The evaluation of the status of marine communities, and especially the monitoring of those heavily exploited by fisheries, is a key, challenging task in marine sciences. Fishing activities are a major source of disruption to marine food webs, both directly, by selectively removing components at specific trophic levels (TL), and indirectly, by altering habitats and production cycles. Food web analysis can be very useful in the context of an Ecosystem Approach to Fisheries, but food web reconstructions demand large and expensive data sets, which are typically available only for a small fraction of marine ecosystems. Recently, new technologies have been developed to easily, quickly and cost-effectively collect environmental DNA (eDNA) during fishing activities. By generating large, multi-marker metabarcoding data from eDNA samples obtained from commercial trawlers, it is possible to produce exhaustive taxonomic inventories for the exploited ecosystems, which are suitable for food-web reconstructions. Here, we integrate and re-analyse the data of a recent study in which the α diversity was investigated using the eDNA opportunistically collected during fishing operations. Indeed, we collect highly resolved information on species feeding relationships to reconstruct the food webs at different sites in the Strait of Sicily (Mediterranean Sea) from eDNA and catch data. After observing that the trophic networks obtained from eDNA metabarcoding data are more consistent with the available knowledge, a set of food web indicators (species richness, number of links, direct connectance and generality) is computed and analysed to unravel differences in food webs structure through different areas (spatial variations). Species richness, number of links and generality (positively) and direct connectance (negatively) are correlated with increasing distance from the coast and fishing effort intensity. The combined effects of environmental gradients and fishing effort on food web structure at different study sites are then examined and modelled. Taken together, these findings indicate the suitability of eDNA metabarcoding to assist and food web analysis, obtain several food web-related ecological indicators, and tease out the effect of fishing intensity from the environmental gradients of marine ecosystems.