Estimation of Juvenile Coho Salmon Occurrence, Abundance, and Detection Using N-Mixture Models Across Multiple Spatial Scales to Define Population Spatial Structure

The data necessary for resource assessment and species targeted restoration should be synonymous.  An understanding of habitat patch size, utilization, connectivity, and colonization and extinction dynamics, must be considered in the context of viability assessment. However, from a restoration and conservation perspective, occupied habitats should be targets for protection as potential sources of recolonization while currently unsuitable habitat is restored.  We applied the hierarchical n-mixture abundance model of Royle (2004) to spatially replicated stream pool dive counts of summer rearing juvenile coho salmon to estimate the probability of occupancy and local abundance simultaneously at two spatial scales over four populations. Within each population, the larger scale corresponds to the average abundance of animals within the sampled pools of the reach (λ) and probability of occupancy (ψ) at the sample reach (ψ=1-exp(-λ)) , whereas the smaller scale corresponds to the average abundance of animals per pool (λ_p), and probability of occupancy at the sample pool (θ), given the species occurred in the sample reach.  Detection probability (p) was modeled at both the reach and pool scale based on two independent and replicated snorkel counts. Our results suggest this study design is effective at defining population spatial structure of coho salmon.