Oryzomys perenensis (Allen, 1901)

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 : 150-155

Oryzomys perenensis (Allen, 1901)

TYPE LOCALITY: ‘‘ Perene´ , Department of Junin, Peru ; altitude 800 m’’; given by Musser et al. (1998: 51) as ‘‘ Valle Perene´ , Colonia del Perene´ ; a coffee plantation at junction of ríos Paucartambo and Chanchamayo’ [Stephens and Traylor, 1983: 161], 10°51̍S/ 75°13̍W, 1000 m.’’

DESCRIPTION: See discussion, above, tables 38 and 40, and figure 101. Musser et al. (1998: 18–42) provide thorough analyses of qualitative and quantitative morphological variation along with excellent illustrations of crania and teeth of this, and other species of the O. ‘‘ megacephalus ’’ complex.

MORPHOMETRIC VARIATION: The proportion of variation in mensural dimensions of the skin and cranium of adults (age classes 3 and above) due to sex, age, or geographic locality was examined by an ANOVA nested by lo­ cality, sex, and age (table 42). Sexual dimorphism is virtually non­existent, accounting for an average of less than 1% of the total pool of variation. In a similar fashion, variation due to locality is also minimal, with an average of only 2.6% due to differences among localities along the 1000 km length of the Rio Jurua´. However, a substantial degree of the total variation (31.4%) is related to an individual’s age, regardless of sex or locality. Indeed, all variables exhibited a significant relationship with age (p <0.001) except IOC and MTRL (fig. 102). While pooling sexes is justifiable because of the lack of demonstrable sexual dimorphism, ordinarily some adjustment for the large age effect might be required in comparisons between localities (see, for example, Myers et al., 1990, for such an ‘‘adjustment’’ in geographic analyses of Andean mice of the ge­

nus Akodon ). However, as there were no differences among those samples of reasonable sizes (n> 20) in their respective age distributions (X 2 = 13.417, df = 13, p = 0.3395), we pooled all individuals regardless of their ages and sex. Interlocality comparisons based on simple pooled or on age­corrected data did not differ in their results.

Musser et al. (1998) examined the pattern of variation among our samples from the Rio Jurua´, and for a set of cranial variables similar to those we measured. Based on a standard principal components analysis (PCA), they found no pattern of differentiation, and particularly none associated with samples from opposite banks of the river. That is, there was no measurable ‘‘riverine’’ effect. We used the multiple groups principal components analysis (Thorpe, 1983; Thorpe and Baez, 1987) to ask the same question. This method has the advantage over ordinary PCA in that it does not confuse the within­ and among­group variation when several groups are used; rather, MGPCA gives pooled within­group components, with the first axis derived from the variance­covariance matrix usually an index of overall size The results of the multiple groups PCA (fig 103) are identical to those of Musser et al (1998). There is no statistical difference in pairwise comparisons between opposite­bank samples, nor is there any difference between samples for the four geographic regions (p> 0.05 in all comparisons of individual scores for any of the MGPCA axes). Furthermore only 42.4% of individual specimens are correctly allocated to their appropriate locality based on posterior probabilities stemming from a discriminant analysis. Not surprisingly, therefore, there is also no relationship between matrices of morphometric Mahalanobis distances and the geographic distances among localities (Mantel’s matrix r = 0.0905; t = 1.220; p = 0.888).

MOLECULAR PHYLOGEOGRAPHY: The lack of demonstrable geographic differentiation in morphometric variables among samples of O. perenensis collected along the nearly 1000 km length of the Rio Juruá is completely concordant with the pattern of mtDNA cytochrome­b haplotype variation these same samples exhibit (Patton et al., 1996a). For the latter, hierarchical analyses showed that the majority (89.2%) of the molecular variance was apportioned among individuals within populations, and that only a small fraction could be either attributed to differences among localities within one of the regional samples (6.6%) or even between the regions themselves (4.2%). There was also no evidence of isolation by distance, as there was no significant relationship between logM and logDistance (Mantel’s matrix r = 0.094, t = 0.802, p> 0.788; Patton et al., 1996a). Either populations of this species throughout the Rio Juruá have recently entered and expanded within the river basin, and thus have not yet achieved genetic equilibrium, or local populations are, and have been, linked by considerable gene flow (see discussion in Patton et al., 1996a). The second possibility seems likely, since estimates of the gene flow parameter, M (Slatkin, 1993), were uniformly high, averaging 17.34 across all among­population comparisons. To put this value into perspective, it only takes a single successful migrant per generation (M of 1.0) to prevent two populations from diverging by drift alone (see review by Mills and Allendorf, 1996). Both the large number of haplotypes found (47 within a total sample of 158 individuals) and the large number of unique, but low­frequency haplotypes recovered from each population and geographic region suggest that the genetically effective population of this species has been historically large and remains so today. These conclusions are in harmony with what few data are

