Résumé
Recent theoretical considerations have highlighted the importance of the pelagic-benthic coupling in marine food webs. In continental shelf seas, it was hypothesized that the trophic network structure may change along an inshore-offshore gradient due to weakening of the pelagic-benthic coupling from coastal to offshore areas. We tested this assumption empirically using the eastern English Channel (EEC) as a case study. We sampled organisms from particulate organic matter to predatory fishes and used baseline-corrected carbon and nitrogen stable isotope ratios (delta C-13 and delta N-15) to determine their trophic position. First, hierarchical clustering on delta C-13 and delta N-15 coupled to bootstrapping and estimates of the relative contribution of pelagic and benthic carbon sources to consumers' diet showed that, at mesoscale, the EEC food web forms a continuum of four trophic levels with trophic groups spread across a pelagic and a benthic trophic pathway. Second, based on the same methods, a discrete approach examined changes in the local food web structure across three depth strata in order to investigate the inshore-offshore gradient. It showed stronger pelagic-benthic coupling in shallow coastal areas mostly due to a reorganization of the upper consumers relative to the two trophic pathways, benthic carbon sources being available to pelagic consumers and, reciprocally, pelagic sources becoming accessible to benthic species. Third a continuous approach examined changes in the mean and variance of upper consumers' delta C-13 and delta N-15 with depth. It detected a significant decrease in delta C-13 variance and a significant increase in delta N-15 variance as depth increases. A theoretical two-source mixing model showed that an inshore-offshore decrease in the pelagic-benthic coupling was a sufficient condition to produce the delta C-13 variance pattern, thus supporting the conclusions of the discrete approach. These results suggest that environmental gradients such as the inshore-offshore one should be accounted for to better understand marine food webs dynamics. (C) 2014 Elsevier Ltd. All rights reserved.