Improving the Design of Acoustic and Midwater Trawl Surveys through Stratification, with an Application to Lake Michigan Prey Fishes

Abstract

Reliable estimates of fish biomass are vital to the management of aquatic ecosystems and their associated fisheries. Acoustic and midwater trawl surveys are an efficient sampling method for estimating fish biomass in large bodies of water. To improve the precision of biomass estimates from combined acoustic and midwater trawl surveys, sampling effort should be optimally allocated within each stage of the survey design. Based on information collected during fish surveys, we developed an approach to improve the design of combined acoustic and midwater trawl surveys through stratification. Geographic strata for acoustic surveying and depth strata for midwater trawling were defined using neighbor-restricted cluster analysis, and the optimal allocation of sampling effort for each was then determined. As an example, we applied this survey stratification approach to data from lakewide acoustic and midwater trawl surveys of Lake Michigan prey fishes. Precision of biomass estimates from surveys with and without geographic stratification was compared through resampling. Use of geographic stratification with optimal sampling allocation reduced the variance of Lake Michigan acoustic biomass estimates by 77%. Stratification and optimal allocation at each stage of an acoustic and midwater trawl survey should serve to reduce the variance of the resulting biomass estimates.