Advanced Spatial Modeling to Inform Management of Data-Poor Juvenile and Adult Female Rays

Abstract

Chronic overfishing has depleted numerous elasmobranch stocks in the North East Atlantic, but addressing this issue has been hampered by management complications and lacking data. Spatial management approaches have thus been advocated. This work presents a novel application and further development of an advanced spatial modeling technique to identify candidate nursery grounds and spawning areas for conservation, by subsetting already limited data. Boosted Regression Tree models are used to predict abundance of juvenile and mature female cuckoo (Leucoraja naevus), thornback (Raja clavata), blonde (Raja brachyura), and spotted (Raja montagui) rays in the Irish Sea using fish survey data and data describing fishing pressure, predation and environmental variables. Model-predicted spatial abundance maps of these subsets reveal distinct nuances in species distributions with greater predictive power than maps of the whole stock. These resulting maps are then integrated into a single easily understood map using a novel approach, standardizing and facilitating the spatial management of data-limited fish stocks.