Contrasting indirect effects of an ant host on prey-predator interactions of symbiotic arthropods.

Indirect interactions occur when a species affects another species by altering the density (density-mediated interactions) or influencing traits (trait-mediated interactions) of a third species. We studied variation in these two types of indirect interactions in a network of red wood ants and symbiotic arthropods living in their nests. We tested whether the ant workers indirectly affected survival of a symbiotic prey species (Cyphoderus albinus) by changing the density and/or traits of three symbiotic predators, i.e., Mastigusa arietina, Thyreosthenius biovatus and Stenus aterrimus, provoking, respectively, low, medium and high ant aggression. An ant nest is highly heterogeneous in ant worker density and the number of aggressive interactions towards symbionts increases with worker density. We, therefore, hypothesized that varying ant density could indirectly impact prey-predator interactions of the associated symbiont community. Ants caused trait-mediated indirect effects in all three prey-predator interactions, by affecting the prey capture rate of the symbiotic predators at different worker densities. Prey capture rate of the highly and moderately aggressed spider predators M. arietina and T. biovatus decreased with ant density, whereas the prey capture rate of the weakly aggressed beetle predator S. aterrimus increased. Ants also induced density-mediated indirect interactions as high worker densities decreased the survival rate of the two predatory spider species. These results demonstrate for the first time that a host can indirectly mediate the trophic interactions between associated symbionts. In addition, we show that a single host can induce opposing indirect effects depending on its degree of aggression towards the symbionts.