Modelling the spatiotemporal distribution of fisheries discards: A case study on eastern Ionian Sea trawl fishery

The mitigation and incorporation of fisheries discards into the spatial management perspective call for a well-planned approach considering their high spatiotemporal variability. This study describes an approach to map the spatiotemporal distribution of bottom trawl discards taking into account on-board observer data, environmental information and high-resolution estimates of fishing effort derived from Vessel Monitoring System data. Two were the main objectives: i) mapping the Discards Per Unit Effort (DPUE: kg/h) and ii) mapping the total discards quantities (in kg) in a predefined spatial unit (grid cell of 5 × 5 km). For this purpose, discards data from commercial bottom trawlers (seasonal basis, period 2003-2006, 2008) operating on the continental shelf of the eastern Ionian Sea were analyzed. The DPUE for a) all species with Minimum Landing Size (MLS), b) all species without MLS and c) all fish species, were modeled by means of Generalized Additive Models (GAMs) using satellite environmental parameters, bathymetry and season as independent covariates. Mapping discards quantities in a predefined spatial unit was based on the spatial overlapping between GAM estimates for the year 2010 and fishing effort distribution for the same year. Spatial clusters of high discard quantities (hot spots) and their spatiotemporal persistence were also investigated. Results revealed how the environmental and temporal factors influence the DPUE distribution of each discard group and indicated increased discards quantities in areas of high fishing effort. The potential implications of the resulting DPUE maps for advancing discards spatial mitigation are further discussed. Mapping discards quantities using both the DPUE and the fishing effort distribution provide valuable insights that can be useful to investigate the broader ecosystem consequences of the EU landing obligation and other potential fishery management policies.