Novel ant-plant mutualisms affect communities: Indirect interactions between extrafloral nectar-bearing plants and honeydew-producing insects via invasive ants

Background/Question/Methods: Indirect interactions can occur when a single mutualist species is shared by multiple partners. While interactions among taxonomically similar guilds of mutualist species have recently received increased study, much less is known about the dynamics of 'interguild' mutualisms. Ants commonly form facultative associations with both extrafloral nectary (EFN) bearing plants and honeydew-producing insects (HPI). Ants consume plant and/or insect-derived carbohydrates and provide partners with protection against natural enemies. Therefore, ants have the potential to mediate indirect interactions between EFN-bearing plants and HPI, possibly shifting the direction of the plant-HPI interaction from (-/+) to (+/+).  The consequences of such shifts could cascade to influence the structure of both plant and arthropod communities. We evaluated these dynamics across several sites comprising communities of invasive ants and native, EFN-bearing plants. To explore the context-dependency of these indirect interactions, we also tested how the quantity of EFN resources affected changes in ant behavior toward HPI. We simultaneously manipulated ants (ambient / reduced), HPI (ambient / reduced), and extrafloral nectar levels (ambient / reduced) on EFN-bearing Morinda citrifolia plants at the plot-level (4mX4m). This experiment was replicated across sites dominated by the highly invasive ant species, Anoplolepis gracilipes (n=6 replicates) and sites dominated by other ant species (n=5 replicates). Three and six months after treatment application, we recorded the identity and abundance of ants, HPI, and other arthropods on M. citrifolia plants. We also measured the growth of M. citrifolia plants. Finally, we conducted a plant-level experiment to evaluate ant-tending behaviors toward HPI across a gradient of plant nectar availability (0-200% ambient levels).

Results/Conclusions:  Here, we report the responses of HPI, A. gracilipes, other ants, and M. citrifolia plants. Anoplolepis gracilipes responded more strongly to plant nectar than HPI abundances, with strong reductions in abundance when nectar levels were reduced, but only a small, non-significant reduction in abundances in plots with reduced HPI abundances. Conversely, other ant species had higher abundances in plots with reduced extrafloral nectar, but decreased abundances in plots with HPI reduced.  Furthermore, ants had a positive influence on M. citrifolia growth, but only at sites dominated by A. gracilipes.  Finally, ant tending of HPI significantly declined as nectar levels increased at sites dominated by A. gracilipes.  However, insect tending by other ant species was consistently high, with no significant changes due to nectar availability. These results demonstrate that invasive ants can alter complex interactions between EFN-bearing plants, their ant bodyguards, and HPI.