Population density and reproduction of two Peruvian leaf-eared mice (Phyllotis spp.)

Population density and reproduction were studied in two leaf-eared mice ( Phyllotis spp.) in the Peruvian locality known as Lomas de Lachay. Density varied between 0-3 individuals ha -1 in P. andium, and between 0-12 individuals ha -1 in P. amicus. An annual cycle of population abundance was observed in both species. Phyllotis amicus howed no seasonal reproductive activity, but P. andium had seasonal reproduction with pregnant females principally observed between July and September. Differences in juvenile weight were related to litter size. It is likely that two young was the optimal litter size in the harsh and seasonal xeric environment of Lomas de Lachay. Our results suggest that P. amicus maintains low, highly dispersed populations in a harsh environment, and can reproduce at any time of the year, whereas P. andium appears to be a more irruptive species closely tied to rainfall patterns.


INTRODUCTION
Leaf-eared mice (genus Phyllotis, Sigmodontinae) are one of the most abundant small mammals in the Peruvian Andes.Eight of 13 species inhabit this area, where a predominance of xeric habitats exists with sparse vegetation (Pearson 1958, Steppan 1995, 1998).Ecological studies of the Peruvian species of Phyllotis are scarce and reproduction data have been only occasionally reported.Two species of leaf-eared mice (P.amicus and P. andium) can be found in a unique seasonal habitat, the Lomas of Lachay, located on the central coast of Peru, 105 km north of Lima.
Here, we present information on population Population density and reproduction of two Peruvian leaf-eared mice (Phyllotis spp.) density, movements, and reproduction (in natural and captive conditions) for these mice species.

MATERIAL AND METHODS
The fieldwork was done in the Teatinos pass, located on the western side of the Lachay National Reserve (11°21'00"-11°21'58" S, 77°21'18"-77°22'25" W).Mean annual precipitation in the last four years (1998)(1999)(2000)(2001)  Reproductive activity of both Phyllotis species was assessed from examination of captured animals inside and outside grids.In addition, individuals of both species of Phyllotis (captured outside grids) were kept and bred in captivity.Each animal was housed in small cages 30 x 18 x 13 cm using wood shavings and paper as bedding materials.Animals were fed with fresh carrots, lettuce, sweet potato, sprouted wheat, and corn.Water was not provided to the animals after the first months as they never drank.Animals were maintained at an average temperature of 23 ± 5 (± SD) °C.For reproduction, one male was placed in a large cage (36 x 32 x 20 cm) for at least two days before introduction of the female.Pairs were maintained together for two weeks, after which the females were returned to small cages whereas the males were maintained in the large cages until a new mate was introduced.Animals were checked daily to record births and deaths in the colony.To analyze the effect of litter size on the weight of young, we recorded body weight each 10-20 days for 90 or more days after birth.Regression analyses between weight versus age were conducted for each litter size.Homogeneity of slopes were tested, and analyses of covariance were conducted to test for statistical differences (P < 0.05) among the regressions.

