The Brazilian Amazon comprises around 40% of all tropical forests in the world, harbouring a gigantic part of the world’s biodiversity. Yet, alarming rates of deforestation are spread across the Amazon. Increasing rates of resource demands harm this ecosystem and conservationists need precise information about biodiversity and ecosystem processes, as well as socioeconomic data. Within the needings of this agenda, a research initiative called Sustainable Amazon Network was developed, aiming to generate data that can foster conservation plans for the regions. This thesis brings results regarding some of the biodiversity aims of this research network. In the first chapter we aim to provide the most comprehensive species list of ants, orchid bees and dung beetles to date to the region of Paragominas, as well as describe patterns of species diversity across the different land-use types in the region. In the second chapter, we aimed to answer how land-use changes affect ant species composition and richness in Paragominas? We also aim to explore which environmental variables regulate ant species richness at local and regional scales. Finally, for the third chapter we ask whether there is an ongoing process of biotic homogenization happening with forest degradation and conversion to production landscapes. We surveyed five taxa (plants, birds, dung beetles, ants and orchid bees) in this chapter, across the municipalities of Paragominas and Santarém. In the first chapter we provided the list, which we hope will be a baseline for the monitoring of forest conservation initiatives taking place in the region. We also found that species richness of the three groups is being reduced by deforestation and forest degradation. In the second chapter, we observe a marked shift in species composition with land-use changes, as well as a process of species loss, highly associated with primary forest cover, being this the variable with the highest importance in explaining species richness. Lastly in the third chapter, we found a general pattern of steady decrease in local diversity with land-use intensification whereas at larger scales this loss can only be noticed for forest versus non-forest comparisons. We found that β diversity is affected only at smaller scales, being higher in forests than in non-forest areas. There is, however, a marked increase in the contribution of nestedness to β diversity in non-forest habitats at both scales. Thus, as a general conclusion, we found that land-use changes are eroding biodiversity and if any action is readily implemented in the region to protect the extant forest patches, we can witness a severe biodiversity loss that is already being observed at the local scale. We do expect as a multidisciplinary network that our results can nurture joint strategies, considering natural and socioeconomic sides, which can culminate in a more wise conservation planning to the region.
