Use of silvicultural landscapes by small mammals in the state of São Paulo, Brazil Studies in agricultural landscapes have shown that anthropogenic environments may be relevant to wildlife species. Given the extent of these areas efficient research methods are needed to assess its ecological value in terms of local diversity, such as the use of biological indicators. The main goal here was to evaluate the possible relationship between the native vegetation biomass and diversity of small mammals in a silvicultural landscape of São Paulo. The study was developed at Fazenda Três Lagoas, which is part of the "Programa de Pesquisa em Biodiversidade" (PPBio), at Angatuba municipality, State of São Paulo, and lasted nine months. From May 2015 to March 2016 nine sampling campaigns were carried over a grid of 15 pitfall traps, five in Eucalyptus plantations, five in native vegetation and five in abandoned pastures. In the Chapter 1 the distribution and abundance patterns of small mammals on a silvicultural landcape during the early second cycle of Eucalyptus plantation in Southeastern Brazil was evaluated. A total of 672 individuals from 15 species of the orders Rodentia (9 spp.) and Didelphimorphia (6 spp.) were captured and five new species were detected in the study area in comparison to the first cycle. Although these species are predominantly generalists, there were differences in the abundance of small mammals in the three environments. The presence of native vegetation patches associated with riparian areas and abandoned pastures and the role of small mammals as a food resource for predators on a trophic structure, gives a certain conservation value to silvicultural landscapes in southeastern Brazil. In the Chapter 2 the relationship between landscape metrics and patterns of diversity, distribution and abundance of small mammals in a silvicultural landscape were evaluated. The species richness and family diversity of small mammals have been predominantly determined by the percentage of native vegetation. However, the abundance of small mammals can be determined by an indirect estimate of the native vegetation biomass (i.e., NDVI), which can be considered a 3D landscape metric as it is related to volume. Therefore, to succeed in assessing the value of agricultural landscape, future studies should consider landscape context, as well as the landscape metrics associated with geospatial data analysis and the distinct responses of species.