RESULTS
Population densities of both Phyllotis species are shown in Fig. 1.Very small numbers of individuals per ha were observed during almost the entire study.Only P. amicus in the low grid showed a higher population density (12.3 individuals ha -1 ) between May and July of 2001.Phyllotis andium was more abundant in the high grid than in the low grid, and exhibited the highest numbers between July and August of 2000.Six P. amicus individuals were recaptured for more than three months and one for eight months.Using data from these individuals, movement distances varied between 22.8 and 78.8 m (X = 44.5 ± 17.9 m).Only one P. andium individual was recaptured one month later.Using the three P. andium individuals captured at least two times in the same month, the movement distance between captures varied between 40.8 and 68.8 m (X = 50.5 ± 12.9 m).
In captivity, 76 and 12 matings were conducted between individuals of P. andium and P. amicus, respectively.Twenty litters with a total of 45 young were obtained in P. andium, whereas four litters with eight young were obtained in P. amicus.The gestation time in both species was 24 days.Litter size varied between one and four young in P. andium and between one and three pups in Phyllotis amicus.Fig. 2 shows the number of pregnant females captured in Lomas de Lachay each month.Phyllotis amicus showed no seasonal reproduction, with pregnant females throughout the year.In contrast, P. andium showed clear seasonal reproductive activity between July and September.In the laboratory, 35 % of female P. andium that became pregnant had litters with two young, whereas 30, 20 and 15 % had litters of one, four and three young, respectively.An ANCOVA showed significant differences in the regressions between weight and age with different litter size (Fig. 3), indicating a relationship between weight and litter size.Heavier animals were from litters of one young whereas smaller animals came from litters with four young.Young (irrespective of litter size) were weaned 30 days after birth.Mean weight of young at weaning was 14.4 ± 2.6 g.Young reached mean weights of 17 ± 3.4 and 20.8 ± 4.7 g during their second and third months of life, respectively.In captivity, females less than seven months of age were not reproductively active.

DISCUSSION
This is the first report on the population ecology of two species of leaf-eared mice in the Lomas de Lachay, and the second that comes from Peruvian lomas (Péfaur et al. 1979, Pearson 1975).Pearson & Ralph (1978) monitored Phyllotis limatus in a 2.02-ha grid in a loma near Tacna, 1,200 km south of Lima.They found that the population density  Número de hembras preñadas capturadas mensualmente en las lomas de Lachay.La parte negra de las barras indica el número de hembras preñadas y la parte blanca muestra el número de hembras no preñadas.
POPULATION DENSITY AND REPRODUCTION OF PHYLLOTIS was 6.5 individuals ha -1 , but considered this value to be an overestimate because the study occurred in November when the breeding season was ending and many young Phyllotis were in the population.
After excluding these young, the population density was 3.9 individuals ha -1 .Our results with P. andium and P. amicus showed similar values of density and suggest an annual cycle (Fig. 1), a pattern that has been previously reported for several South American rodents (Crespo et al. 1970, González et al. 1982, Murúa 1983, Murúa & González 1985, Murúa & González 1986, Murúa et al. 1986).Higher population densities were observed during the wet season (June-August) in both species.One interesting feature was the absence of individuals of P. andium during most of the study period in both grids.When traps were placed in other areas of Lomas de Lachay, the result for P. andium followed the same trend observed in the grids.Therefore, only occasional individuals appear to be present in the lomas except between July and August when their numbers increase.Captures of P. amicus outside of the grids were very low, and the individuals trapped were always in the most xeric and sparse habitats at low elevations in Lomas de Lachay.This suggests some spatial segregation between these two species of Phyllotis.Phyllotis andium occupies most of the area when in low numbers whereas P. amicus occurs in more restricted, xeric areas in the lowest parts of Lomas de Lachay.This segregation agrees with findings of Pizzimenti & De Salle (1980) who also captured P. amicus isolated from its congeners.Both species of Phyllotis had nocturnal activity, and showed an aggregated distribution throughout the year suggesting a rather social type of spatial organization, which could be the result of some limiting factors such as substrate or shelter.Mean distances traveled were 44.5 and 50.5 m for P. amicus and P. andium, respectively.Pearson & Ralph (1978) found a movement distance of 36 m for P. limatus at Morro Sama, a loma in southern Peru.Although this value is within the range of distances recorded in P. amicus (22.8-78.8 m) and close to the minimum value found in P. andium (40.8 and 68.8 m), it is likely an underestimate due to the grid being trapped for only six nights in Morro Sama.Nevertheless, these results suggest that Phyllotis spp.restrict their movements in lomas-type ecosystems.
Reproductive activity differed between the two species of Phyllotis (Fig. 2).Pregnant females of P. andium were found mainly between July and September, whereas pregnant females of P. amicus were found throughout the year.These results Fig. 3: Regression lines between weight and age for different litter size in P. andium.

