Community-wide consequences of modeling mutualistic benefits as more direct than they are

Background/Question/Methods

Mutualistic benefits are often modeled as direct positive effects of one species on another. However, in reality they often involve species or resources outside the mutualism: benefits provided by ant protectors for example depend on the abundances of natural enemies (i.e. consumers/predators), and benefits provided by mycorrhizal fungi depend on soil nutrient pools. These structures place mutualisms within the realm of ‘modified interactions’ in food webs, and may affect not only mutualists themselves but also species in the wider community. We examine food webs containing mutualisms (defined here by having positive effects on one another's instantaneous growth rate) in which benefits to a focal partner are either: 1) direct, 2) mediated through reduced interaction between mutualists and their consumers, or 3) mediated through increased mutualist uptake of resources. We compare the effects of each type of mutualism on equilibrium populations of the focal mutualist's resources and consumers, and in turn on competitors (species that share resources with the mutualist), and apparent competitors (which share consumers with the mutualist), as well as more distant species. We also explore how each type of mutualism affects system responses to various disturbances.

Results/Conclusions

We show that the different types of benefits can have qualitatively different effects on species outside the mutualistic pair, which in turn can depend on the web in which the mutualism is embedded. For example, direct benefits generally reduced the density of the focal mutualist's resource when mutualists were resource-limited, but had no effect when mutualists instead experienced top-down control by consumers. In contrast, resource-mediated benefits more consistently reduced resource density when mutualists experienced top-down control than when they were resource-limited. We further show that these effects often have counterintuitive outcomes, such as mutualists' competitors being the only species to experience reduced population size when a mutualism is disrupted or weakened. Benefits types also have different implications for assembly processes and cascading effects of species extinctions or invasions in large communities. These models show that while direct mutualistic benefits can at times have effects qualitatively similar to benefits mediated through additional species, they are generally quite different. This should be considered in future work incorporating mutualisms into models of large communities.