Effects of attack frequency on the tolerance to herbivory of Neotropical savanna trees

Background/Question/Methods Tolerance is the ability of a plant to maintain the same level of growth and reproduction following damage. Experimental studies typically measure tolerance in response to the intensity of herbivory (i.e. the amount of leaf tissue removed in one attack). However, the effects of attack frequency (i.e. how many times plants are attacked during a growing season) on plant tolerance is virtually unexplored. Using experimental defoliations that mimicked patterns of attack by leaf-cutter ants (Atta spp.), we examined how frequency of herbivory influenced plant tolerance traits in six tree species of the Brazilian Cerrado. For two consecutive years we completely and repeatedly defoliated treatment trees in two attack frequencies, monthly and quarterly, while also maintaining an untouched control group. Following our first defoliation, we quantified how the frequency of clipping influenced individual survivorship, relative growth rate, foliar nitrogen content, architecture (i.e. branching pattern), and flowering.

Results/Conclusions Our results showed little evidence that our focal tree species are tolerant to repeated and consecutive attacks by leaf-cutter ants. The content of leaf N increased among clipped individuals of most species, suggesting that Atta influences the allocation of resources in damaged plants. In addition, our clipping treatments affected tree architecture in ways thought to promote tolerance (i.e. increased tree branching). However, none of our focal species exhibited a compensatory increase in growth (increment in trunk diameter) in response to herbivory – relative growth rates were significantly lower in clipped than in unclipped individuals. Furthermore, the probability of survival was much lower for clipped plants, and lower for plants clipped monthly than for those clipped quarterly. The consequences of leaf-cutter herbivory for survival are magnified in the dry season, during which their activity continues unabated. For plants that did survive, simulated herbivory dramatically reduced the probability of flowering. That our results were similar across a phylogenetically distinct suite of species suggests a potential extendability of these findings to other plant species in this system. More importantly, given that Atta spp. attacks increase mortality, reduce growth, and reduce flowering, ant activity could contribute to lower population growth rates and potentially lead to a shift in community composition towards non-preferred species.