Assessing survey design changes of long-term fishery-independent groundfish trawl surveys in the Gulf of Mexico

Long-term fishery-independent surveys provide a wealth of information on fisheries stocks that inform stock assessments. One of the strengths of these surveys is that the design and methods are consistent through time. However, maintaining an unchanged time series can pose several potential issues as management needs change resulting in the need to alter either the survey design or its spatial extent. In the United States Gulf of Mexico, bottom trawl surveys targeting groundfish and shrimp (hereafter, groundfish surveys) have been conducted since the 1950s, with standardized surveys beginning in 1972. The resulting data can provide a great deal of information on commercially and recreationally important species. However, many of the alterations to the survey design have been buried in gray literature or otherwise poorly noted. The history of these surveys is discussed, along with the rationale behind these changes and the impacts they had on stock assessments in the region. Starting in 1981, the Southeast Area Monitoring and Assessment Program began groundfish surveys, continuing the time series. Over time, the survey’s spatial extent and design have been altered to meet changing management needs. Delta-lognormal models are used to draw inferences concerning the effects of the survey design change on the relative abundance and their associated coefficients of variation for several commercially and recreationally important species. The expansion of the surveys across the Gulf of Mexico is examined in relation to stock assessments. Overall, the design changes and spatial expansion have been beneficial from a stock assessment standpoint, resulting in an increase in the number of indices used for single-species stock assessments and the utility of survey data in support of ecosystem modeling efforts. Finally, a discussion around the lesson learned (i.e., the critical need for overlap of survey designs), emphasizing the potential impacts of these changes on the overall time series concerning stock assessments, is presented.