Proechimys brevicauda Gunther, 1877

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 : 213-224

Proechimys brevicauda Gunther, 1877

TYPE LOCALITY: ‘‘Chamicuros, Huallaga River,’’ Departamento de Amazonas, Peru´.

DESCRIPTION: Proechimys brevicauda is medium­sized with relatively short ears, hind feet, and tail, although by its size alone, it cannot be distinguished from other sympatric species, especially P. cuvieri (table 60). The length of the tail is approximately twothirds that of the body; the dorsum of the tail is a darker brown than the venter, and may be almost entirely unicolored brown or grade from cream to light brown towards the tip in some specimens; hair on the tail is sparse and the scales are conspicuous to the eye (approximately 10 annuli per cm). The color of the dorsal surface of the hind foot varies from cream to dark brown, and in most specimens the paler color of the inner surface of the hind limbs extends across the tarsal joint The aristiform hairs are medium in length and narrow, with a whiplike tip (see da Silva, 1998: fig. 3). One of the most distinctive features of this species is the characteristically fulvous lateral stripe and the rufous wash which extends ventrally over the chin, throat and abdominal region, covering the entire venter in the most extreme examples. The

dorsal color is uniformly dark reddish­brown streaked with black.

The baculum is elongated and broad, with well­developed apical wings (fig. 137) (for characterization and list of references in which the baculum of this species has been previously presented, see Patton, 1987).

The skull is relatively large (figs. 138 and 139) with a long and narrow rostrum (table 64) and well­developed supraorbital ridges but with barely perceptible temporal ridges The incisive foramen of P. brevicauda is as described and figured by Patton (1987: 323 and fig. 13) for members of the longicaudatus ­ group. It is lyrate in shape, slightly to strongly constricted posteriorly, with flanged posterolateral margins (in 52 out of 53 specimens) that form a groove extending onto the

anterior palate (fig. 140). The premaxillary portion of the septum is long, the maxillary portion is keeled in most specimens (46 out of 53) and both are in contact in all individuals examined; the vomer is visible in all but one individual. The groove on the floor of the infraorbital foramen that accommodates the infraorbital nerve is barely visible and the floor of the foramen is rather smooth (51 out of 53 specimens). The mesopterygoid fossa is broad and shallow (fig. 141), matching closely specimens from central Peru´ , southern Perú, and northern Bolivia described by Patton (1987: 330 and table 4). It penetrates into the posterior palate to the posterior margins of M 3 in 12 specimens, to the posterior half of M 3 in 33, and the anterior half in 7 individuals. Most specimens have three folds in all upper and lower teeth (n = 50 and 46 upper and lower toothrows, respectively)

However, a small number of individuals (11) have four folds in pm4 and two folds in m3, a characteristic of samples from southern Perú and adjacent Bolivia (Patton, 1987: table 5). Finally, a few individuals (8) have four or five folds in a given tooth.

SELECTED MEASUREMENTS: Representative external and cranial measurements are summarized in table 64.

COMPARISONS: Externally, P. brevicauda can be readily distinguished from all other sympatric species within the Rio Juruá basin by its dark overall body color, short, dark tail, and dark dorsal surfaces of the hind feet. It is also the only species along the Rio Juruá with a fulvous lateral stripe and venter, features clearly evident even in younger­aged individuals (also see table 60). It also differs from all other species in bacular shape (fig. 137), which can be seen in live animals in the field by eversion of the phallus. In cranial characters, P. brevicauda is unique in its combination of lyrate and distally flanged incisive foramen with a complete and keeled septum (fig. 140), strongly grooved palate with a median ridge, wide and shallow mesopterygoid fossa (fig. 141), generally smooth floor of the infraorbital foramen, and relatively simple cheekteeth with three folds in each, above and below. Only P. cuvieri can be confused with P. brevicauda , as it is similar in size and has dark dorsal coloration and a lyrate incisive foramen (fig. 140) However, P. cuvieri has a white, as opposed to rufous venter, white rather than dark hind feet, a short and broad baculum with distinct apical extensions (fig. 137), and a somewhat narrower and more deeply penetrating mesopterygoid fossa (fig. 141).

