Species distribution modeling for king mackerel (Scomberomorus cavalla) and its prey species in the Gulf of Mexico
MetadataShow full item record
Ecosystem based fisheries management (EBFM) has been broadly recognized throughout the world as a way to achieve better conservation. Therefore, as an important part of EBFM, mapping multi-species interactions or spatial distributions has been strongly needed. Species distribution models are widely applied since information regarding the presence of species is usually only available for limited locations due to the high cost of field surveys. Furthermore, a large proportion of the fisheries survey data have only presence records instead of regular presence and absence records, Thus, presence-only species distribution models are needed. In this study, four presence-only species distribution algorithms (Bioclim, Domain, Mahal and Maxent) were applied using 12 environmental parameters as predictors to model the distribution of king mackerel (Scomberomorus cavalla) and 31 of its prey species in the Gulf of Mexico. Based on the results, 10 major distribution patterns were proposed to describe the distribution of the 32 species. Post hoc with Tukey’s test shows that area under curve (AUC) for the Maxent-based models were significantly (p<0.05) higher than those for Bioclim and Domain based models, but insignificantly different from those for Mahal-based models (p=0.955); while correlation coefficients (r) for the Maxent-based models were significantly higher than those for all the other three types of models (p<0.05). Thus, Maxent-based models were concluded to have the best performance. Generalized linear models (GLM), generalized additive models (GAM) and random forest models (RF) were applied to model the abundance distribution of three shrimp species throughout the Gulf. Results show that abundance distributions predicted were quite close to the species distribution predicted by the presence-only models, which validated the good performance of the presence-only models. Evaluation of the models by correlation shows that the GAM models had the best performance for brown shrimp abundance modeling, while the RF models had the best performance for the other two shrimp species. Good performance of the species distribution/abundance models shows that interesting distribution patterns, especially the special zones (eg. the dead zone), can provide some insights for scientists or government managers to better manage fisheries resources in the Gulf of Mexico.