available on reproductive potential and habitat range, as summarized next.

DISTRIBUTION AND HABITAT: Individuals of O. perenensis were found in virtually every terrestrial habitat present along the river, even occasionally in the dense grass growing on exposed sand bars during low water seasons. However, this species was twice as common in várzea forest than terra firme, or other habitats (table 43). Its numbers in relatively open, disturbed habitats, such as those along the river margins, were exceeded only by Oligoryzomys microtis . It was also uniformly the most common terrestrial rodent in both our terra firme and várzea standardized plots, being exceeded in numbers at particular sites only by one or more species of spiny rats, Proechimys . The species was captured only in traps placed on the ground never in those placed a meter or two in the lower vegetation, nor in the canopy platform traps on the standardized lines. The high densities and broad habitat tolerances of O. perenensis along the Rio Juruá is typical of our and others, experiences with this species elsewhere within western Amazonia (e.g., the Río Cenepa, Departamento de Amazonas and Balta, Departamento de Ucayali, Perú [J. L. Patton, personal observations], and Cuzco Amazónico, Departamento de Madre de Dios, Perú [Woodman et al., 1995]).

REPRODUCTION: We caught reproductively active individuals of both sexes at each site and during every survey month from August through June. At each locality and regardless of month, nearly every adult male had scrotal testes and enlarged vesicular glands (> 16 mm in length, maximum 24 mm) indicative of breeding activity. This suggests that males are competent throughout the year. More than 75% (79/105) of the total sample of adult females (those of age class 3 or older) were pregnant, and pregnant females comprised more than 50% of all adults at each site. The modal litter size, based on fetal counts, was 4; range 2–5. Pregnancy rates were lowest during the months of August through September in the Upper Central Region (13 of 23 adult females, 56.5%), but much higher in all other sampling periods and areas ( Lower Central Region , October and November , 19 of 24, 79.2%; Headwaters Region , February and March , 33 of 40, 82.5%; Mouth Region , May and June , 14 of 18, 77.8%). All remaining nonpregnant adult females were either lactating or parous, with evident placental scars indicating relatively recent pregnancies. As with males, therefore females probably breed yearround. Finally juveniles of both sexes in sparse, gray pelage, were captured at all sampling periods throughout the year, although not at every site, again supporting rather continuous reproductive activity spanning both dry and rainy seasons. A similar pattern of reproductive activity was observed for O. megacephalus in French Guiana (Henry, 1994) .

Both males and females begin breeding at early ages, at least based on the association of reproductive state and toothwear age classes. Males apparently do not reach reproduc­ tive competency until age class 2, and even a small fraction of older individuals were apparently not breeding (fig. 104). Females, however, entered what appeared to be their first estrous even before their third molars were fully in place (age class 1), and more than 25% of age class 2 individuals were pregnant (fig. 104). If growth in rice rats is at all similar to that of deer mice (e.g., Peromyscus truei ; Hoffmeister, 1951), breeding commences within one month after birth in both sexes.

Males were invariably more commonly trapped than females in all months (and therefore at all individual sites), although the proportion of females in each sample was higher when the pregnancy rates were highest. Perhaps males are more exploratory, or have larger home ranges than females, and females with young in the nest make fewer nightly movements than do those that are pregnant.