MOLECULAR PHYLOGEOGRAPHY: We have sequence data from the cytochrome­b gene for 10 individuals from seven localities in northern and southern Perú as well as the Headwaters Region of the Rio Juruá (fig. 142, left; table 65). Two clearly delineated clades differ by an average of 12% (fig. 142, right). All specimens from the Rio Juruá cluster closely (average divergence of 2.3%) with individuals from the Río Cenepa and Río Santiago from Departamento de Amazonas in northern Peru´, while those specimens from southern Perú along the Río Madre de Dios (Departamento de Madre de Dios) form the second clade. As noted below, our Rio Juruá samples share the rufous venter with specimens from the upper Rio Marañón region of northern Perú, not the pure white venters of specimens from southern Perú.

We also examined relationships among 18 individuals from all Headwaters Region localities, based on 450 bp of cytochrome­b sequence. In this analysis, we were interested in any evidence for haplotype segregation relative to the side of the river from which samples were taken. We do not present these results, but simply note that no pattern of haplotype segregation was found (i.e., there was no tendency for haplotypes to cluster either by individual locality or by river margin). The average Kimura two­parameter distance among all separate haplotypes recovered from the 18 individuals was only 1.116% (0.106 standard error).

MORPHOMETRIC VARIATION: Our samples of this species are limited both in number and geographically, to the Headwaters Re­ gion of the Rio Juruá in Estado do Acre. The total number of adults (n = 53) precluded an examination of nested patterns of variation in mensural variables with regard to locality sex, and age. However, in one­way ANO­ VA’s, we found no significant locality or sex effects for any external or cranial variable (p> 0.05 in all cases). Consequently, the extent of either geographic differentiation or sexual dimorphism within P. brevicauda over this limited region is trivial.

DISTRIBUTION AND HABITAT: Proechimys brevicauda has been recorded at localities from northern Perú south to northern Bolivia (Patton, 1987: map, fig. 3; Anderson, 1997) Our data provide the first records of this species for western Brazil, based on specimens examined by us in museums in the United States and England.

We trapped individuals in approximately equal numbers on both sides of the Rio Juruá (60 and 52, on the right and left banks, respectively), but only in the Headwaters Region. The majority (78%) were obtained in terra firme forest, including naturally disturbed forest with either high densities of bamboo or in second growth, while 22% came from forest that floods only occasionally, not seasonally (table 64). We collected a few specimens with shotguns (9%), but most were trapped either in Sherman (33%) or Tomahawk live traps (58%). Young and subadults were captured in about the same numbers in both kinds of traps, but 91% of adults were caught in Tomahawk traps.

REPRODUCTION: We obtained all specimens of P. brevicauda during the rainy season in the months of February and March, during our survey in the Headwaters Region. Hence, we have no information regarding reproductive seasonality. All reproductively active males were of toothwear age classes 7 or higher; most (15 of 17) belonged to ages 9 or 10. However, while all individuals of younger age classes were nonreproductive, so were two individuals of age 8. Of the total number of females collected, 39.1% were pregnant (27 of 69). While pregnancies were recorded in animals as young as age 5, most (18) were from females in classes 9 or 10. Seven other females were parous, as evidenced by a vascularized uterus and placental scars, with two of these lactating. No pregnant individuals were also lactating. The modal litter size was 2, (range 1–4). These results agree with data for P. brevicauda from northern Perú (Patton and Rogers, 1983) where females began breeding by age class 5 and males by age 7. Young (age classes 1 to 5) and subadult (ages 6 and 7) individuals were found at each of the four standard sites in the Headwaters Region (localities 1 through 4), suggesting that breeding minimally commenced by the end of the previous dry season two to three months earlier. However, since we have no samples of this species from sites visited at other seasons, we do not know whether reproduction is seasonal or largely aseasonal.

