Some species may have a larger role than others in the transfer of complex effects of multiple human stressors, such as changes in biomass, through marine food webs. We devised a novel approach to identify such species. We constructed annual interaction‐effect networks (IENs) of the simulated changes in biomass between species of the southeastern Australian marine system. Each annual IEN was composed of the species linked by either an additive (sum of the individual stressor response), synergistic (lower biomass compared with additive effects), or antagonistic (greater biomass compared with additive effects) response to the interaction effect of ocean warming, ocean acidification, and fisheries. Structurally, over the simulation period, the number of species and links in the synergistic IENs increased and the network structure became more stable. The stability of the antagonistic IENs decreased and became more vulnerable to the loss of species. In contrast, there was no change in the structural attributes of species linked by an additive response. Using indices common in food‐web and network theory, we identified the species in each IEN for which a change in biomass from stressor effects would disproportionately affect the biomass of other species via direct and indirect local, intermediate, and global predator–prey feeding interactions. Knowing the species that transfer the most synergistic or antagonistic responses in a food‐web may inform conservation under increasing multiple‐stressor impacts.