Proechimys simonsi Thomas, 1900

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 : 240-248

Proechimys simonsi Thomas, 1900

TYPE LOCALITY: ‘‘ Perené River , Junin Province, Peru. Altitude 800 m’ ’; Departamento de Junín, Peru´.

DESCRIPTION: A thorough description of this species was provided by Patton and Gardner (1972) under the name P. hendeei . Patton (1987) included P. simonsi as the sole member of his simonsi ­ group, based on details of the baculum and qualitative features of the cranium, with hendeei listed as a synonym. This is one of the largest species of spiny rats to occur on the Rio Jurua´, equaled or exceeded in size only by individuals of P. echinothrix and P. steerei (table 64). It is characterized by an elongated body, long and narrow face, long ears, proportionately and absolutely long tail, and large hind feet (tables 60 and 64). The tail is bicolored, with a distinct dark dorsal stripe and white ventrum; it is covered by sparse, fine hair, although the scales remain conspicuous to the eye (9–13 annuli per cm). As in most species of Proechimys , the middorsal color is somewhat darker than the sides of the body, with the rump the darkest portion of the body, coarsely streaked with black hairs and interspersed dark brown aristiforms. The aristiform hairs are long and thin with a distinctly whiplike tip (da Silva, 1998: fig. 3). The venter, chin, sides of the upper lips, undersurfaces of forelimbs, and hind limbs are pure white. The white of the inner leg extends across the tarsal joint onto the foot in most specimens, but in some, dark hair forms a ring around the ankle, interrupting the white stripe. The hair of the dorsal surface of the hind foot is usually white. Occasional specimens, however, have the tarsals, the entire first and the second digits, or just the distal end of the digits, in combination or not, covered with brownish hair.

The baculum of P. simonsi from the Rio Juruá is identical to that described and figured by Didier (1962), Patton and Gardner (1972), and Patton (1987); it is long and narrow, with a rounded and slightly broadened base (fig. 137).

Our specimens of P. simonsi do not differ in any detail from the craniodental descriptions given by Patton (1987) for this species, based on materials that he examined from Colombia, Ecuador, Perú, and Bolivia. The skull is large, the rostrum is distinctly long and narrow (figs. 138 and 139 and table 64) and supraorbital ridges are well developed but do not extend across the parietals. The incisive foramen is ovoid in shape, sometimes slightly elongated but never strongly constricted posteriorly; its posterolateral margins are mostly flat, rarely with a weakly developed flange, but even then lacking grooves extending onto the anterior palate (fig. 140). The premaxillary portion of the septum is short, rounded, and usually no more than half the length of the foramen; the maxillary portion is attenuate and usually not in contact with the premaxillary portion. The floor of the infraorbital foramen is usually grooved, with moderately developed lateral flanges. The anterior border of the mesopterygoid fossa is acutely angled (30°–35°) and penetrates deeply into the palate, reaching the anterior half of M3 or the middle of M2 (fig. 141). The counter fold pattern of both upper and lower teeth from the Rio Juruá is the same for other samples of this species (Patton, 1987), with three folds in PM4 and M1 and three to four folds in M2 and M3 as the general condition. Overall, the character trends observed in the Rio Juruá specimens are in complete agreement with those described by Patton (1987) in specimens from other localities outside the Rio Juruá Basin but throughout the range of P. simonsi .

SELECTED MEASUREMENTS: See table 64.

COMPARISONS: This is one of the two largest spiny rats found in the terra firme forests of the Rio Jurua´, the other being P. echinothrix (table 64). Externally, it can be distinguished from all other species by a combination of its large body size, absolutely and proportionately long tail, large ears and feet and relatively soft aristiforms (table 60). The tail is sharply bicolored and sparsely covered by hair; the undersurface of the body and the dorsal surface of the foot is mostly pure white, and the baculum is long and narrow Cranially, P. simonsi is easily separable from all other spiny rats by a long and distinctly narrow skull and rostrum, ovoid incisive foramen with largely incomplete septum, and smooth anterior palate, as well as by the very narrow and deeply penetrating mesopterygoid fossa.

