Modeling the interaction between salmon management and consumption by coastal brown bears

Abstract

Abstract Harvest management policy for species with strong trophic connections can reverberate through food webs and cause unintended consequences, such as altering the abundance of a harvested species' predators or prey. Pacific salmon (Oncorhynchus spp.), a key food for many predators and an economically valuable harvested species, is generally managed for maximum sustained harvests without explicit consideration for the freshwater and terrestrial food webs that they support. The density of brown bear (Ursus arctos) populations in Alaska, USA, is correlated with the amount of salmon they can access and consume, so it seems likely their populations are inadvertently affected by salmon management. We simulated the effect of salmon management policy on brown bears by customizing a general bear–salmon model using empirical data from three watersheds in southwest Kodiak, Alaska. Our goal was to quantify the effect of current salmon management policy (i.e., escapement goals and early/late run allocations) on salmon consumption by brown bears. Bears in the individually‐based model evaluated the value of each foraging site based on salmon abundance, salmon vulnerability, and competition with other bears and made movement decisions (among salmon spawning sites) accordingly. A validation of the model based on empirical brown bear foraging data revealed that simulated bears selected the same salmon spawning locations but visited more sites and fished for more days compared to real bears. In simulations across variables (salmon abundance, phenological variation, bear competition, and bear density), consumption of salmon by bears was remarkably resilient to changes in salmon abundance within the range of current high and low escapement goals, as long as all run‐timing variation and salmon sub‐populations were preserved. Mean salmon consumption increased by ~20% as escapement approximately doubled from the regional lower escapement goal of 625,000 to the upper goal of 1,270,000, but rapidly declined if salmon abundance decreased below the minimum escapement goal. In this system, salmon management policy, designed to achieve maximum sustained yield, also provides adequate salmon for the current bear population. This model creates a potentially useful tool for assessing the knock‐on effects of future changes to salmon management policy.

Keywords

• EBFM
• ecosystem‐based management
• fisheries
• phenology
• predation
• salmon management