Habitat suitability modeling based on a spatiotemporal model: an example for cusk in the Gulf of Maine

Habitat use and distribution is a critical aspect in the management and conservation of a species, particularly for those in decline. Habitat suitability indices (HSI) are a common method of habitat mapping dependent on empirical data that can easily lead to misunderstanding the spatiotemporal dynamics of marine species experiencing population decline and density-dependent catchability within surveys. This is especially true when only a single monitoring program with limited spatiotemporal coverage is used. A delta-generalized linear mixed model was used to combine trawl and longline surveys to predict density estimates for cusk (Brosme brosme) in unsampled locations for use in HSIs. Catchability was estimated for longline and trawl gear without having an estimate of area fished for the longline survey. HSIs performed better using model-based density estimates from multiple surveys compared with sample-based abundance indices from a single survey. The increased spatial resolution can better inform the HSIs by providing information where the survey programs did not sample. This study provides a novel approach for integrating data from different monitoring programs for habitat modeling.