Bottom trawling is one of the most pervasive forms of fishing disturbance in marine ecosystems, yet its effects on the trophic organisation of benthic-demersal assemblages remain difficult to detect using taxonomic or biomass-based descriptors alone. Here, we integrate stable isotope analysis, Bayesian isotopic niche metrics and functional trait clustering to assess whether chronic trawling acts as a trophic-functional filter on demersal and benthic assemblages of the central Mediterranean shelf (Sicily Malta platform), across gradients of depth and swept area ratio (SAR). Trait-based ordination identified six ecologically interpretable functional clusters, separating seafloor-associated, low-mobility benthic invertebrates from mobile demersal predators along gradients of locomotion, feeding strategy and reproductive investment. Functional clusters showed coherent differences in mean δ¹⁵N values, whereas their broader differentiation in trait space than in isotope space indicated that traits and isotopes captured complementary dimensions of assemblage organisation. Environmental fitting further showed that isotopic structure was more closely associated with local productivity and near-seafloor physical conditions, particularly chlorophyll, temperature and bottom stress, than with depth. Fishing intensity was associated with a marked non-linear response in isotopic niche structure. Low-SAR assemblages occupied broader isotopic space, whereas assemblages exposed to intermediate-to-high SAR values showed the strongest contraction across multiple niche metrics, consistent with trophic-functional compression. At the highest SAR values, isotopic space partially re-expanded rather than continuing to contract, suggesting trophic reassembly by resistant, opportunistic or trophically flexible taxa rather than ecological recovery. Our findings indicate that combining functional traits and isotopic niche metrics can reveal food-web reorganisation that may remain hidden when trawling impacts are assessed only through taxonomic composition, abundance or biomass. This integrated framework adds a food-web integrity dimension to SAR-based and trait-based assessments, helping to identify candidate zones where chronic trawling coincides with trophic-functional simplification or reassembly. Trait-informed isotopic metrics therefore offer a promising tool for ecosystem-based fisheries management and for monitoring the ecological consequences of reducing or redistributing trawling pressure.