Influence of mesoscale oceanographic features on pelagic food webs in the Gulf of Mexico

This study examined the influence of mesoscale oceanographic features (anticyclonic; warm core and cyclonic; cold core) on offshore pelagic food webs in the Gulf of Mexico. Mean total biomass (wet weight) of all consumers was significantly higher in samples collected within cyclonic features (mean 3.78 g per 10 min tow) than anticyclonic features (mean 0.51 g per 10 min tow) during each survey date. Using stable isotope ratios of carbon (δ13C) and nitrogen (δ15N), we contrasted the two main primary producers in this ecosystem: phytoplankton (based on particulate organic matter, POM) and Sargassum spp. over a 2-year period. In addition, consumers (zooplankton, six invertebrate species, and eight fish species) collected in upper surface waters were analyzed for δ13C and δ15N. Both producers and ten of the fifteen consumer species had significantly enriched 15N in cyclonic relative to anticyclonic features in year one and each of the six selected ‘model taxa’ collected during both years showed this same pattern. Model taxa included POM, Sargassum spp., zooplankton, glass shrimp (Leander tenuicornis), Sargassum crab (Portunus sayi), and blackwing flyingfish (Hirundichthys rondeleti). δ13C values were more variable and dependent upon feature and survey date. Contributions for the two primary producers were estimated using a two-source Bayesian mixing model. Results support equal contributions of organic matter from phytoplankton and Sargassum spp. to consumers, but estimates were species and feature dependent and nitrogen-fixing Trichodesmium was likely important. For example, contribution estimates of Sargassum-derived organic matter to zooplankton in anticyclonic features ranged from 68 to 76%, in contrast to cyclonic features that varied from 29 to 83%. This study highlights the differences in δ13C and δ15N among producers and consumers collected within mesoscale oceanographic features in the Gulf of Mexico and demonstrates the need to obtain feature-dependent baseline estimates for calculating contribution estimates using stable isotope mixing models.