Higher-order interactions and indirect effects: a resolution using laboratory Drosophila communities

The importance of higher-order interactions to community structure has been debated for decades. We attempt to resolve the conflict by placing higher-order interactions in their appropriate functional relationship with indirect effects. Indirect effects are changes in the outcome of a pairwise relationship caused by other species. These changes can be mediated through changes in species densities (altering the frequency of encounters) or through a change in the nature of the relationship itself (such as a change in the per capita competition coefficients in Lotka-Volterra models). These latter forms of indirect effects, limited by their specific mechanistic definition, are higher-order effects. They can be resolved empirically only if density effects can be accounted for, and this additional analytical requirement may explain their scarcity. We also test the competitive relationships among mycophagous drosophilids and describe an indirect effect. However, there was no evidence of a significant higher-order effect; the nonadditive effect is apparently mediated through changes in species densities. Previous laboratory studies of multispecies Drosophila communities failed to document nonadditive effects. We suggest that the multigenerational approach of earlier studies may obscure indirect effects that could be important to natural communities.