A GENERAL TEST FOR INTERACTION MODIFICATION

How can measurements of population dynamics be used to deduce the mechanisms of interaction in an assemblage of species? We present a technique to use such measurements to distinguish among assemblages of species with no direct interactions, sets of species influenced predominantly by pairwise interactions, and communities with significant interaction modification. We define the interaction coefficients and show that their dependence on the population sizes of the various species reveals and pinpoints interactions. Our technique distinguishes non-additivity in the statistical sense from interaction modification in the biological sense.