Proechimys cuvieri Petter, 1978

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 : 224-228

Proechimys cuvieri Petter, 1978

TYPE LOCALITY: ‘‘Saül (S 21), Guyane française,’’ French Guiana.

DESCRIPTION: Proechimys cuvieri is similar to P. brevicauda with its medium­sized body and relatively short ears, hind feet, and tail (tables 60 and 64). The tail is approximately two thirds the length of the body; in most specimens it is distinctly bicolored, with the ventral coloration varying from creamy to pale brown. In no specimen does the tail appear uniformly dark, as is often the case in P. brevicauda . The tail hairs are prominent but scales remain visible to the eye; these range from 9–12 per centimeter. Overall, this species has a relatively dark­colored body, tail, and feet. As in other species, the color of the midline of the dorsum appears darker than the sides of the body, especially over the rump. In contrast especially to P. brevicauda , P. cuvieri is brightly colored with a strong reddish overall hue. The aristiform hairs are stiff to the touch, larger and wider than those of all other species except P. echinothrix , but with a whiplike tip (da Silva, 1998: fig. 3). The dorsal surface of the hind foot is brownish, and in most specimens the pale color of the inner surface of the hind limbs extends across the tarsal joint as a ful­ vous stripe. No lateral line is present and the reddish color of the sides contrasts sharply with the almost pure white venter. In addition to its medium­sized body and moderately stiff aristiforms, a helpful character for identifying P. cuvieri males in the field is the very short and massive baculum, which has a broad shaft, expanded base, and apical extensions (fig. 137).

The skull is relatively large, with a long narrow rostrum (figs. 138 and 139; table 64 and well­developed supraorbital ridges, but with weakly developed temporal ridges (Patton, 1987: fig. 21d). The overall shape of the incisive foramen is lyrate, but only moderately constricted posteriorly in contrast to P brevicauda (see Patton, 1987: figs. 13 and 14). The posterolateral margins of the foramen are flanged (fig. 140) in all individuals (n = 25), but the flanges are not as developed as in P. brevicauda , forming only weak grooves that extend into the palate (in 33 of 35 specimens). The premaxillary portion of the septum is long and almost always (31 of 35) in contact with the maxillary portion which may be equally either slightly keeled or smooth; the vomer is slightly to well exposed ventrally in most specimens (n = 25) but in some (n = 9) it is completely enclosed within the premaxillary sheath. The groove on the floor of the infraorbital foramen is slightly developed in the majority of specimens (26 of 35), but the degree of development of the lateral flanges is weak; in two of the remaining individuals, the floor is smooth without any groove, and in seven it has a groove present with moderately developed lateral flanges. The mesopterygoid fossa penetrates the posterior palate to a moderate degree (fig. 141), extending to the posterior margins of M 3 in 16 specimens, but in approximately equal number of individuals it reaches into the posterior half of M3 (17 specimens) or to its anterior half (2 specimens). The counterfold pattern is typically three folds in all four upper and lower teeth except for pm4 which has four folds in most specimens examined (24 of 34). A smaller number of individuals (n = 7) has either two four, five, two to three, three to four or four to five folds in a given tooth.

SELECTED MEASUREMENTS: Selected exter­ nal and cranial measurements are summarized in table 64.

COMPARISONS: In overall morphology, P. cuvieri is most similar to P. brevicauda . Both are characterized by a medium­sized body, ears, and hind feet and relatively short tail (table 60). In comparisons of series of both species, P. cuvieri is brighter and more reddish dorsally than P. brevicauda , which tends to be duller and more brownish; it also has a white venter that lacks the rufous patches typical of P. brevicauda . In the latter species, the coloration of the ventral surface of the tail tends to be darker towards the tip, but appears uniformly paler than the dorsal surface in the specimens of P. cuvieri we studied. The aristiform hairs of P. cuvieri are stiffer to the touch, being much larger and wider than those of P. brevicauda , but less so than in P. echinothrix (da Silva, 1998) . As in other species of spiny rats, except P. brevicauda , P. cuvieri has no lateral stripe and the reddish color of the sides contrasts sharply with the mostly pure white ventral coloration. The baculum of P. cuvieri is most similar in general shape to that of P. echinothrix and P. pattoni (fig. 136), although the three differ in size (see Patton, 1987).

MOLECULAR PHYLOGEOGRAPHY: We have cytochrome­b sequence data for individuals from 11 localities throughout the extensive geographic range of P. cuvieri (fig. 144, above; table 67). We examined relationships among the 16 haplotypes for which 798 bp were available from these 10 localities by maximum parsimony. There are two major clades, with an average divergence of 9.8%, one of which is divided into three equally divergent groups (fig. 144, below). The first clade groups all specimens from the Rio Jurua´, from the Headwaters through the Lower Central sections of the river. We did not find apparent subgrouping by either regional area or river bank; haplotypes from the four available localities differ by an average of less than 1%. The second major clade includes a group of widely scattered localities in eastern Amazonia, from the Guianan region of eastern Venezuela and French Guiana to the Carajás region of Estado do Pará south of the Rio Amazonas and the left (= east) bank of the Rio Negro near the city of Manaus in central Amazonia. Haplotypes from these five localities differ from each other by an average of 3%. Also included in the second clade are single localities in the upper Rio Negro of northwestern Brazil and the Río Santiago of the Departamento de Amazonas in northern Perú. Collectively, these three groups differ by an average of 7.3%. Clearly P. cuvieri is comprised of a series of quite divergent clades that replace one another across Amazonia, but additional sampling is needed to determine whether any of these correspond to separate species. Morphologically, this taxon is relatively uniform with little evidence of geographically distinguishable populations.

