Land use changes affect ant assemblage and ecological function in different way in brazilian savanna

Abstract

Tropical savannas are highly diverse in terms of species richness and ecosystem services, yet are subject to intense anthropogenic pressure with high rates of habitat loss and degradation. In the Cerrado (Brazilian savanna) more than half of native areas have been transformed into pasture and agriculture, and most studies have evaluated biological communities structuring and impacts of land use changes within one vegetation type and focusing only on deforestation process. Such fact raises difficulties advances in understanding alterations in biological communities of native fauna from Cerrado in general. Thus, this thesis aimed to evaluate factors structuring biological community in local and landscape scales, and the impacts of land use changes (Eucalyptus plantation and planted pasture) on ant community and ecological function in three Cerrado vegetation types (grassland, savanna and savanna-forest) through both tree less and afforestation (gain of tree) processes. I sampled ants using epigaeic pitfall and ants removing seeds (ecological function) in areas of native Cerrado, Eucalyptus and planted pasture in each vegetation type. In chapter 1 I assessed the factors influencing ant assemblage in native habitats in different Cerrado vegetation types at local and landscape scales. I observed that ant assemblage in native habitats is predominantly influenced by landscape factors, indicating the importance of surrounding native habitats and to include landscape factors in management and conservation strategies. In chapter 2 I used taxonomic and guild approaches to evaluate if land use changes, by tree loss and afforestation, affect ant species and guilds in a similar way. I found that response of taxonomic structure to land use changes do not necessarily lead to the same response of guild structure regardless of the type of change. This shows that both approaches, taxonomic and guild, contribute to improve our understanding about community ecology in conversion of native habitats into agroecosystems. In the chapter 3 I examined the impacts of land use changes on seed removal by ants and if these impacts were correlated with changes in habitat attributes. I observed that these impacts on seed removal depend on the type of land use change (afforestation or tree loss), and are correlated to the similarity in habitat attributes between agroecosystems and native habitat. Overall, this thesis shows that ant assemblage is dependent on landscape factors and that land use change, when evaluated simultaneously across different vegetation types, have different impacts on ant assemblage and their ecological function of seed removal. Therefore, identifying factors structuring ant assemblage, as well as the spatial scale they operate, and the impacts of land use change in different vegetation types may contribute to improve our understanding about the response of ecological communities and applied management and conservation strategies in the Cerrado.