MOLECULAR PHYLOGEOGRAPHY: We have extensive sequence data for the cytochromeb gene for this species from a substantial portion of its range, including the entire Rio Juruá drainage area and the upper Rio Urucu in western Brazil, as well as for localities in northern and southeastern Perú and northern Bolivia (fig. 148, left; table 71). The cladogram (fig. 148, right) is based on 801 base pairs for 20 haplotypes taken from localities throughout the species’ range. Sequence divergence is moderate over most of this range, although the sample from the Río Santiago in the Departamento de Amazonas in northern Perú diverges by nearly 8% from all those to the east and south. The latter samples differ by an average of nearly 5% among themselves.

Within the Rio Jurua´, we have examined haplotype variation in a 399 base pair frag­ ment of the cytochrome­b gene for 150 individuals from the 11 terra firme forest localities. A total of 47 different haplotypes were recovered, with an average divergence of 3.6%. These data are being analyzed separately by Marjorie Matocq, in conjunction with M. N. F. da Silva and J. L. Patton, with regard to the pattern of haplotype apportionment among localities along the river. A synopsis is presented in the section beyond on riverine diversification patterns.

MORPHOMETRIC VARIATION: We summarize variation in mensural characters for adult individuals according to locality, sex, and age by a nested ANOVA in table 72. For the four external and 21 cranial variables, there is virtually no geographic component, as locality explains an average of only 4.4% of the total variation. Both sexual dimorphism and age (toothwear scores of 8, 9, and 10) are some­ what more important, contributing 16.7% and 14.1%, respectively. However, when analyses were restricted to the two largest samples (Condor [locality 6], n = 39 adults, and Penedo [locality 7], 43 adults), and a nested ANOVA used to determine the relative contributions of sex and age on total variation, no variable exhibited significant sexual dimorphism, but 16 of the 25 characters showed a significant age effect. While sexual dimorphism within localities appears to be limited or nonexistent, cranial age score contributes importantly to within­locality variation. This result is consistent with the observation that growth is essentially indeterminant in other species of Proechimys , and related genera, as size continues to increase even among adult toothwear age classes (Patton and Rogers, 1983; dos Reis et al., 1990; Lara et al., 1992).

The rather small amount of variation attributable to locality comparisons is mirrored by the uniform pattern of haplotype variation in cytochrome­b sequences within the Rio Juruá basin and, in general, in qualitative morphological features throughout the species range (Gardner and Emmons, 1984; Patton, 1987). There is negligible differentiation among populations assignable to this species other than weak character clines, making P simonsi perhaps the ‘‘most consistently recognizable group of spiny rats’’ (Patton, 1987 337). However, slight but significant geographic differentiation is apparent within P simonsi along the Rio Jurua´, based on multivariate analyses. In a principal components analysis, samples pooled into the four regional sample areas overlap broadly in multivariate space (fig. 149, top), yet average individual scores for each region differ significantly along both the first and second axes (F 3,203 = 13.932, p <0.001; F 3,203 = 3.177, p <0.025, respectively). Table 73 provides the factor coefficients for the first two axes which combine to explain 73.9% of the total variance. The first PC axis is a multivariate

representation of overall size, as indicated by the high and positive factor coefficients (table 73) and the strongly positive correlations of individual scores with their respective mensural variables (data not shown). The correlation coefficient for the relationship between PC­1 scores and individual values for condyloincisive length, a univariate index of overall size, is 0.973, for example. Overall size, however, does decrease clinally from localities in the Headwaters Region to those in the Lower Central Region, with samples in the Mouth Region reversing the trend by being the largest of all (fig. 150).

Although regional samples exhibit some pattern to the character variation along the Rio Jurua´, the river itself does not appear to represent a barrier between opposite­bank populations. A nested ANOVA with region as the main effect and river bank as a sec­ ondary one yielded no significant effect due to either river bank (F 1,199 = 0.544, p = 0.461, for PC­1) or to the interaction between region and river bank (F 3,199 = 1.344, p = 0.261). The lack of differentiation due to riv­ er bank is readily apparent in a bivariate plot of principal component scores for the first two axes, with left and right bank samples pooled and plotted separately (fig. 149, bottom). However, if samples are pooled by riv­ er bank and subjected to a discriminant function analysis, which maximizes betweengroup variance while minimizing that within groups, slight segregation of opposite­bank populations is apparent. A one­way ANOVA of individual scores on the first discriminant axis yields a significant river bank effect (F 1,205 = 54.057, p <0.001), and histograms of these scores (fig. 151, left) illustrate the slight differences among the samples. Only 64.7% and 76.2% of the 207 individual specimens included in the analysis are allocated to their correct side of the river by a posteriori probabilities of group membership. Table 74 provides the standardized character coefficients from the discriminant analysis.

