Predictive substratum modelling for juvenile gadoid distribution and abundance

No abstracts are to be cited without prior reference to the author.

Successful spatial management of species and their habitats requires a good understanding of their distribution. Coastal areas are proposed to be of particular importance for juvenile demersal fish recruitment. However, in many cases the distribution of seabed types and their role for species are poorly understood. Data on fish relative abundance, size and substratum type were collected using Stereo Baited Remote Underwater Video cameras (SBRUV). A multinomial model was used to predict and map substrata around a recently designated Marine Protected Area (MPA) in the Firth of Clyde, west of Scotland. The predicted substratum map was used to understand the relationship between seabed type and extent and gadoid distribution. The multinomial model performed well with an area under the curve (AUC) score of 0.87. Substratum type and heterogeneity had varying effects on gadoid relative abundance. Atlantic cod (Gadus morhua) was associated with relatively more rugose and heterogeneous substrata than haddock (Melanogrammus aeglefinus) or whiting (Merlangius merlangus).