Composition and structure of anthropogenic landscapes shape the interactions among plants, frugivorous bats and ectoparasites

Abstract

The loss of natural areas for agriculture has resulted in landscapes with different mosaics of land use, causing drastic changes in biodiversity. These environmental disturbances affect the conservation of wild populations, resulting in the loss of ecological interactions and the consequent reduction in ecosystem functions. In previous studies, it has been observed that populations of frugivorous bats can be impacted by landscape characteristics. These animals are extremely important for the maintenance and regeneration of native forests, because they promote the dispersion of seeds, allowing plants to reach degraded areas and to increase the gene flow in their populations. Another important factor that can affect populations of frugivorous bats is the occurrence of ectoparasites, which is also impacted by the characteristics of the landscapes. In the light of these facts, the present study has two main objectives. First, we intend to evaluate the influence of composition and structure of anthropogenic landscapes on mutualistic bat-plant and antagonist bat-ectoparasite interaction networks, aiming to understand if the characteristics of landscapes interfere distinctively with these types of interaction. Secondly, we expect to assess parasite prevalence and mean intensity of infestation in three species of frugivorous bats — Artibeus lituratus (Olfers, 1818), Carollia perspicillata (Linnaeus, 1758) and Sturnira lilium (É. Geoffroy, 1810) — and also to determine whether there is a typical response of the parasitic load associated with the ectoparasite group (mite or bat-fly). Our results indicated that antagonistic interactions were more affected by the landscape characteristics than mutualistic interactions. Forest cover was of remarkable importance: its increase reduced the diversity of interactions and increased specialization in bat-plant networks. In bat-ectoparasite networks, forest cover caused a decrease in the diversity of interactions, in the modularity, and in the richness of the two taxa, while nestedness increased. The proportion of edge perimeter per sampled patch area influenced only bat-ectoparasite networks, and increased perimeter caused a reduction in the diversity of interactions and an increase in specialization. The heterogeneity of the landscape also influenced only bat-ectoparasite interactions, with increased heterogeneity resulting in more nestedness and less modularity. Regarding the parasitic load, there was great variation between the interactions studied, and we could not identify any typical variation associated with the ectoparasite group (mite or bat-fly). The proportion of infested hosts was more affected by the landscape than the mean infestation values were. An increase in landscape heterogeneity resulted in very different responses among the evaluated interactions. On the other hand, an increase in forest cover reduced the parasitic load in most of the interactions. An increase in the proportion of edge perimeter per sampled patch area had an influence apparently related to the taxa of the host. It is likely that the influence of landscape on parasitic load occurs due to characteristics that interfere in the rates of encounter between hosts (for mite or bat-fly) and between hosts and ectoparasites in roosts (for bat-fly). These findings demonstrate the great importance of the characteristics of anthropogenic landscapes in the organization of ecological interactions between plants, bats and ectoparasites.