Understanding the ecology of mutualisms becomes a particularly important task when considering agroecosystems, as many ecosystem services are associated with mutualistic interactions. Here we report on experiments associated with an indirect pest control mutualism between the arboreal nesting ant Azteca sericeasur and coffee. This system is particularly interesting because the indirect Azteca-Coffea mutualism emerges from an Azteca-scale insect (Coccus viridis) mutualism that takes place on the coffee plant which results in the protection of the coffee against herbivores and in particular against the Coffee Berry Borer (CBB) (Hypothenemus Hampei). Here we ask whether the benefit A. sericeasur provides to coffee increases with increasing density of the intermediary mutualist (scale insects, Coccus viridis). In a 45 ha plot we located 30 plants with both C. viridis and A. sericeasur. In each plant we chose six branches with varying densities of C. viridis, and excluded A. sericeasur on three of them. We then added 20 adult CBBs to simulate and after 24 hours we returned to see the number of fruits damaged on the control and exclusion branches.
Results/Conclusions
With high C. viridis density treatment consistently showing an effect from ants across both years, we suggest that this experiment provides what we understand as the first evidence of the reward of one mutualism to be dependent on the density of an intermediary mutualist. Furthermore, we also found that at low Coccus density Azteca only benefits Coffea in the beginning of the rainy season, and this effect is likely due to the fact that Coccus produces less sugars with higher precipitation thus altering the dynamics of the mutualism. The role of seasonality on the outcome of the Azteca-Coccus-Coffea complex is not so surprising, as the role of conditionality or context dependency in the outcomes of ecological interactions has become a prominent feature of ecology. We suggest that management practices in agroecosystems may provide a useful mechanistic framework for unpacking contingency in these systems. Finally we caution against management recommendations, as our study suggests someone studying the dynamics of the Azteca-Coccus-Coffea interactions in the dry season may come to very different conclusions than someone studying it in the rainy season, and any management suggestions that may arise from either year individually could have unseen consequences for producers.