Linking net interaction effects with network topologies in food webs

Aim The difficulty in predicting responses to press perturbation is largely due to the tangled nature of indirect interactions in natural ecological networks. Although previous theories have paved the way for evaluating net interaction effects by incorporating direct and indirect effects related to press perturbation, there are still critical knowledge gaps regarding the relationships between net interaction effects and network topologies because of lack of comparable interaction strengths across ecosystems. Location Global. Time period 1984-2020. Major taxa studied Aquatic organisms. Methods We compiled a global dataset of food webs with a consistent modelling method to estimate direct interaction strength. We revealed the general patterns of net interaction effects and their associations with network topologies by applying matrix theory and network analysis. Results In the examined food webs, the net (absolute) interaction strengths followed a right-skewed pattern (many weak and few strong net strengths) and were significantly associated with various network- and species-level topological structures. Indirect effects led to the emergence of net mutualism and competition effects rather than simply net predator-prey interaction effects, and might play an important role in changing signs from direct to net interaction effects in species-rich and highly connected food webs and for species with a large number of interaction partners. Main conclusions Our results highlight the ecological significance and necessity of linking net interaction effects with network topologies, which can help target ecosystems and species that are more vulnerable to sustained perturbation in the application of conservation and biodiversity management.