La Peyre MK, Sable S., Taylor C, Watkins KS, Kiskaddon E, Baustian M. 2021. Effects of sample gear on estuarine fish assemblage assessments and food web models. Ecological Indicators 133:108404 https://doi.org/10.106.j.ecolind.2021.108404.
Abstract
Long-term fisheries-independent sampling data inform population status and trends of species specific biomass and are often used to drive biomass-based food web models such as the Comprehensive Aquatic Systems Model (CASM). Indicators such as total biomass and mean trophic level derived from these data and from CASM outputs inform management and facilitate assessments of on-going and predicted coastal change and restoration activities on fisheries, but rely on consistent sampling to enable comparisons across space and time. Changes in coastal estuarine gradients, combined with the availability of new sampling technologies, highlight a need to assess the potential consequences of changing sampling technologies on fisheries data and the cascading impact on model outputs. In Louisiana, USA, CASM models are used to inform coastal restoration projects, relying on 40 years of fisheries independent data derived from 50’ seine sampling. However, alternative use of electrofishers as a sampling method has been proposed to replace the seine sampling. In this study, we examine data from concurrent seine and electrofisher sampling in Barataria Basin, Louisiana, and compare biomass, assemblage data and CASM outputs related to species biomass, food web structure and energy cycling. In a paired comparison of data in 2018-2019, the electrofisher captured higher total catch and diversity compared to the seine. The electrofisher samples were dominated by shrimp (grass, white, brown) and larger bodied fish, while seine samples were dominated by smaller-bodied fish (i.e., anchovy, menhaden). Ecosystem indicators derived from running the CASM using biomass data from seine and electrofisher sampling separately provide contrasting results. Although initializing the CASM biomasses for key species using seine and electrofisher data to begin the model simulations did not cause large or long-running changes in the simulated biomasses over time, results using the electrofisher biomasses substituted into the original calibrated CASM on day 500 in a different simulation resulted in changes in biomass structure, with a doubling of total food web biomass due to the increased shrimp count, and a 13% increase in total energy flow (TST) through the food web. Conversions based on area and gear efficiency for overall catch may be useful in maintaining the continuity of historical data. However, differences in species-specific catch due to gear selectivity could have large consequences for constructing and calibrating fish and ecosystem models and remain difficult to reconcile. These differences in assemblages, and estimated biomasses for key food web species, suggest careful consideration in changing gears.