Evaluation of Carbon and Nitrogen Stable Isotopes of Red Drum in the Northern Gulf of Mexico

Red drum Sciaenops ocellatus are conspicuous members of the ichthyofauna in the coastal zone of the northern Gulf of Mexico and support a popular inshore recreational fishery. Red drum exhibits differences in distribution where younger fish (age <= 4 years, y) are found primarily in inshore estuarine locations, and older individuals are generally found in offshore locations near Mississippi's barrier islands. In this study, we evaluate stable isotope values of carbon and nitrogen in muscle tissue of red drum to (1) understand how these vary with the location (bearing) of capture, age (y), and total length (TL, mm), (2) describe how patterns of isotopic niche space and overlap vary with age group, and (3) evaluate if location category (estuary and island) of capture could be predicted using a machine learning approach. We found that patterns of delta C-13 and delta N-15 were variable with bearing, TL, and age. Contrasting patterns of age-specific delta C-13 and delta N-15 values are evident in younger fish (age <= 2 y and age 3 to 4 y). Fish from estuaries had lower delta N-15 and lower delta C-13. To understand age group-specific patterns of isotopic niche space and overlap, we employed a bootstrapping algorithm using SIBER and nicheRover. We observed a minimum overlap of isotopic niche space between the age group comprised of age 1 to 6 y individuals and the age group comprised of age 7 + y individuals. The area of the isotopic niche space was greatest for the age group comprised of age 1 y individuals and for the age group comprised of age 1 to 7 y individuals. To understand if the location category of capture could be predicted, we constructed a random forest model. We found that the variables delta C-13, delta N-15, and TL are the most important predictors of the location category of capture. The out-of-bag error rate for the classification was 5%. The movement dynamics of red drum continue to challenge managers charged with the sustainability of the stock. This work highlights the utility of stable isotope analysis for understanding changes in diet and habitat use of red drum and the use of length as a predictive variable for the prediction of location category.