Winter weather versus group thermoregulation: what determines survival in hibernating mammals?

For socially hibernating mammals, the effectiveness of huddling as a means of energy conservation should increase with group size. However, group size has only been linked to increased survival in a few hibernating species, and the relative importance of social structure versus winter conditions during hibernation remains uncertain. We studied the influence of winter weather conditions, social group composition, age-structure, and other environmental factors and individual attributes on the overwinter survival of hoary marmots (Marmota caligata) in the Yukon Territory, Canada. Juvenile hoary marmot survival was negatively correlated with the mean winter (November to May) Pacific Decadal Oscillation (PDO) index. Survival in older age-classes was negatively correlated with PDO lagged by 1 year. Social group size and structure were weakly correlated with survival in comparison to PDO. The relationship between winter PDO and survival was most likely due to the importance of snowpack as insulation during hibernation. The apparent response of hoary marmots to changing winter conditions contrasted sharply with those of other marmot species and other mammalian alpine herbivores. In conclusion, the severity of winter weather may constrain the effectiveness of group thermoregulation in socially hibernating mammals.

The impact of predation on burrow use by arctic ground squirrels in the boreal forest.

In sedentary animals, the choice of a suitable home site is critical to survival and reproductive fitness. However, habitat suitability may vary with predation risk. We compared habitat use of Arctic ground squirrels (Spermophilus parryii plesius) living in the boreal forest under conditions of fluctuating predation pressure. In our study area, predators show ten-year cycles in numbers that track that of their primary prey, the snowshoe hare (Lepus americanus). In 1993, we compared burrows that continued to be occupied following the period of intense predation during the hare decline of 1990-1992 with those that became vacant, and with random locations. We contrasted these sites to those in a predator exclosure where predation pressure was minimized. Burrows on control sites were located on sloped sites with high visibility. Burrows that remained occupied during the period of intense predation were more likely to be in open areas with fewer fallen trees than burrows that became vacant. We used discriminant functions derived from the control sites and found that 89% of the burrows on the predator exclosure were classified as being similar to the random locations on control sites. We conclude that the distribution of Arctic ground squirrels in the boreal forest is a direct function of predator presence.

Social structure and facultative mating systems of hoary marmots (Marmota caligata).

Mate-choice theory predicts different optimal mating systems depending on resource availability and habitat stability. Regions with limited resources are thought to promote monogamy. We tested predictions of monogamy in a social rodent, the hoary marmot (Marmota caligata), at the northern climatic extreme of its distribution. Mating systems, social structure and genetic relationships were investigated within and among neighbouring colonies of marmots within a 4 km(2) valley near Kluane National Park, Yukon, Canada, using 21 microsatellite loci. While both monogamous and polygynous populations of hoary marmots have been observed in the southern reaches of this species' range; northern populations of this species are thought to be predominantly monogamous. Contrary to previous studies, we did not find northern hoary marmot social groups to be predominantly monogamous; rather, the mating system seemed to be facultative, varying between monogamy and polygyny within, as well as among, social groups. These findings reveal that the mating systems within colonies of this species are more flexible than previously thought, potentially reflecting local variation in resource availability.

Concurrent density dependence and independence in populations of arctic ground squirrels.

No population increases without limit. The processes that prevent this can operate in either a density-dependent way (acting with increasing severity to increase mortality rates or decrease reproductive rates as density increases), a density-independent way, or in both ways simultaneously. However, ecologists disagree for two main reasons about the relative roles and influences that density-dependent and density-independent processes have in determining population size. First, empirical studies showing both processes operating simultaneously are rare. Second, time-series analyses of long-term census data sometimes overestimate dependence. By using a density-perturbation experiment on arctic ground squirrels, we show concurrent density-dependent and density-independent declines in weaning rates, followed by density-dependent declines in overwinter survival during hibernation. These two processes result in strong, density-dependent convergence of experimentally increased populations to those of control populations that had been at low, stable levels.