We also examined the relationship between the morphometric and genetic distances among our samples of P. simonsi , as well as the relationship between each of these variables and geographic distance. We used the Mahalanobis D 2 matrix generated from a discriminant function analysis that specified localities as the a priori groups as a measure of morphometric distance, and a matrix of genetic similarities (Slatkin’s [1993] M­statistic) generated from the population cytochrome­b haplotypes by the AMOVA program of Excoffier et al. (1992). These were compared to the log 10 of the straight­line geographic distances among localities given in table 1. Morphometric distances increase significantly with an increase in geographic distance among locality pairs (fig. 152; Mantel’s matrix correlation coefficient r = 0.551, p = 0.0014), and genetic similarity decreases equally significantly with geography (fig. 152; r = ‾0.518, p = 0.0021). Not surprisingly, morphometric distance and genetic similarity are intercorrelated (r = ‾0.521, p = 0.012). Populations of P. simonsi along the Rio Juruá thus exhibit a clinal, isolationby­distance pattern in both morphometric and genetic traits.

DISTRIBUTION AND HABITAT: Proechimys simonsi occurs throughout the western Amazon Basin from southern Colombia to northern Bolivia (fig. 148, and Patton, 1987; Anderson, 1997). The specimens we record here, as well as those obtained by us on the upper Rio Urucu southeast of Tefé in Estado do Amazonas represent the first records for the species from Brazil known to us. We found P. simonsi to be common along the length of the Rio Jurua´, and well represented in the fauna of all four regional areas (table 63). The great majority of individuals was collected in terra firme forest (78.3%; 310 out of 396), 23 (5.8%) in várzea forest, and 63 (15.9%) in a variety of terra firme habitats, such as secondary or disturbed forest gardens, or terra firme–várzea edge. Interestingly, all but one of the individuals trapped in the várzea are from the Headwaters region, where the flooding regime of the várzea forest is sporadic across years, not annual Most specimens (79%) were caught in Tomahawk traps, with only 22% in Sherman traps; 6% were shot and 3% were caught in snap traps or by hand. Of the animals captured in live traps, 39% of young and sub­

adults were caught in Sherman traps and 61% in Tamahawk traps; 90% of the adults were collected in Tomahawk traps and only 10% in the smaller Sherman traps.

REPRODUCTION: We obtained specimens of P. simonsi along the entire river and at all seasons of the year. Of the 117 males with complete autopsy data, 66 were reproductively active by our criteria. These ranged in age class from 6 to 10, although the majority (79%) were relatively old adults (age classes 9 and 10). All young individuals (ages 1–5) and most subadults (ages 6 or 7) were reproductively inactive, although this category also included some age class 9 animals. Detailed reproductive data are available for 154 individual females (table 75); of those, 65% were either pregnant or postpartum whereas 35% had apparently not yet reproduced. Pregnant, lactating, or postpartum females were obtained at all sites, suggesting that at least some females breed in all seasons over the year. Of the parous individuals, almost all (90 of 93) are adults (age classes 8–10, with only four of age class 8), and only three are subadults (age class 6). Sixty­seven females were pregnant. The majority of these (91%) were full adults of ages 9 or 10, although pregnancies were observed at ages as young as 6. Four of the 67 pregnant females were also lactating, suggesting that a postpartum estrous is possible, if uncommon, in this species. The modal litter size was 2, with a range from 1–3. Individuals recorded as nulliparous were present in all age categories except 10; subadults and adults comprise almost half of those (23 of 50), with the proportion of nulliparous adults (age class 8 and 9) around 30%. Both males and females appear to reproduce only after they are fully grown, and a relatively large proportion of adults in a population does not reproduce at a given time.

KARYOTYPE: 2n = 32; FN = 58. The autosomes consist of two pairs of large submetacentrics, eight pairs of medium­sized to small metacentrics and submetacentrics, one pair of large and three pairs of medium­sized to small subtelocentrics (one of which has a secondary constriction on the long arm), and a single pair of medium­sized acrocentrics. The X chromosome is a medium­sized ac­ rocentric and the Y is a minute one. This karyotype was previously described and figured by Patton and Gardner (1972), and is apparently invariant throughout the species range. Table 76 provides a summary of karyotype data for specimens which we have personally examined. In addition to these specimens, the same karyotypes have been published for specimens presumptively of this species from southern Colombia (Reig and Useche, 1976). We have not examined these specimens.

