Proechimys kulinae da Silva, 1998

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 : 236-238

Proechimys kulinae da Silva, 1998

TYPE LOCALITY: ‘‘ Brazil: Amazonas ; Seringal Condor, left bank Rio Jurua´ , 70°51̍W, 6°45̍S.’’

DESCRIPTION: Proechimys kulinae , P. gardneri ,and P. pattoni are the smallest spiny rats occurring in the Rio Juruá basin. In general morphology these specimens of P. kulinae are relatively slight in build, have small ears, and moderately short tail and hind feet (tables 60 and 64). Overall, the color of the body is uniform reddish brown, coarsely streaked with varying amounts of black on both the dorsum and sides. The dorsal pelage is interspersed with thick, dark brown aristiform hairs that form a darker medial band contrasting with the sides of the body. The venter, chin, and undersurfaces of fore and hind limbs are pure white; the upper lips are dark, generally lacking patches of white hair; the tarsal joint is either ringed by dark and rusty­colored hair, or the tarsal ring is interrupted by white hair confluent with that of the undersurface of the hind limbs and feet; the hind foot, including the digits, is white, with some golden tones in most individuals. The tail appears almost naked, distinctively bicolored with dark brown dorsum and white venter. In juveniles, the color of the pelage varies from dark grayish brown (age class 2) to pale brownish mixed with rusty­colored hair (age class 5). Five plantar tubercles are present in the majority of specimens (12 of 16), with the lateral metatarsal tubercule missing. The baculum is elongate and relatively narrow, with stout and short apical extensions; the proximal and distal ends are about equal in width (fig. 137).

The skull is relatively small, with a short, narrow rostrum (figs. 138 and 139) and a well­developed supraorbital ledge extending onto the anterior portion of the parietals. The postorbital process of the zygoma is well developed and formed mostly by the squamosal. The floor of the infraorbital foramen is generally smooth, without a demonstrable groove for the maxillary nerve. The incisive foramen is mostly square to oval in shape (fig. 139), with nearly flat posterolateral mar­ gins; the anterior palate is smooth, lacking grooves extending posteriorly from the incisive foramen, and lacking a median ridge the premaxillary portion of the septum is short, extending for less than half the length of the foramen; the maxillary portion is variable, attenuate to expanded anteriorly, and usually in contact with the premaxillary portion; the vomer is either completely enclosed or only barely visible. The mesopterygoid fossa is narrow (fig. 141), with an angle of indentation into the posterior palate averaging 57°; it is moderately deep, usually extending to the anterior half of M3. All upper cheekteeth have three folds except M3 which usually has three but may have only two.

SELECTED MEASUREMENTS: Means and ranges of selected external and cranial measurements are given in table 64.

COMPARISONS: Overall, P. kulinae is most similar to the other small species, P. pattoni and gardneri (figs. 135 and 136). Interestingly, these three apparently replace one another along the Rio Jurua´, as no combination of these species occurs sympatrically at our various sample sites— P. pattoni is known only from the Headwaters Region, P. gardneri from the right bank in the Lower Central Region, and P. kulinae is known only from left bank localities in the Upper and Lower Central regions; fig. 133). Proechimys kulinae does co­occur with P. cuvieri , P. simonsi and P. steerei , all three of which are much larger in body size with absolutely larger hind feet, ears, and tail (fig. 133; table 64) The baculum of P. kulinae is most similar in shape to those of P. simonsi and P. steerei (fig. 137), especially so relative to P. simonsi , although shorter and with somewhat more expanded apical wings. Along with P. pattoni , this is the smallest species of Proechimys currently known from Amazonia, with a total length less than 330 mm. It has wider and longer aristiform hairs than either P gardneri or P. pattoni (da Silva, 1998: table 4 and fig. 3); usually has only five, instead of the six plantar tubercles characteristic of the other two small species; and a stronger more spinose postorbital process of the zygoma. Proechimys kulinae differs from all known species, including both P. gardneri

and P. pattoni , in its 2n = 34, FN = 52 karyotype.

MOLECULAR PHYLOGEOGRAPHY: Sequence data are available from only the two locations along the central portion of the Rio Jurua´. Haplotypes from individuals from the same population are closely similar, varying by no more that 0.1 to 0.4%, but individuals from the two known localities are quite divergent, differences averaging nearly 10% (fig. 147). This is a striking degree of divergence, particularly for localities only 238 km apart and on the same side of the river (table 1 and fig. 133). Additional data from other populations will be needed to assess the significance of this level of divergence. The species presumably is more widely distributed as, based on morphological characters, da Silva (1998) assigned individuals from northeastern Peru´, south of the Amazon, to this species. Unfortunately, we lack tissue samples from these individuals so that their degree of molecular divergence remains unknown.

DISTRIBUTION AND HABITAT: This species is known from only two localities in the Upper and Lower Central regions (Condor [locality 6] and Barro Vermelho [locality 12]) on the left bank of the Rio Juruá in the Estado do Amazonas, Brazil, and two localities in the Departamento de Loreto, Perú (da Silva, 1998). All specimens collected along the Rio Juruá are from primary or second­growth terra firme forest (table 63). In our sample of 33 individuals, the majority (52%) were caught in Tomahawk live traps, 27% in Shermans, and 21% were caught in snap traps Of those captured in both types of live traps twice as many young and subadults (4 out of 6) were caught in Shermans, and three times the number of adults were captured in Tomahawk traps (15 out of 20).

REPRODUCTION: We obtained all specimens of P. kulinae towards the end of the dry season in the months of September and October Eight of the 14 male specimens we classified as reproductively active, the other six as inactive. All reproductively active males were full adults of age classes 9 or 10 whereas the inactive ones ranged in age from 3 to 9. Five of the 18 females were pregnant, and all were adults (age classes 8 to 10). Evidence of estrous activity (swollen and vascularized uteri) was observed in one age class 7 female, and all parous females were old adults The modal litter size was 1, and the range 1– 2

KARYOTYPE: We karyotyped 20 of the 33 individuals trapped, two from Barro Vermelho and 18 from Seringal Condor. The karyotype of P. kulinae (illustrated in da Silva 1998) has a 2n = 34 and FN = 52. The autosomes comprise a pair of very large metacentrics, six pairs of medium­sized to small metacentrics and submetacentrics, one pair of large and one pair of medium­sized subtelocentrics (the latter has a secondary constriction on the longer arms), one pair of large and five pairs of small to minute acrocentrics. The X­chromosome is a moderately small metacentric and the Y­chromosome is a minute submetacentric. This is the first Proechimys from Amazonia reported with 34 chromosomes. This same karyotype is reported for specimens of Proechimys sp. from northern Perú by Aniskin (1993). These likely represent P. kulinae , as da Silva (1998) allocated specimens from nearby localities to this species.

Specimens Examined: (n = 33): (6) 13 m, 16 f, 1 unknown — INPA 2553–2557; JLP 15534, 15612, 15660 ; MPEG 25502–25506 MVZ 187184–187190, 187192; JUR 178–

182, 185; MNFS 546, 552, 554; (12) 1 m, 2 f — INPA 2558; JUR 186; MVZ 187193.