Best practices for defining spatial boundaries and spatial structure in stock assessment

Abstract

The ‘stock concept’ in fisheries science conforms to theoretical assumptions of stock assessment models, including negligible movement across stock boundaries, relatively homogeneous vital rates, and extensive mixing within stock areas. Best practices for representing population structure in stock assessment involve 1) interdisciplinary stock identification to delineate spatially discrete populations or more complex population structure; 2) stock boundaries that are aligned with the most plausible population structure; 3) spatially-explicit sampling, fleet structure or spatial structure in assessment models to account for heterogeneity, fishing patterns, and movement within stock areas; 4) routine stock composition sampling and analysis for spatially overlapping populations; and 5) simulation testing the performance of assessments with mis-specified or uncertain population structure. Practical assessment units that do not accurately represent population structure may not provide sufficient information to achieve fishery management objectives, so practical constraints should be addressed through iterative advances in routine stock identification, delineation of stocks to meet unit-stock assumptions, and stock assessment modeling.