Scale of effect matters: Forest cover influences on tropical ant-plant ecological networks

Abstract

In the last decade, the number of studies dealing with how species interaction networks are structured by landscape has increased in the literature. However, many studies quantify landscape metrics at one single spatial extent, without considering the effects at multiple spatial extents. To the best of our knowledge, no study has explicitly addressed and discussed the spatial extent at which landscape structure best predicts the organization of ecological networks, which is key to understanding how ecological interactions are affected by landscape modifications. Therefore, two important research questions arise: can the spatial scale at which landscape structure best predicts an ecological response (i.e., scale of effect) magnify our ability to understand ecological networks or can the suboptimal scale just blind our view about what really happens? Motivated by these concerns, we used ant-plant relationships obtained from 16 landscapes in a human-dominated tropical rainforest of Mexico to assess and discuss the spatial extent at which forest cover (%) best predicts ant-plant interaction network patterns (i.e., network size, interaction diversity, network specialization, and nestedness). In general, we observed that smaller scales best predicted the interaction diversity (50 m) and specialization (250 m) of ecological networks involving ants and plants, which are descriptors associated to abundance and the distribution of interactions between plant and ant species. This is possibly because ant-plant relationships are being modulated by processes that act at the microclimatic level due to the biotic and abiotic conditions that forest provides. However, we found that network size and nestedness (i.e., descriptors associated to species detectability and emergent properties) were not associated with forest cover at any analyzed spatial extent. These findings indicate that the ant-plant ecological network structure can be influenced by processes that occur at different spatial extents. Moreover, the results presented here have a direct impact on our theoretical and empirical understanding of how the landscape structure shapes biotic interactions at different spatial extents in highly diverse tropical rainforests.