Mortality sensitivity in life-stage simulation analysis: A case study of southern sea otters

Currently, there are no generally recognized approaches for linking detailed mortality and pathology data to population-level analyses of extinction risk. We used a combination of analytical and simulation-based analyses to examine 20 years of age- and sex-specific mortality data for southern sea otters (Enhydra lutris), and we applied results to project the efficacy of alternative conservation strategies. Population recovery of the southern sea otter has been slow (rate of population increase λ = 1.05) compared to other recovering populations (λ = 1.17-1.20), and the population declined (λ = 0.975) between 1995 and 1999. Age-based Leslie matrices were developed to explore explanations for the slow recovery and recent decline in the southern sea other population. An elasticity analysis was performed to predict effects of proportional changes in stage-specific reproductive or survival rates on the rate of population increase. A life-stage simulation analysis (LSA) was developed to evaluate the impact of changing age- and cause-specific mortality rates on λ. The information used to develop these models was derived from death assemblage, pathology, and live population census data to examine the sensitivity of sea otter population growth to different sources of mortality (e.g., disease and starvation, direct human take [fisheries, gun shot, boat strike, oil pollution], mating trauma and intraspecific aggression, shark bites, and unknown). We used resampling simulations to generate random combinations of vital rates for a large number of matrix replicates and drew on these to estimate potential effects of mortality sources on population growth (λ). Our analyses suggest management actions that are likely and unlikely to promote recovery of the southern sea otter and more broadly indicate a methodology to better utilize cause-of-death data in conservation decision-making.