Fishers' knowledge improves the accuracy of food web model predictions

Ecosystem-based management advice requires the development of food web models and Ecopath with Ecosim (EwE) has become one of the most widely used approaches. However, models such as EwE have high data requirements that usually results in important uncertainties in data input. In these cases, fishers’ knowledge can be a valuable but under-utilised source of information. To investigate the drivers behind ecosystem changes in the Irish Sea, an EwE model was constructed. Due to a significant gap in scientific knowledge on fishing effort prior to 2003, we demonstrate a new approach to incorporating fishers’ knowledge into EwE models. A new Bayesian approach was developed to optimise the magnitude of historical effort trends provided by fishers within ± 99% uncertainty ranges based on the largest deviations between scientific and fishers’ effort estimates. The Irish Sea model was then fitted and results compared using fishing effort time-series based on: 1) scientific knowledge, 2) fishers’ knowledge, 3) adjusted fishers’ knowledge, and 4) a combination of (1) and (3), termed ‘hybrid knowledge’. The hybrid model produced the best overall statistical fit, capturing the biomass trends of commercially important stocks. Importantly, the hybrid model also replicated the increase in landings of groups such as ‘crabs & lobsters’ and ‘epifauna’ (e.g. common whelk, sea urchin, common mussel) which, using only scientific knowledge, had no drivers for fishing effort. Incorporating environmental drivers (primary production) and adjusting vulnerabilities in the foraging arena further improved model fit. The final model suggests that both fishing and environmental drivers have historically influenced trends in finfish and shellfish stocks in the Irish Sea. Such co-production of knowledge not only improved the overall food web model but the process strengthened the trust and relationships between the fishers and scientists. This approach may prove to be fundamental for developing more robust and credible ecosystem-based management advice in a global context.