Demography of the Pacific walrus (Odobenus rosmarus divergens): 1974–2006

Abstract

Global climate change may fundamentally alter population dynamics of many species for which baseline population parameter estimates are imprecise or lacking. Historically, the Pacific walrus is thought to have been limited by harvest, but it may become limited by global warming-induced reductions in sea ice. Loss of sea ice, on which walruses rest between foraging bouts, may reduce access to food, thus lowering vital rates. Rigorous walrus survival rate estimates do not exist, and other population parameter estimates are out of date or have well-documented bias and imprecision. To provide useful population parameter estimates we developed a Bayesian, hidden process demographic model of walrus population dynamics from 1974 through 2006 that combined annual age-specific harvest estimates with five population size estimates, six standing age structure estimates, and two reproductive rate estimates. Median density independent natural survival was high for juveniles (0.97) and adults (0.99), and annual density dependent vital rates rose from 0.06 to 0.11 for reproduction, 0.31 to 0.59 for survival of neonatal calves, and 0.39 to 0.85 for survival of older calves, concomitant with a population decline. This integrated population model provides a baseline for estimating changing population dynamics resulting from changing harvests or sea ice.

Trend in Pacific walrus (Odobenus rosmarus divergens)tusk asymmetry, 1990–2014

“…This population trend, as well as, the age composition of females, and model derived reproductive rates and calf survival rates are also consistent with expectations based on the Pacific walrus population reaching or exceeding K n early in the time series. The negative trend in FA1 for males and females combined and males alone from 1990 to 2014 is consistent with the modeling results of Taylor and Udevitz () and other evidence (Fay et al . , MacCracken et al .…”

“…The magnitude and rate of population change cannot be reliably determined from these surveys due to the imprecision of the estimates and sampling issues (Hills and Gilbert , Udevitz et al . , Taylor and Udevitz ). However, Taylor and Udevitz () integrated the aerial survey population estimates (recalculated based on a log‐normal distribution) and ship‐based age and sex composition counts that were conducted annually in 1981–1984, 1998, and 1999 (Citta et al .…”

“…, Taylor and Udevitz ). However, Taylor and Udevitz () integrated the aerial survey population estimates (recalculated based on a log‐normal distribution) and ship‐based age and sex composition counts that were conducted annually in 1981–1984, 1998, and 1999 (Citta et al . ) with Bayesian methods.…”

“…, MacCracken et al . , Taylor and Udevitz ). Density dependent competition for food and space at that time likely stressed individuals and resulted in population level effects (Fay et al .…”

“…Pacific walruses are long‐lived (Fay ) and their primary prey (mollusks and worms) would also take several years to recover once the walrus population declined and predation rates on those food items also declined. Taylor and Udevitz () estimated that survival and reproductive rates in Pacific walruses did not begin to increase from lows associated with high population densities until a decade later.…”