MORPHOMETRIC VARIATION: Our samples of adults are limited in number, although the species was trapped at localities in the Headwaters and both Upper and Lower Central regions (fig. 133). Across this region, however, only one cranial dimension exhibits any interlocality differentiation (OCB; one­way ANOVA, F 5,32 = 4.954, p = 0.0018); all oth­ er cranial and all external dimensions display uniform variation across their sampled range within the Rio Jurua´. Sexual dimorphism is also nonexistent (p> 0.05 for all variables in one­way ANOVAs).

DISTRIBUTION AND HABITAT: Proechimys cuvieri is known from scattered localities throughout Amazonia from the coastal Guianan region west along the Amazon Riv­ er from near its mouth to northern Perú (Patton, 1987: fig. 4). Our specimens from the Rio Juruá extend this distribution about 400 km to the south. All individuals were collected in the Headwaters, Upper, and Lower Central regions, but not in the Mouth Region. It is likely that the species is present in the latter region, since specimens of P. cuvieri have been taken localities along the length of the Rio Solimões–Rio Amazonas from near Iquitos to Belém. The majority of specimens were trapped in terra firme forests (28 of 57), but the species was also collected in locally inundated forest (5 individuals) or secondary upland forest and abandoned gardens (22 individuals; table 63). We captured animals in Tomahawk and Sherman traps in all habitats as well as by hunting and with the use of snap traps (Victor rat traps and a similar all­metal trap made and sold in Manaus for domestic use). Although the total number of each kind of trap varied greatly, the most of our sample (66%) were caught in Tomahawk traps with only 20% in Sherman traps; 4% were shot, 10% were caught with snap traps. Of the animals captured in Sherman and Tomahawk traps, young and subadults were caught in nearly equal numbers in both kinds of traps, but 89% of adults were caught in the larger Tomahawk traps.

REPRODUCTION: We caught specimens of P. cuvieri during the interval from August through March, a period that includes most of the dry season and the early rainy season. We classified 13 of the 17 male specimens for which autopsy data are available as reproductively active; all belonged to age classes 9 or 10. The four reproductively inactive individuals were of age classes 3, 5, 7, and 9. We caught pregnant females (n = 4) only in the months of February and March; one lactating individual was collected in August. Finally, we found young animals (ages 1 to 5) in all sample regions. These data are too limited to determine whether breeding is largely confined to the wet season or extends through the dry period as well. Eight of nine pregnant and parous females were old adults (ages 9 or 10); one pregnant female was of age 7. Litters consisted of two in all cases.

KARYOTYPE: We karyotyped 35 specimens of P. cuvieri from six different localities (locality 1 [n = 5], locality 3 [n = 3], locality 4 [n = 3], locality 6 [n = 8], locality 7 [n = 13], and locality 12 [n = 3]). All specimens were 2n = 28, FN = 48 (fig. 143B). The autosomes include a pair of large metacentrics, eight pairs of medium­sized to small metacentrics and submetacentrics, two pairs of large and one pair of small subtelocentrics (of which the small pair possess terminal satellites on the long arms), and a single pair of medium­sized acrocentrics. The X chromosome is a small acrocentric and the Y is a minute one. There is some karyotypic variation in P. cuvieri over its sampled range (table 68), and we illustrate karyotypes from specimens from the Rio Jaú (Estado do Amazonas, Brazil) and the Rio Xingu (Estado do Para´, Brazil) in figure 143 along with one from the Rio Juruá to show the known extent of variation present. Diploid number remains constant at 2n = 28, but FN varies between 46 and 50, due to differences in the number of subtelocentric and acrocentric autosomes. The karyotype of P. cuvieri is similar, if not identical in nondifferentially stained preparations, to some samples of P. brevicauda (compare figs. 143A and 143D and tables 66 and 68).

COMMENTS: Patton (1987) placed two taxa within his cuvieri ­ group based on similarities in bacular structure, although he recognized that these were not closely related. One of these corresponds to P. cuvieri and the other has been recently described as P. pattoni (da Silva, 1998) . Our cytochrome­b analyses corroborate the distinctiveness of both taxa and, as they do not form a monophyletic clade (fig. 134), suggesting that their bacular similarity is convergent rather than indicative of close relationship. Although P. pattoni should not be considered a member of the cuvieri ­ group, P. cuvieri itself is likely to prove a composite. While there is general cranial similarity in P. cuvieri throughout its range as mapped by Patton (1987: fig. 4), our geographic samples of this species exhibit divergent regional units with levels of sequence divergence equivalent to those we found between other species of the genus. Clearly, a more refined examination of detailed geographic sampling is needed to assess the possibility that each of these clades may represent a distinct species.

SPECIMENS EXAMINED (n = 55): (1) 2 m, 7 f — MNFS 1080, 1085, 1070, 1075–1077 1192–1193, 1385; (2) 1 m — MNFS 1283 (3) 1 m, 3 f — MNFS 1580, JUR 207, 236 238; (4) 2 f — MNFS 1486, 1644; (6) 5 m 4 f — JLP 15603, 15636, 15638, 15642 15664, MNFS 531, 534–535, 551; (7) 14 m 10 f, 1 unknown — JLP 15258, 15260 15267, 15271, 15278, 15284–15285, 15308 15310, 15369, 15405, 15460–15463, 15478 MNFS 331–332, 346, 353, 363, 365, 408 499, 512; (12) 2 m, 3 f — JLP 15886, 15903 15890, 15922, JUR 187.