Metapopulation ecology of Vancouver Island marmots (Marmota vancouverensis)




Bryant, Andrew Albert

Journal Title

Journal ISSN

Volume Title



Vancouver Island marmots (M. vancouverensis) rank among the world's most critically endangered mammals. There were probably fewer than 100 marmots in 1998, with 90% distributed south of Alberni Inlet, and the remainder on or near Mount Washington. This represents a 60–70% decline in numbers during the past 10 years, and a considerably reduced geographic range during the past several decades. I used data from marked animals, radio-telemetry and population counts to test whether population dynamics were consistent with predictions made under five hypotheses: habitat tracking, sink-connectivity, weather, predators and disease. Estimates of demographic rates from intensive mark-recapture work and population counts were generally consistent, although estimation of adult survival from counts was problematic because of the difficulty of distinguishing surviving marmots from immigrants. There was no apparent influence of mark-recapture on survival or reproduction, and intensively studied colonies showed similar dynamics to colonies that were visited infrequently. There was little evidence for habitat tracking in natural habitats. Few colonies showed chronically low reproduction or survival, which would be the predicted result of a gradually deteriorating environment. Declines were more often abrupt and catastrophic. Marmots did not colonize clearcuts in proportion to their temporal or spatial availability, and ultimately colonized only a minuscule fraction of the potential habitat. However, marmots already inhabiting clearcuts represent a special case of habitat tracking; survival rates were significantly lower at clearcuts of more advanced seral age (i.e., >11 years after harvest). Evidence for source-sink and landscape connectivity processes was relatively strong. Marmots inhabiting clearcuts had chronically lower survival rates (by 5–10%). Per female reproductive contribution in clearcuts was half that of females inhabiting natural environments. However not all clearcuts acted as sinks, or acted as sinks in all years. Colonizations of clearcuts were spatially concentrated and none occurred at distances greater that 5 km from an existing natural colony. Apparent adult survival was significantly associated with isolation but juvenile survival was not, which is consistent with the prediction that isolated colonies should receive fewer immigrants. However the spatial pattern of extinctions was unexpected. Isolated and closely-clustered colonies had similar probabilities of extinction. Weather significantly influenced marmot survival and reproduction but explained only small amounts of variation. Survival was significantly associated with rainfall, temperature and snowpack depth. Reproduction was negatively associated with snowpack and temperature. Slope aspect was significantly associated with survival, perhaps suggesting the importance of snowmelt patterns. Natural and clearcut colonies responded differently to weather. Indices of wolf and cougar abundance were inconsistent and probably do not reflect true population sizes. Deer abundance was weakly associated with marmot survival in natural habitats, which could suggest switching of predator hunting effort. Marmot survival was spatially correlated, which is consistent with the idea that a few individual predators may focus hunting efforts at adjacent colonies. Field observations and radio-telemetry corroborated the importance of predators. In natural habitats, disappearances were uniformly distributed throughout summer, as predicted. In clearcuts, disappearances, were more heavily skewed towards late summer, suggesting that winter mortality was more important. Spatial correlation of survival is also consistent with the disease hypothesis. Survival was lower in colonies with high relative density of adults, which is a predicted result given the prediction of increased risk of disease transmission. The incidence of high mortality events increased during the 1990s, and the degree of spatial correlation also increased despite a more fragmented population structure. These trends are consistent with a hypothesis of a new disease organism or increased risk of infection. Forestry appears to be the primary cause of recent population dynamics in the Nanaimo Lakes region. Logging reduced overall marmot survival, inhibited their ability to re-colonize sites, and concentrated the population, making colonies more susceptible to predators and disease. The prognosis for continued survival remains hopeful provided that current plans for captive-breeding and captive-breeding and reintroduction are pursued aggressively.



Marmots, Habitat conservation