Evaluating the use of acoustic bottom typing to inform models of bottom trawl sampling efficiency

Abstract

The sampling efficiency of survey bottom trawls can vary in response to a variety of biotic and abiotic factors. In a previously published model of the sampling efficiency of snow crab (Chionoecetes opilo), based on a trawl comparison experiment, the efficiency of the bottom trawl used in the eastern Bering Sea varied with water depth and sediment size. The sediment size used in the model, however, was not directly sampled at the trawl locations but instead was interpolated from an existing sediment data base. We examine whether bottom type attributes estimated with commercially available software applied to acoustic data collected by the vessels during the experiment was more informative to the model than the interpolated estimates of sediment size. Based on increases in explained deviance, the model fits for males (6% higher) and females (35% higher) were both improved using the acoustically derived estimates of bottom type.