has a somewhat more slender rostrum when paired with

similar aged O. bicolor , a narrower interorbital region, and distinctly shorter toothrows (mean MTRL = 3.58 vs 4.03; p <0.01 by 2­tailed t ­test). The incisive foramen is teardrop in shape rather than oval as in O. bicolor , widest posteriorly rather than at its midlength. The anterior border of the mesopterygoid fossa is flatter, less arched into the palate, and the parapterygoid fossae are deeper. The upper molar teeth are similar in occlusal details, except that distinct cuspules are evident at the entrance to both the proto­ and hypoflexi in all three specimens. Although these cuspules are occasionally observed in the teeth of O. bicolor , they are never as distinctly well developed.

SELECTED MEASUREMENTS: See table 29.

MOLECULAR PHYLOGEOGRAPHY: Each of the three individuals came from separate localities. Interestingly, the two specimens from the left bank localities of Condor (locality 6) and Vira­Volta (locality 14), located some 660 km apart (table 1), are nearly identical in their respective cytochrome­b sequences. These two differ at only 6 sites of the 414 base pairs examined (Kimura two­parameter distance of 0.24%; fig. 85). The third specimen, however, is quite distinct, differing from the other two by an average Kimura two­parameter distance of 7.29% (fig. 86). It was trapped on the right bank at Vai­Quem­ Quer (locality 15), opposite Vira­Volta in the Mouth region of the river. These data suggest that two deeply divergent mitochondrial clades, perhaps signaling separate species status, are separated by the Rio Jurua´, at least in its lower reaches. Additional collections of this mouse are badly needed from within the Rio Juruá Basin, as well as elsewhere, before its status can be adequately understood.

DISTRIBUTION AND HABITAT: This species was obtained only in terra firme forest in Sherman traps; one specimen was on the ground, the other two in canopy platforms positioned at 9.3 m and 12.5 m above the ground. We obtained so few specimens of this species that all data must be taken as preliminary, but it appears to occupy a distinctly different habitat than its similar sized relative, O. bicolor , which is found nearly exclusively in flooded várzea forest within the Rio Juruá basin.

REPRODUCTION: All three individuals collected were females, two subadults that were nulliparous and the third pregnant with three embryos when trapped in May during the wet season.

KARYOTYPE: 2n = 86; FN = 98 (fig. 88) We karyotype each of the three specimens of this species that we obtained. Despite deep differences in mtDNA sequences (see fig 86), each of these appeared to have the same karyotype, although the specific determination of the number of arms for many of the very small autosomal elements is, admittedly, equivocal. The autosomal complement consists of seven pairs of small metacentrics and submetacentrics and 35 pairs of medium to small acrocentrics. Although no male specimens were examined, we identify the X­chromosome as a very large subtelocentric, an identification that is consistent with the X­chromosome of other species of Oecomys that we, and others, have karyotyped

SPECIMENS EXAMINED (n = 3): (6) 1f — JLP 15675; (14) 1f — JUR 480; (15) 1f — JUR 354.