Search

A major question in ecology is how age-specific variation in demographic parameters influences population dynamics. Based on long-term studies of growing populations of birds and mammals, we analyze population dynamics by using fluctuations in the total reproductive value of the population. This enables us to account for random fluctuations in age distribution. The influence of demographic and environmental stochasticity on the population dynamics of a species decreased with generation time. Variation in age-specific contributions to total reproductive value and to stochastic components of population dynamics was correlated with the position of the species along the slow-fast continuum of life-history variation. Younger age classes relative to the generation time accounted for larger contributions to the total reproductive value and to demographic stochasticity in "slow" than in "fast" species, in which many age classes contributed more equally. In contrast, fluctuations in population growth rate attributable to stochastic environmental variation involved a larger proportion of all age classes independent of life history. Thus, changes in population growth rates can be surprisingly well explained by basic species-specific life-history characteristics.

Despite recent advances in biodemography and metapopulation ecology, we still have limited understanding of how local demographic parameters influence short- and long-term metapopulation dynamics. We used long-term data from 17 local populations, along with the recently developed methods of matrix metapopulation modeling and transient sensitivity analysis, to investigate the influence of local demography on long-term (asymptotic) versus short-term (transient) dynamics of a yellow-bellied marmot metapopulation in Colorado. Both long- and short-term dynamics depended primarily on a few colony sites and were highly sensitive to changes in demography at these sites, particularly in survival of reproductive adult females. Interestingly, the relative importance of sites differed between long- and short-term dynamics; the spatial structure and local population sizes, while insignificant for asymptotic dynamics, were influential on transient dynamics. However, considering the spatial structure was uninformative about the relative influence of local demography on metapopulation dynamics. The vital rates that were the most influential on local dynamics were also the most influential on both long- and short-term metapopulation dynamics. Our results show that an explicit consideration of local demography is essential for a complete understanding of the dynamics and persistence of spatially structured populations.

We investigated mechanisms of energy conservation during hibernation. The amount of time torpid was significantly less for groups of three young marmots than for marmots hibernating singly. Mean daily mass loss (DML; as mg d(-1) g(-1) immergence mass) averaged 1.33 for single marmots and 1.46 for grouped young. Animals were active 17.3% of the time, which used 82.4% of the energy, and were torpid 82.7% of the time, which used 17.6% of the energy expenditure. During longer torpor bouts, more time was spent in deep torpor, which decreased the hourly cost of a complete bout. Bout oxygen consumption V dot o2, percent time in deep torpor, and body temperature (T(B)) during deep torpor changed seasonally and were curvilinearly related to when in the hibernation period the measurements were made and probably represent a stage in the circannual metabolic cycle. The decrease of environmental temperature (T(E)) to 2 degrees C significantly increased metabolism. Potential costs of low T(E) were reduced by allowing T(B) to decrease, thereby reducing the T(B) to T(E) gradient. Average monthly metabolic rate was high early and late in the hibernation period when time spent euthermic was greater and when VO2 was higher. Over the hibernation period, energy saved averaged 77.1% and 88.0% of the costs for winter and summer euthermic metabolism, respectively. Hibernation costs were reduced by the seasonal changes, the high percentage of time in torpor, the rapid decline in V dot o2 following arousal, and allowing T(B) to decline at lower T(E). Asynchrony in the torpor cycles increased energy expenditures in group hibernators, which negated possible beneficial effects of group hibernation.

While sociality has been hypothesized to drive the evolution of communicative complexity, the relationship remains to be formally tested. We derive a continuous measure of social complexity from demographic data and use this variable to explain variation in alarm repertoire size in ground-dwelling sciurid rodents (marmots, Marmota spp.; prairie dogs, Cynomys spp.; and ground squirrels, Spermophilus spp.). About 40% of the variation in alarm call repertoire size was explained by social complexity in the raw data set. To determine the degree to which this relationship may have been influenced by historical relationships between species, we used five different phylogenetic hypotheses to calculate phylogenetically independent contrasts. Less variation was significantly explained in contrast-based analyses, but a general positive relationship remained. Social complexity explained more variation in alarm call repertoire size in marmots, while sociality explained no variation in repertoire size in prairie dogs and no variation in phylogenetically based analyses of squirrels. In most cases, substantial variation remained unexplained by social complexity. We acknowledge that factors other than social complexity, per se, may contribute to the evolution of alarm call repertoire size in sciurid rodents, and we discuss alternative hypotheses. Our measure of social complexity could be used by other researchers to test explicit evolutionary hypotheses that involve social complexity.