Manuscript Report 176

175. Boulanger J. and A. Gunn. 2007. Exploring Possible Mechanisms for the Decline of the Bathurst Herd of Barren-Ground Caribou Using Demographic Modeling. 56 pp. In Press. (PDF = 683 KB)

ABSTRACT

The Bathurst herd of barren-ground caribou (Rangifer tarandus groenlandicus) declined between 1986 and 2006 at a mean annual rate of about 5% (Gunn at al. 2005a). We use a demographic model to explore possible mechanisms that were responsible for this decline. The modeling exercise had three components.  First, we compiled demographic data from previous research on the Bathurst herd and other migratory tundra caribou herds.  Second, we developed a population model that could be objectively fitted to the field data. Third, we undertook further demographic analysis using results of the initial population model to investigate effects of harvesting and potential scenarios for herd recovery. The field data most supported a model with calf survival and fecundity declining while adult male and female survival was constant.  However, low sample sizes of collared caribou reduced survival estimate precision therefore reducing the power to detect trends in survival.   Elasticity analysis suggested that the population could tolerate a larger degree of variation in productivity compared to adult survival.  The model with declining calf survival and fecundity was then used to explore potential effects of harvest and herd recovery. We simulated potential increases in adult survival rates due to reduced hunting (with productivity constant).   The model predicted that the population would still decline unless the adult survival rate was increased by more than 7.5%, suggesting that herd productivity (calf survival and fecundity) must increase for population recovery to occur.  We ran simulations using 2007 parameter estimates of fecundity or calf survival and found that the herd would still decline unless calf survival increased to levels (i.e. 0.5 to 0.6).  The principal challenge for our modeling was objectively determining plausible demographic model parameters and model formulations given the relatively few field measurements of demographic rates. Our modeling indicated that the Bathurst herd declined because of a trend toward reduced calf survival (and/or fecundity), which was likely exacerbated by reduced survival of adult females. Although a reduction in hunting would improve adult survival, our modeling suggested the herd will not be able to recover until calf survival improves.  In addition, our results highlight the need for continuous monitoring of population parameters such as adult survival and productivity.