KARYOTYPE: 2n = 52, FN = 62. The autosomal complement consists of 25 acrocentric chromosomes grading in size from large to small and six pairs of small metacentric or submetacentric elements. The X­chromosome is a medium large acrocentric chromosome, the Y is small and acrocentric. This karyotype was described and figured by Gardner and Patton (1976) based on specimens from localities in eastern Perú, and individuals with the same karyotype have been reported from central Perú and eastern Ecuador (Musser et al., 1998: table 13). Seventy­six individuals were karyotyped, from the following localities: Porongaba (locality 1), n = 18; Nova Vida (locality 3), n = 6 Sobral (locality 4), n = 3; Sacado (locality 5), n = 4; Condor (locality 6), n = 4; Penedo (locality 7), n = 3; Boa Esperança (locality 9a), n = 1; Jainu (locality 11), n = 3; Barro Vermelho (locality 12), n = 6; Vira­Volta (locality 14), n = 22; Vai­Quem­Quer (lo­

cality 15), n = 3; and Ilhazinha (locality 16), n = 3.

SPECIMENS EXAMINED (n = 466): (1) 21m, 17f — MNFS 1100, 1115–1116, 1120, 1143–1146, 1148–1149, 1168–1170, 1173, 1204–1207, 1224–1227, 1229, 1259, 1264, 1268, 1296–1297, 1309, 1329–1330, 1381– 1382, 1400, 1418, 1421–1423; (2) 23m, 24f — MNFS 1180, 1239, 1241–1244, 1246, 1248–1249, 1251, 1276–1282, 1304–1307, 1334–1335, 1340–1346, 1348, 1367–1375, 1388, 1390–1391, 1405–1408; (3) 13m, 16f, 1 unknown — JUR 213, 228, 237, MNFS 1554, 1558–1560, 1582–1587, 1597, 1609, 1612–1614, 1622, 1629, 1631–1632, 1641, 1648–1649, 1651, 1675–1678; (4) 9m, 10f — JUR 216, 218–219, 233, 246, MNFS 1436, 1454, 1462, 1464, 1466, 1480, 1492, 1497–1498, 1534, 1566, 1574–1575, 1669; (5) 27m, 19f, 4 unknown — JUR 132, 135, 144–156, 158–169, 171, 173–175, MNFS 582, 586–588, 621, 629–635, 643–646, 655– 657; (6) 15m, 10f — JLP 15529, 15536, 15564, 15604, 15609, 15646–15647, 15663, 15667, 15687–15689, 15694–15697, 15706– 15707, 15719, 15723, 15726–15729, 15740; (e) 1f — JLP 15713; (7) 34m, 11f, 4 unknown — JLP 15229–15232, 15238–15240, 15243, 15248–15249, 15256, 15259, 15263, 15272, 15274, 15291, 15302, 15304, 15311, 15322, 15330–15332, 15362, 15427, 15441– 15442, 15455–15456, 15480, 15499, 15507, MNFS 329–330, 385, 389, 404–405, 420– 423, 488–490, 498, 510, 518, 520–521; (8) 18m, 11f, 2 unknown — JLP 15415–15422, 15447, JUR 1, 4, 6, 10–11, 14, 33–34, 36, 38–40, 42–43, 47, 72, 76–79, 114, MNFS 440; (9) 4m, 5f — JLP 15967–15968, 15989, 16026–16029, 16067, 16081; (9a) 6m, 3f — JUR 190, MNFS 922, 924, 930, 937–938, 960–962; (10) 4m, 3f — JLP 16030, MNFS 896–897, 917, 952–954; (11) 13m, 11f — JLP 15752, 15758, 15822–15825, MNFS 693–694, 696, 698–699, 705–707, 712–713, 716, 751, 765–768, 786–787; (12) 40m, 15f — JLP 15748, 15762–15763, 15768–15773, 15782–15783, 15790, 15813, 15828, 15839– 15845, 15865, 15871–15872, 15875–15877, 15881–15884, 15892–15893, 15901, MNFS 682, 685–686, 736–738, 750, 761, 774–779 807–808, 818, 821–822, 827–828; (13) 1f — JUR 263; (14) 31m, 16f — JUR 418–427 441–444, 446–447, 454–456, 465, 473, 475 481–482, 490–493, 514–516, 522–523, 525 531, 536, 552, 554, 556, 558, 561–562, 568– 570; MNFS 1685, 1791; (15) 2m, 2f — JUR 288, 295, 300, 394; (16) 10m, 10f — JUR 510–513, 527–530, 537–539, 551, 553, 563– 564, MNFS 1785–1786, 1794.