KARYOTYPE: Available data on karyotypic diversity for P. brevicauda (sensu Patton 1987) are summarized in table 66, which only includes data for localities from which we personally examined specimens. Data are available for nearly the entire range of P. brevicauda , from northern to southern Perú and western Brazil. Diploid number varies from 28 to 30, with differences among samples in the number of morphological classes of autosomes present. We karyotyped 88 specimens combined from all sampled localities within the Headwaters Region of the Rio Jurua´. The karyotype of all individuals was consistently 2N = 28 and FN = 48, with a single pair of large metacentrics, eight pairs of medium­sized to small metacentrics and submetacentrics, two pairs of large subtelocentrics, and two pairs of medium­sized to small acrocentrics; the X chromosome is a medium­sized acrocentric and the Y is a minute one (fig. 143) This karyotype appears identical to that published by Aniskin et al. (1990, 1991) for animals from Pucallpa, Departamento de Ucayali, Perú, which surely represent P brevicauda although we have not examined the relevant specimens. All differences in the numbers of each type of autosome listed in table 66 involve chromosomes whose small size sometimes makes their correct morphological classification difficult (see Gardner and Emmons, 1984).

COMMENTS: Proechimys brevicauda was one of two species listed by Patton (1987) in his longicaudatus ­ group. He suggested that P. brevicauda and P. longicaudatus were geographic replacements of one another, documented by ‘‘... sharp character transition, particularly in pelage color and color pattern but also in bacular measurements...’’ (Patton, 1987: 338) between the upper Río Iténez and Río Mamoré of southern Beni and Santa Cruz departments, Bolivia. Gardner and Emmons (1984) also suggested that Ecuadoran specimens referable to gularis Thomas, 1911 were specifically distinct from northern Peruvian P. brevicauda based on karyotypic differences (2n = 30, FN = 48, without large subtelocentric autosomes, versus 2n = 30, FN = 48, with two pairs of large subtelocentric autosomes). They also suggested that the central and southern Peruvian popula­ tions might represent a valid subspecies of P. brevicauda , to which the name elassopus Osgood, 1944 would apply, based on both karyotypic (2n = 28, FN = 50) and color differences. Our specimens from the Rio Juruá have the rufous venter characteristic of typical brevicauda from northern Peru´, rather than the white venter of elassopus, and they group with specimens from northern Perú in cytochrome­b sequences (fig. 142). Karyotypically, the Rio Juruá samples are the same as those from the Rio Ucayali in central Perú (table 66), and differ slightly from those from southern Peru´. The deep molecular sequence divergence between the northern Peruvian and Rio Juruá clade relative to that from southern Perú (fig. 142) supports a specific separation of brevicauda and elassopus. As noted by Patton (1987: 338), ‘‘... a thor­ ough analysis of geographic variation within this group is certainly warranted.’’

SPECIMENS EXAMINED (n = 112): (1) 13 m, 27 f — MNFS 1069, 1074, 1079, 1083, 1094–1095, 1097, 1110, 1121–1122, 1127– 1130, 1132, 1137, 1139, 1152–1155, 1158– 1159, 1189, 1202–1203, 1211, 1215, 1218– 1219, 1221–1222, 1312–1313, 1325, 1328, 1397–1398, 1413, 1415; (2) 3 m, 6 f — MNFS 1175, 1177, 1253, 1285, 1300–1302, 1392–1393; (a) 4 m, 2 f — MNFS 1013– 1014, 1030, 1055, 1058–1059; (b) 2 m — MNFS 999, 1044; (c) 2 f — MNFS 1034, 1052; (3) 7 m, 11 f — JUR 204, 224; MNFS 1520, 1541, 1546, 1549–1550, 1581, 1595, 1602–1604, 1606–1607, 1634–1635, 1654, 1681; (4) 14 m, 21 f — JUR 242, 248; MNFS 1429, 1443, 1448, 1450, 1458, 1460– 1461, 1467, 1472, 1482, 1487, 1500, 1502, 1504, 1508–1510, 1522, 1524, 1552, 1562, 1567–1568, 1572, 1618, 1626–1627, 1642, 1645–1646, 1665–1667.