Abstract
Eight spring declines in population size of M. townsendii were monitored by mark-recapture techniques near Vancouver, British Columbia Canada. In light-to-moderate spring declines most of the loss may be due to dispersal; in severe declines most of the loss may be due to death. In moderate spring declines, there is a positive correlation between rate of decline and proportion of heavy-weight voles; in severe declines, the correlation is negative. Males disappear more rapidly than females. Wounding rates are higher in moderate declines. In moderate spring declines, large voles of both sexes survive better than do small voles, but in severe declines, large and small females survive equally poorly. Smaller voles in the post-peak winter lost weight. Intrinsic and extrinsic factors contributed to the post-peak winter decline, but spacing behavior seems the most likely explanation for the rapid decrease in survival at the start of the decline breeding season.