Benefits in a mutualism often involve a trophic interaction.  Ant protectors consume aphid honeydew or extrafloral nectar, mycorrhizal fungi consume plant carbon, and seed dispersers consume fruit.  However, reciprocal benefits are often less direct, and involve a separate consumer-resource interaction.  Ants reduce predation or herbivory on their partners, mycorrhizal fungi improve their partners' access to soil nutrients, and dispersal to more favorable habitats may improve plants' access to resources or reduce their exposure to natural enemies.  Mutualistic benefits, and therefore the profitability of the mutualism, may be more sensitive to fluctuations in the abundances of other species and resources for the partner receiving indirect benefits than for the partner receiving direct benefits.  What does this mean for mutualisms?  We develop models where a focal mutualist is consumed by its mutualistic partner, but in return gains either protection from a separate consumer or enhanced access to resources.  We also explore the consequences of the focal mutualist having varying degrees of control over the association, by being able to limit its consumption by the other mutualist at the cost of forgoing an equivalent amount of the indirect benefit it receives. 

Results/Conclusions

           Short-term interests of the focal mutualist were often at odds with its long-term interests.  In cyclical systems without an additional predator, intermittent periods could occur where investing in mutualistic partners to obtain more resources was profitable, and such intermittent investment could maintain the population of partners, even when population size of the focal mutualist would be higher if its partner were absent.  Interestingly, generosity towards mutualists constituted a selfish (as opposed to altruistic) act in this case, because mutualism benefits individuals in the short-term at the expense of long-term population growth.  The mutualism persisted because endogenously generated transient conditions favored generosity by the focal mutualist.  When benefits involved protection from another consumer, focal mutualists given complete freedom to regulate the association often caused the extinction of their partners, only to suffer lower average population size (due to increased predation by the other consumer) than when they had limited or even no control.  Common to both cases was the result that having control over mutualistic associations could be detrimental to a species when its decisions affected community feedbacks.  We discuss the potential for models like these to make comparisons across mutualisms and between mutualisms and other ecological relationships.