Document: BRE-3-26-7

Emergence of complex, realistic habitat selection patterns from fitness-based rules in a stream trout model.

HARVEY, B.C.* ¹ and S.F.RAILSBACK ²

U.S.D.A. Forest Service, Redwood Sciences Lab, Arcata, CA, USA ¹
Lang, Railsback and Associates, Arcata, CA, USA ²

Abstract:
For individual-based models (IBMs) of mobile organisms, effective rules governing movement are essential for realistic simulation of habitat selection, survival, and growth. Evidence of realistic habitat selection by animals in IBMs should provide greater confidence in the ability of these models to predict population-level parameters.Using an IBM for stream fish, we tested the ability of three different movement rules to reproduce six patterns of habitat selection which have been observed in stream-dwelling trout. The IBM simulates habitat selection in an environment with spatio-temporal variation in food availability and mortality risk. Model fish move in each daily time step to maximize their movement objectives, with no other restrictions imposed on habitat selection. We compared patterns of habitat selection when model fish moved to maximize: 1) instantaneous growth; 2) instantaneous probability of survival; or 3) "Expected Maturity" (EM). EM is the product of: 1) predicted survival over a set time period (incorporating mortality risks from starvation, predation, and extreme physical conditions) and 2) the fraction of reproductive size attained over the time period. Simulations with movement to maximize growth reproduced three of the six patterns of habitat selection, maximizing survival reproduced two patterns, and maximizing EM using a 90-d time horizon reproduced all six patterns. Only the EM movement objective reproduced two particular patterns: 1) a shift with increasing temperature to greater use of habitats with higher water velocities, and 2) shifts to habitats with relatively low probabilities of survival in response to lower overall food availability. "Maximize EM" is a simple movement rule for IBMs which leads to realistic patterns of behavior, in part by successfully reflecting trade-offs between mortality risk and growth which affect habitat selection.

Keywords: Individual-based modeling, movement, habitat selection, trout, emergence

This abstract is being presented at: 9:40 AM in session:
Symposium # 27: Advancing the Individual-Based Modeling Approach: New Tools and Concepts.