Integrated multi-timescale modeling untangles anthropogenic, environmental, and biological effects on catchability

Catchability plays a central role in fisheries stock assessment. Since catchability often varies with time depending on population density, environmental factors, and anthropogenic effects, assuming constant catchability in population models can lead to biased abundance estimates. Here we present a novel way to simultaneously estimate time-varying catchability and abundance by integrating a short-term (month-based) removal method and a long-term (year-based) age-structured population dynamics model. We applied this approach to commercial fishery data for a Japanese pufferfish (Takifugu rubripes) population and found that the models with time-varying catchability greatly outperformed the models with constant catchability in terms of predictive ability and model consistency. The temporal variation in catchability was parsimoniously predicted by fishing effort and population size, indicating the existence of effort- and density-dependent catchability. Our approach, integrating population dynamics at different timescales, will help avoid inadvertent overexploitation and contribute to sustainable harvesting by enhancing the estimation accuracy of time-varying catchability and abundance.