Accounting for movement in spatial surplus production models and case studies of redfish (Sebastes spp., Sebastidae) and yellowtail flounder (Limanda ferruginea) on the Eastern Grand Banks of Newfoundland

This thesis explores the integration of spatial modelling and surplus production models (SPMs) for fisheries stock assessment. Typically, SPMs disregard the spatial dynamics of populations. To address this limitation, we propose a novel approach that utilizes the Gaussian random field to capture spatial heterogeneity. This method enhances spatial representation without explicitly parameterizing movement dynamics, offering a more robust framework. The methodology (i.e., the random field model) builds upon existing surplus production models by adapting a triangular grid and employing stochastic process errors to capture spatial variation. Simulations and case studies demonstrate the model’s effectiveness in estimating stock biomass dynamics, outperforming non-spatial and movement models. The random field model offers a simplified but robust alternative to the explicit spatial movement model. Applied to the 3LN Redfish stock, the random field model highlights significant spatial heterogeneity and a decline in biomass between 2012-2019. Furthermore, the approach was extended to Yellowtail Flounder in 3LNO Divisions, demonstrating stable biomass distributions with spatial preferences for shallower waters. The findings underscore the importance of spatially explicit models in fisheries stock assessment when sufficient spatial data are available. This study contributes to advancing fisheries stock assessment by providing a scalable and adaptable framework for spatial stock assessment.