Plant architecture and habitat variation of the exotic ant-plant Ricinus communis (L.) across tropical landscapes

Background/Question/Methods
Ant-plant protective mutualisms, particularly facultative mutualisms, can be heavily influenced by abiotic factors, and consequently these associations vary greatly across landscapes, habitats, and seasons. Previous work has demonstrated that the attraction of native ant species by an exotic ant-plant can result in the formation of reward-based facultative associations and disrupt pre-existing and coevolved native interactions. The objective of this study was to evaluate the natural history of the facultative interactions with the exotic ant-plant Ricinus communis across habitats in Costa Rica. We examine the following questions: (1) What is the variability of plant architecture (Leaf size and herbivory, extrafloral nectary [EFN] size and abundance) across all habitats?; (2) Is there a relationship between ant diversity, EFN variability and plant herbivory?; and (3) Does ant-exclusion have an effect on herbivory across all habitats? A total of 360 plants were sampled from 20 sites across four habitat types (abandon pasture, roadside, plantation, and forest).  Leaf area, percent herbivory, EFN size and abundance were measured for each individual and three transects were run to determine the ant species abundance and diversity at each site.  In addition, an ant exclusion experiment was performed to examine the role of defense by ants in this exotic ant-plant species.  . 

Results/Conclusions
A Principle Components Analysis revealed that plant architecture varied across habitats, as did ant abundance, along a disturbance regime gradient closely related to the different habitats.  No significant relationship was detected between herbivory levels and either EFN size or EFN abundance individually; however, a surface plot of said variables demonstrated that percent herbivory increased when both EFN size and EFN count decreased. A total of 35 ant species were found visiting R. communis across all habitats sampled in this study. Richness estimates for all Costa Rican study sites suggest that there are between 44 (Chao 2) and 48 ant species (ICE) visiting R. communis EFNs.   The ant exclusion experiment revealed that mean herbivory levels did not differ between treatments.  However, plants without access to ants had herbivory levels that were significantly more variable than plants with access to ants (P<0.001).  These results suggest that the disturbance regime plays an important role in shaping plant architecture as well as determining the strength of plant defense via facultative ant-plant interactions.  Prior studies have focused mainly on the static effects of exotic species on native biodiversity; however, this study documents the dynamics of non-native interspecific interactions across native habitats.