During the austral winter
of 1987 (June-August) in Auco, an outbreak of small mammals was apparently
triggered by a single episode of unusually high rainfall. From October
1987 to November 1990, we monitored the outbreak on a monthly basis on
two equatorial- and two polar-facing slopes. Overall density on equatorial-facing
slopes was 239 individuals/ha in spring 1987, increasing to a peak of
404/ha by summer 1988, and then steadily declining to a crash of 20/ha
(5% of peak density) by spring 1990, with no signs of recovery.
On polar-facing slopes, mammalian
abundances were about half those on equatorial-facing slopes. There were
112 individuals/ha in spring 1987, increasing to a peak of 199/ha by summer
1988, and then steadily declining to a crash of 8/ha (4% of peak density)
by spring 1989. Since then, mammal populations on polar-facing slopes
have been slowly recovering, reaching 11% of their peak density by November
1990. Of the eight species monitored, only three irrupted: the granivorous
cricetid Phyllotis darwini, the omnivorous cricetid Akodon olivaceus,
and the insectivorous didelphid Thylamys elegans. These three irrupted
and declined in phase, simultaneously on the two opposite-facing slopes,
such that their relative frequencies did not shift markedly. Two of the
three folivores (Abrocoma bennetti, Octodon degus, but not Chinchilla
lanigera), one granivore (Oligoryzomys longicaudatus), and one insectivore
(Abrothrix longipilis) disappeared from the site, persisting longer on
equatorial-facing slopes.
We reported the extinction
and colonization rates of five sympatric small mammal species (Phyllotis
darwini, Akodon olivaceus, Thylamys elegans, Octodon degus, and Oligoryzomys
longicaudatus) in Auco. Based on six years of monitoring (1987-1992),
we provide information on how colonization and extinction rates change
according to landscape features (slope aspect) and on their relationship
to populations size, population variability, and body size. We found that:
(a) for all species in the assemblage, extinction rates of subpopulations
from equatorial-facing slopes were significantly lower than those in polar-facing
slopes, (b) population size was the most important factor determining
extinction rates, (c) colonization rates did not vary between slopes,
and were affected by population size only in equatorial-facing slopes,
and (d) most species had higher extinction than colonization rates. Persistence
of the metapopulation system for all five small mammal species appears
to be fueled by repeated colonization events
Four subpopulations
of Phyllotis darwini were monitored through monthly censuses conducted
1988-1991 on opposite slopes of two creeks separated by 2 km in Auco.
Nine response variables were measured to characterize the spatial and
temporal variability of the subpopulations: (a) number of individuals
per grid, (b) mean body mass, (c) coefficient of variation in body mass,
(d) coefficient of variation in body mass of males, (e) coefficient of
variation in body mass of females,(f) number of males, (g) number of females,
(h) number of juveniles, and (i) number of adults. These variables were
analyzed with regard to two main factors (creek and slope). The demographic
characteristics were more variable at a lower spatial scale (slope exposure)
than at a larger spatial scale (creek). In situ reproduction and recruitment
on each slope, as well as the low level of migration between slopes, indicate
that this system is not driven by source-sink dynamics. The spatial and
temporal dynamics of these subpopulations can be characterized as a partially
coupled metapopulation system whose component subpopulations are weakly
interconnected by dispersal between opposite slopes and virtually unconnected
between creeks.
It is widely known that some rodent populations
display dramatic density fluctuations in semiarid regions of western South
America after unusual rainfall levels associated with ENSO (El Niņo Southern
Oscillation) disturbances. These correlated phenomena have led some ecologists
to believe that rodent outbreaks are determined solely by density-independent
factors (e.g. rainfall regime). However, demographic studies have detected
strong delayed density-dependence effects in one of the most irruptive rodent
species, the leaf-eared mouse Phyllotis darwini. We tested the effects of
rainfall and delayed density-dependent factors by constructing a structured
model based on demographic data estimated from a capture-mark-recapture (CMR)
study of this species in Chile. A model including both rainfall and delayed
density-dependence effects predicts the observed population dynamics rather
accurately over a 10 year period. Interestingly, small changes in model parameters
result in large changes in model dynamics, which strongly suggest the importance
of local variations in demographic features for explaining the asynchronous
pattern in outbreak occurrences. These findings suggest that inextricably
intertwined endogenous and exogenous forces cause rodent outbreaks in western
South America. The former are characterized by delayed non-linear feedbacks,
whereas the latter by the positive effects of "El Nino" and the
negative effects of "La Nina" phases of the ENSO disturbance.