ARANA ET AL.
suggest that P. andium reproduce seasonally and that breeding coincides with the wet period (winter; with precipitations above to 1.0 mm monthly).Phyllotis amicus lacked seasonal reproductive activity.A similar result was found by Williams (1990), who captured two pregnant P. amicus females in Lomas de Lachay during May and September.Differences between the reproductive patterns of P. andium and P. amicus may be related to their food habits.Although both species have broadly omnivorous feeding habits, P. amicus eats more insects that any other species of Phyllotis (Pizzimenti & De Salle 1980), and therefore, this species may have a reproductive activity with a lesser dependence on vegetation.This finding is relevant due to the higher population density of P. amicus during May and July, which does not seem to be due to juvenile recruitment.One untested hypothesis is that mobility increases in the population, as it had been reported in other sigmodontine rodents (Murúa & González 1986).
Differences in the weight of juveniles born in captivity are related to the litter size as shown in Fig. 3. Young from a litter of 1-2 individuals reach a higher weight in less time, thus likely increasing their survival rate.This suggests that there is a balance between the number and survivorship of young (as measured by growth and weight).Since the majority of pregnant P. andium females had a litter size of two, such figure could be the optimal litter size under the regular climatic conditions at Lomas de Lachay.However, during anomalous climatic conditions when food resources are over abundant (i.e., after El Niño Southern Oscillations, ENSO events), outbreaks of some Phyllotis species may occur (Pearson 1975, Jaksic et al. 1977, Péfaur et al. 1979, Meserve et al. 1995, Lima et al. 1999a, 1999b).Meserve et al. (1995) reported a dramatic increase in the population size of P. darwini in north-central Chile as a consequence of the 1991-1992 ENSO event.The population size of this species never exceeded 9 individuals 0.56 ha -1 between 1989 and 1991, but reached a maximum of 25 in February 1992 and of 39 in May 1993.On the other hand, P. darwini reached 200 individuals ha - 1 during summer 1988 in Aucó, north-central Chile (Lima et al. 1999a(Lima et al. , 1999b)).Given that P. darwini is capable of producing many litters within a short period of time (Meserve & Le Boulengé, 1987), and that the litter size in this species varies between 4.2 (Walker et al. 1984) and 5.2 (Fulk 1975, Meserve & Le Boulengé 1987), the rapid increase of population size during an outbreak ("ratada") indicates that Phyllotis spp.have a reproductive potential largely constrained by environmental factors.Therefore, in a xeric, harsh, seasonal environment such as lomas ecosystems, a small litter size of two provides greater survival potential than larger litter sizes (4-6).
Our results showed no reproductive activity of females younger than seven months in P. andium.This may explain the relatively delayed response in the population changes of Phyllotis as a biological effect of ENSO events (Pearson 1975, Pefaur et al. 1979, Meserve et al. 1995).Other recent reports have suggested an interaction between delayed density-dependent factors and rainfall at demographic and population dynamic levels as the key element for understanding outbreaks of P. darwini in semiarid regions of Chile (Lima et al. 1999a, 1999b, Lima & Jaksic 1999a, 1999b).Our short-term study did not allow us to evaluate this hypothesis in the species here studied.
The apparent year-around reproduction of P. amicus (in contrast to P. andium) and its very low population level may be a strategy that enables this species to persist in the lomas regardless of environmental conditions.Indeed, our results suggest that P. amicus is a species that maintains low, highly dispersed populations in a harsh environment, and can reproduce at any time of year, whereas P. andium appears to be a more irruptive species closely tied to rainfall patterns.

Fig. 2 :
Fig. 2: Number of pregnant females captured in Lachay lomas by month.The black portion of the bars show the number of pregnant females and the white portion shows the number of non pregnant females.

TABLE 1
Movement distances and length of residence of Phyllotis amicus and Phyllotis andiumDistancia de desplazamiento y tiempo de residencia de Phyllotis amicus y Phyllotis andium