Fishing suitability maps: helping fishermen reduce discards

Discards poses a serious problem for the productivity and sustainability of European fisheries and thus is an important fisheries and ecological management issue to be solved. In this paper we present a statistical tool, based on the random forest technique, that aims to reduce the negative ecological impact of fishing by providing fishermen with near-real-time maps of a fishing suitability index based on haul-by-haul catch and discard rates, indicating the most suitable areas for fishing. These easy-to-interpret maps are to be accessible to users via an online geoportal. Observer data from the Spanish discard sampling program from 2004 to 2008 for several species caught in the Cantabrian Sea (ICES area VIIIc) were used to illustrate the random forest approach. Results in the case study varied among species and seasons, with better results achieved for balanced datasets, such as those for economically valuable target species with segregated life stages. We discuss how this online tool could be useful for fisheries management, particularly in the context of the European Common Fisheries Policy reform and the discard ban on commercial species.