In the last two decades,
several researchers have noted rodent population outbreaks in semiarid
South America, in association with unusually high precipitation that seemingly
concurs with ENSO events (El Niņo Southern Oscillation). To date, no studies
have been conducted to determine the statistical relationships between
ENSO events, increased precipitation, and rodent irruptions.
Here we show that: (a) there
is a close association between ENSO events and increased precipitation
in the semiarid region of northern Chile; (b) the occurrence of rodent outbreaks in that region
is statistically related with the precipitation levels of the same year; (c) the multi-annual
patterns of the total annual precipitation levels and population abundance
of those rodents during the summer are positively associated. The putative
chain of effects seems to start with unusually high rainfall brought by
ENSO to semiarid environments, which thus respond with increased primary
productivity (herbage and seeds), which then fuels the rodent outbreaks.
We reported the extinction
and colonization rates of five sympatric small mammal species (Phyllotis
darwini, Akodon olivaceus, Thylamys elegans, Octodon degus, and Oligoryzomys
longicaudatus) in Auco. Based on six years of monitoring (1987-1992),
we provide information on how colonization and extinction rates change
according to landscape features (slope aspect) and on their relationship
to populations size, population variability, and body size. We found that:
(a) for all species in the assemblage, extinction rates of subpopulations
from equatorial-facing slopes were significantly lower than those in polar-facing
slopes, (b) population size was the most important factor determining
extinction rates, (c) colonization rates did not vary between slopes,
and were affected by population size only in equatorial-facing slopes,
and (d) most species had higher extinction than colonization rates. Persistence
of the metapopulation system for all five small mammal species appears
to be fueled by repeated colonization events
Four subpopulations
of Phyllotis darwini were monitored through monthly censuses conducted
1988-1991 on opposite slopes of two creeks separated by 2 km in Auco.
Nine response variables were measured to characterize the spatial and
temporal variability of the subpopulations: (a) number of individuals
per grid, (b) mean body mass, (c) coefficient of variation in body mass,
(d) coefficient of variation in body mass of males, (e) coefficient of
variation in body mass of females,(f) number of males, (g) number of females,
(h) number of juveniles, and (i) number of adults. These variables were
analyzed with regard to two main factors (creek and slope). The demographic
characteristics were more variable at a lower spatial scale (slope exposure)
than at a larger spatial scale (creek). In situ reproduction and recruitment
on each slope, as well as the low level of migration between slopes, indicate
that this system is not driven by source-sink dynamics. The spatial and
temporal dynamics of these subpopulations can be characterized as a partially
coupled metapopulation system whose component subpopulations are weakly
interconnected by dispersal between opposite slopes and virtually unconnected
between creeks. 
It is widely known that some rodent populations
display dramatic density fluctuations in semiarid regions of western South
America after unusual rainfall levels associated with ENSO (El Niņo Southern
Oscillation) disturbances. These correlated phenomena have led some ecologists
to believe that rodent outbreaks are determined solely by density-independent
factors (e.g. rainfall regime). However, demographic studies have detected
strong delayed density-dependence effects in one of the most irruptive rodent
species, the leaf-eared mouse Phyllotis darwini. We tested the effects of
rainfall and delayed density-dependent factors by constructing a structured
model based on demographic data estimated from a capture-mark-recapture (CMR)
study of this species in Chile. A model including both rainfall and delayed
density-dependence effects predicts the observed population dynamics rather
accurately over a 10 year period. Interestingly, small changes in model parameters
result in large changes in model dynamics, which strongly suggest the importance
of local variations in demographic features for explaining the asynchronous
pattern in outbreak occurrences. These findings suggest that inextricably
intertwined endogenous and exogenous forces cause rodent outbreaks in western
South America. The former are characterized by delayed non-linear feedbacks,
whereas the latter by the positive effects of "El Nino" and the
negative effects of "La Nina" phases of the ENSO disturbance.
