Pacific Salmon Environmental and Life History Models: Advancing Science for Sustainable Salmon in the Future
Modeling the Effect of Stream Network Characteristics and Juvenile Movement on Coho Salmon
Scott G. Liebowitz and Denis White
Abstract.—Simulation modeling can be a valuable tool for improving scientific understanding of the mechanisms that affect fish abundance and sustainability. Spatially explicit models, in particular, can be used to study interactions between fish biology and spatiotemporal habitat patterns and to study the significance of fish movement to population growth and abundance. In this chapter, we present a novel approach for modeling fish populations that generalizes across systems rather than being site-specific. We describe a spatially explicit coho (Oncorhynchus kisutch) simulation model that uses randomly generated stream networks. The use of random landscapes has been extremely valuable for the study of movement, metapopulation dynamics, and extinction risk in terrestrial populations, and should allow similar insights into long-term dynamics and sustainability of salmonids and other aquatic organisms. In particular, random stream networks can help test specific hypotheses by controlling for stream characteristics that could confound analytical results. In addition, random stream networks can allow for more robust inferences across stream networks by addressing network-level variability. We also include in our model a number of different mechanisms for representing fish movement. This allows us to examine how these different mechanisms affect model output, which will be useful for species where these mechanisms are poorly understood. We describe this model in detail, and include results from a number of simple analyses that are meant to examine the novel features of the model. These demonstrate how properties of the stream network can affect model output, and how different ways of representing fish movement can influence results.