Leaf-cutting ants as ecological filters in forest communities in process of restoration: effect of herbivory in the establishment of Croton floribundus Spreng. seedlings

Abstract

Leaf-cutting ants are dominant herbivorous insects in Neotropics and have high potential for modifying and structuring plant communities, being considered as ecosystem engineers. Most damages caused by ants of the genus Atta and Acromyrmex are associated with the leaves of the trees and seedlings, through their herbivorous habits. In the case of areas undergoing restoration, ants have been considered as pests that hamper the establishment and survival of seedlings, and their control is usually necessary during the initial stages of planting. However, after a certain time of community development and their interactions, the action of cutters tends to be minimized, with the establishment of other ants functional groups. Yet, are not known its effects as ecological filters for the establishment of certain species, since they have preferences on their foraging habits. This study will evaluate the performance of leaf-cutting ants in the establishment of capixingui (Croton floribundus Spreng.) seedlings, a pioneer and native species, very common in semideciduous seasonal forests and areas in the process of restoration. Our hypothesis is that leaf-cutting ants act as a biotic filter, preventing the development of Croton floribundus Spreng. (capixingui) seedlings, due to their feeding preference on this species, and that the location of the seedlings on the nests and the habitat characteristics will also affect their degree of herbivory. So, eight nests will be selected, four inside the plantations and four in neighboring pastures. On each nest will be installed two plots of 5 seedlings, and two more paired plots outside the nests, at least 10 m away from them. The seedlings will be surveyed biweekly, being framed on a growing scale of herbivory (0 to 4) and evaluated for mortality and survival rates. Every three months the seedlings will be measured for the total height and diameter of the stem. In this same period, the incidence of PAR (photosynthetically active radiation) will be estimated over the plots between 12 and 13 h with simultaneous measurements in the open area to estimate the relative intensity of radiation. To test the effect of treatments on the degree of herbivory, we will use non-parametric tests, analysis of variance and regression analysis, depending on the type of variable analyzed.