Mutualisms are highly context dependent, and their effects can be changed dramatically by the presence of additional mutualistic partners. Many studies have tested interactions among multiple mutualists, finding that their interactions range from antagonistic to synergistic. Antagonistic interactions are expected when mutualistic partners compete for resources, resulting in reduced benefits from multi-mutualist interactions than expected from two-species interactions. In contrast, synergistic effects are expected when mutualists do not compete for resources and provide complementary benefits to their partners. This conceptual framework is strongly linked to ANOVA statistical tests, with the null hypothesis being additive effects of multi-mutualist interactions, and alternative hypotheses being antagonistic or synergistic effects. While the conceptual theory and empirical research in this field is well developed, multi-mutualist interactions are not well linked to mathematical theory or models. Here we modify Lotka-Volterra models of mutualist interactions to model multi-mutualist effects, and we compare model predictions to multi-mutualist interactions in a model plant system, Chamaecrista fasciculata. We test the fitness effects of mycorrhizal fungi, rhizobia, and ant mutualists on the seed production of this annual legume.
Results/Conclusions
Models of mutualistic interactions demonstrate that additive per capita effects of mutualisms can lead to synergistic responses in the populations of mutualistic partners. However, if these mutualistic partners compete for resources, the potential for synergistic interactions is diminished. At high levels of competition, our models predict that the benefits of two mutualists are reduced to the average effects of the mutualistic partners. Additive effects, as modeled by ANOVA interactions, only emerge with precisely balanced mutualistic and competitive effects. Consequently, we suggest that the null hypothesis of additive multi-mutualist effects in an ANOVA are not connected to theoretical expectations, and we ought to shift our focus toward measuring per capita effects as tests of multi-mutualist theory. In our empirical tests of mutualistic effects on Chamaecrista, we find that mycorrhizal fungi, rhizobia, and ants all increase the fitness of their host, while their pairwise effects are similar to their effects in isolation. These pairwise effects suggest strong competition among these mutualists for resources from their plant host. However, much to our surprise, the combined effect of all three mutualists far exceeds what we would have predicted from the effects of two mutualists, highlighting the potential difficulty of predicting complex species interactions.
