Northern Prairie Wildlife Research Center
The only weather data that appeared extreme were those indicating that winter began considerably earlier than usual. Temperatures, precipitation, and snow cover in late August and September brought winter conditions five to six weeks earlier than the norm. Because daily minimum temperatures do not usually rise above freezing until mid-June (Environment Canada 1998), most vegetation does not begin growing until late June. Thus, the usual summer replenishment period for herbivores is approximately from 1 July to about 1 October, or three months.
However, the early onset of winter conditions in 1997 would have ended the replenishment period in mid-August, reducing it by about 50%. This shortened feeding period appears to be the best explanation for the demographic problems observed in muskoxen and arctic hares during summer 1998.
Conceivably the relative dearth of these animals in 1998 might be attributable to some unusual distribution pattern. However, there is little reason to believe that both species would show such a pattern, for there is no evidence that the two are regularly associated (Schaefer et al., 1996). Furthermore, the lack of young of both species, when at least some adults were seen, evinces a population phenomenon that is real and bolsters the conclusion that the low numbers of adults seen reflected a reduction in population.
Because there was no evidence of either ice crusts or abnormally deep snow in spring, the population phenomenon I observed fit none of the usual hypotheses about muskox population declines (Thomas et al., 1981; Gunn et al., 1989, Forchhammer and Boertmann, 1993). Therefore I present a new hypothesis: that early onset of winter that might cause adverse demographic effects in arctic herbivores by shortening their summer replenishment period.