Evaluating potential changes to the US Chukchi Sea bottom trawl survey design via simulation testing

Abstract

The US Chukchi Sea consists of the waters off the northwest of Alaska and is a naturally dynamic ice-driven ecosystem. The impacts from climate change are affecting the Arctic marine ecosystem as well as the coastal communities that rely on healthy marine ecosystems. In anticipation of increased ecosystem monitoring in the area, there is an opportunity to evaluate improved sampling designs for future ecological monitoring of the Chukchi Sea, an area that is sampled less comprehensively compared to other regions in Alaska. This analysis focused on standardized NOAA-NMFS-AFSC bottom trawl surveys (otter and beam trawls) and three types of survey designs: simple random, stratified random, and systematic. First, spatiotemporal distributions for 18 representative demersal fish and invertebrate taxa were fitted using standardized catch and effort data. We then simulated spatiotemporal taxon densities to replicate the three survey design types to evaluate design-based estimates of abundance and precision across a range of sampling effort. Modest increases in precision were gained from stratifying the design when compared to a simple random design with either similar or lower uncertainty and bias of the precision estimates. There were often strong tradeoffs between the precision and bias of the systematic estimates of abundance (and associated variance) across species and gear type. The stratified random design provided the most consistent, reliable, and precise estimates of abundance indices and is likely to be the most robust to changes in the survey design. This analysis provides a template for changing bottom trawl survey designs in the Chukchi Sea and potentially other survey regions in Alaska going forward and will be important when integrating new survey objectives that are more ecosystem-focused.