COMMENTS: Patton (1987) remarked that the P. simonsi group showed relatively low levels of morphological character variation throughout its geographic range, and available data suggested karyotypic uniformity as well. Our samples from the Rio Juruá and upper Rio Urucu extend the geographic distribution of this species approximately 1000 km to the east, as mapped by Patton (1987), and are fully consistent with this view of uniform character variation. As noted above, the cytochrome­b sequence data also generally support the interpretation of uniformity throughout the majority of the species’ range However, the single sample from northern Perú is rather divergent, and additional samples from the northern portion of the range of this species are needed to determine the

extent to which P. simonsi is structured into regional, reciprocally monophyletic lineages.

SPECIMENS EXAMINED (n = 416): (1) 21m, 31f — MNFS 1071–1073, 1078, 1081–1082, 1084, 1086, 1089–1092, 1099, 1108, 1112, 1114, 1124–1126, 1133–1136, 1140–1142, 1156–1157, 1162–1164, 1197, 1199, 1209, 1212–1214, 1216–1217, 1288–1289, 1314– 1317, 1324, 1357, 1359, 1364, 1383, 1386– 1387, 1414; (2) 6m, 4f, 1 unknown — MNFS 1176, 1178–1179, 1284, 1336, 1349– 1351, 1376–1378; (b) 2m — MNFS 1000– 1001; (3) 6m, 11f — JUR 205, 222, 229– 230, 239, MNFS 1542, 1544, 1551, 1593– 1594, 1600–1601, 1605, 1653, 1656–1657, 1680; (4) 16m, 29f — JUR 217, 232, 234– 235, 243, 250, MNFS 1427, 1431–1434, 1440–1442, 1449, 1451, 1456–1457, 1468– 1471, 1473, 1483–1485, 1488–1489, 1501, 1503, 1505, 1511, 1523, 1525, 1536, 1561, 1563–1564, 1570–1571, 1647, 1661, 1663– 1664, 1668; (6) 28m, 25f, 1 unknown — JLP 15530–15531, 15533, 15539–15540, 15547, 15560, 15573, 15576, 15578, 15589–15595, 15610–15611, 15613, 15617–15618, 15621– 15622, 15639–15641, 15643–15644, 15649, 15656–15659, 15668–15669, 15681, 15684, 15702, JUR 177, 184, MNFS 538–540, 542– 544, 547, 555–559, 561; (7) 40m, 18f — JLP 15261, 15264, 15270, 15276–15277, 15279– 15280, 15286–15287, 15293–15300, 15307, 15309, 15315–15318, 15320, 15333–15339, 15347–15352, 15363, 15370–15374, 15464, 15506, MNFS 339–340, 344–345, 347, 350– 352, 355, 362, 364, 406, 476; (9) 18m, 23f — JLP 15928, 15991–15998, 16015–16018, 16038, 16041, 16055, 16082–16083, 16086– 16090, JUR 195, MNFS 852, 855–856, 858, 860–864, 878–880, 884, 886–887, 912–913; (12) 13m, 24f — JLP 15747, 15764, 15778, 15785–15787, 15797, 15809–15812, 15817– 15820, 15834, 15855, 15873–15874, 15887– 15888, 15904 –15905, 15907, JUR 189, MNFS 723–724, 726, 732, 741, 743–744, 762–764, 819–820; (13) 1f — JUR 346; (14) 14m, 43f — JUR 428–429, 431–432, 434, MNFS 1684, 1687–1693, 1695–1697, 1702, 1705–1713, 1720–1722, 1726–1733, 1740– 1741, 1746–1748, 1758–1762, 1773–1776, 1778, 1781–1782, 1788–1790; (15) 23m, 16f — JUR 270–272, 286, 289, 294, 299, 302– 303, 318, 320, 323, 344, 359, 362, 365–368, 371–373, 376, 379–380, 382–383, 387–389, 391–392, 400, 402, 405, 407–410; (16) 2f — MNFS 1751, 1754.