95th ESA Annual Meeting (August 1 -- 6, 2010)

COS 119-4 - Utilizing hierarchical models of individual movement to construct maps of spatial predation risk

Friday, August 6, 2010: 9:00 AM
412, David L Lawrence Convention Center
Eric J. Ward, Northwest Fisheries Science Center, Seattle, WA, Phil Levin, NOAA-Fisheries, Seattle, WA and Alejandro Acevedo, Department of Biology, Western Washington University, Bellingham, WA
Background/Question/Methods

While there is strong evidence that some predators can play important roles in the dynamics of food webs, resource managers may be faced with complex tradeoffs regarding management actions. Since the 1970s, populations of harbor seals on the western coast of the U.S. have increased exponentially. In Washington state, populations appear to have leveled off only recently. Understanding how an increase in these consumers shapes the Puget Sound – Georgia Basin food web has implications for both biology and management.  While harbor seals are themselves no longer at risk, one of the most common prey species in the diet of harbor seals are at-risk fish species with high commercial value (salmonids).

Results/Conclusions  

In order to understand the spatial role of seal predation, we satellite-tagged harbor seals over several winters. These spatial positions were used to develop Bayesian state-space switching models of individual movement. These models have become widely used in ecology in recent years; here we extend previous models to a hierarchical framework, which allows us to explicitly quantify variation between individuals.  Two metrics of interest are the distribution of dispersal parameters across individuals, and the proportion of time spent foraging.  The posterior distributions on the individual state vectors suggests that these animals may travel much further to forage than previously thought. To answer our final question, we focus on the management implications of seal predation. We combine the posteriors from the movement model with observational data of seal densities at haul out sites to generate spatial distributions or risk-maps of predation pressure. With the addition of diet data, these maps can be constructed on a species-specific basis, for salmonids or other fish species at-risk.