Friday, August 6, 2010 - 8:20 AM

COS 122-2: Defensive floral traits shape ant-flower networks in Hawai’i: Patterns and consequences for endemic species

Robert R. Junker1, Curtis Daehler2, Richard Bleil1, and Nico Blüthgen1. (1) University of Würzburg, (2) University of Hawaii

Background/Question/Methods

Invasive alien species are often well integrated in interaction webs with varying effects on native species. On the Hawaiian Islands, which represent the largest landmass that evolved in absence of ants, we observed interactions between invasive ants and flowers of native and introduced plants. We formulated null-model expectations of the interaction networks by considering the nectar quantity and quality and the ant species’ composition, which enabled us to compare the interaction strength between ants and flowers of native versus introduced plant species. Furthermore, we examined floral morphology, responses to floral scents and nectar palatability to evaluate the influence of these features on the network structure. Additional to the network approach, we examined the effects of ants on other flower visitors and on the reproduction of two endemic plant species.

Results/Conclusions

We found that ants more heavily exploit floral nectar of native Hawaiian plant species than the nectar of introduced plant species. This uneven visitation rate was the result of floral defence mechanisms in flowers of introduced plant species that share an evolutionary history with ants. Native Hawaiian plants often lacked such defenses. These results, along with a phylogenetic analysis of the involved plant species, suggest that ants select for floral features that protect nectar against these antagonists. Flowers of Metrosideros polymorpha (Myrtaceae) were visited by nectar- and pollen-collecting invasive honeybees (Apis mellifera) and by endemic Hylaeus spp. bees that collected pollen only. On ant-visited flowers, honeybees shifted from nectar- to pollen-collection as a result from exploitation and interference competition with ants. Honeybees turned out to be more efficient pollinators compared to Hylaeus bees and their mutualistic service was not affected by the presence of ants measured as fruit set. On flowers of Vaccinium reticulatum (Ericaceae), the presence of ants decreased fruit parasitism by larvae of a plume moth (Stenoptilodes littoralis) and therefore strongly increased fruit set. We conclude that flowers of Hawaiian plant species are not adapted to ants in contrast to introduced plants that protect their flowers. This often leads to a massive exploitation of floral nectar by ants – which is not necessarily always detrimental to the Hawaiian flora.