Does habitat or depth influence catch rates of pelagic species?

Abstract

The efficiency of a pelagic longline fishing operation and the species composition of the resulting catch is influenced primarily by the relationship between the distribution of hooks and species vulnerability, with vulnerability described by either depth or some suite of environmental variables. We therefore fitted longline catch rate models to determine whether catch is estimated better by vertically distributing a species by depth or by environmental conditions (e.g., temperature, thermocline gradient, and oxygen concentration). Catch rates were estimated by two methods: (i) monitoring longlines where the vertical distribution of hooks and catch in relation to depth and environmental conditions is known, and (ii) applying a statistical habitat-based standardization (statHBS) model to fishery and environmental data to develop relative abundance indices for bigeye tuna (Thunnus obesus) and blue shark (Prionace glauca). Results indicated that an understanding of gear dynamics and environmental influences are important for analyzing catch-per-unit-effort (CPUE) data correctly. Analyses based on depth-specific catch rates can lead to serious misinterpretation of abundance trends, despite the use of sophisticated statistical techniques (e.g., generalized linear mixed models). This illustrates that inappropriate inclusion or exclusion of important covariates can bias estimates of relative abundance, which may be a common occurrence in CPUE analysis.