Redes de interações entre aves e frutos em habitats antropizados no interior do estado de São Paulo, Brasil

Abstract

Interaction networks represent species that interact, and their analysis provides a systematic view of community properties and trophic interactions. When evaluating anthropogenic disturbances in interaction networks involving frugivory, it is also important to consider the presence of exotic plant species, as understanding their effects can inform conservation strategies. This study aimed to characterize and analyze the topology of bird-fruit interaction networks in rural and urban areas in the interior of São Paulo State, considering the anthropized landscape, the presence of exotic species, and the role of ecosystem services. We hypothesized that network topology would vary depending on the predominant landscape matrix, revealing: (i) greater generalization, connectance, and interaction homogeneity in urban areas; (ii) persistence of these characteristics even in networks composed exclusively of mutualistic interactions and native plants; (iii) a higher proportion of exotic species in urban areas; and (iv) an increase in specialization when exotic species are excluded. Data were collected in three urban green areas and three rural vegetation fragments in the Sorocaba region, through transect walks and focal observations of fruiting plants to record bird-fruit interactions. Quantitative interaction networks were constructed and characterized using metrics such as total number of interactions, weighted connectance, interaction evenness (E2), network specialization (H2'), weighted nestedness (wNODF), and modularity (Q). Our results partially confirmed the hypothesis: urban areas showed lower persistence of specialists and higher prevalence of generalists, as indicated by modularity, specialization, and nestedness metrics, while interaction evenness and connectance did not differ significantly. Contrary to expectations, exotic species were not more frequent in urban areas, indicating their substantial impact in rural areas as well. As predicted, exotic species had a generalist role, and their exclusion increased network specialization. These findings help clarify how anthropogenic changes and spatial context influence interaction network topology, highlighting their impact on community composition in green areas embedded in human-modified landscapes.