Multiple hypothesis testing and the declining-population paradigm in Steller sea lions

Area of Science:

• Ecology
• Population Dynamics
• Conservation Biology

Background:

• Traditional regression methods struggle with sparse, noisy data and multiple hypotheses for population declines.
• Understanding drivers of population decline is crucial for effective conservation and management.
• The western Alaskan population of Steller sea lions (Eumetopias jubatus) faces significant decline.

Purpose of the Study:

• To introduce and validate a novel, spatially and temporally detailed, likelihood-based approach for identifying causes of population declines.
• To apply this method to the Steller sea lion population in western Alaska.
• To simultaneously consider multiple hypotheses regarding factors affecting vital rates.

Main Methods:

• Developed a stochastic population model incorporating vital rates (fecundity, pup survival, non-pup survival) as functions of local conditions.
• Utilized a likelihood-based framework, unbiased by sample size, to analyze rookery- and year-specific census data.
• Employed backward iteration and a beta-binomial structure to calculate likelihoods and estimate parameters, constructing marginal likelihood curves.

Main Results:

• Found strong support for a positive effect of total prey availability on pup recruitment.
• Identified negative effects of prey species composition (pollock fraction) on fecundity and pup survival.
• Detected a positive effect of harbor seal density (predation risk proxy) on non-pup survival.

Conclusions:

• The likelihood-based approach effectively disentangles multiple factors influencing population dynamics, outperforming traditional regression.
• Results highlight key environmental and ecological drivers impacting Steller sea lion vital rates.
• The framework supports adaptive management, such as making fishery quotas contingent on prey surveys.