The partitioning of gamma diversity into its alpha and beta components has been used to quantify the effects of natural habitat reduction on species diversity and distribution. Ant assemblages respond to anthropogenic changes, allowing their use as an indicator of biodiversity conservation in human-modified landscapes. Additionally, ant species present different habitat-uses, such as forest specialist, open-habitat specialist and generalist that can live in both habitat types. These different ant groups can present distinct responses to the same human impact. Southwestern Brazilian Amazon, has experienced high levels of forest shifting to human land-use systems, mainly pastures, since 70-80s. Thus, we investigated how gamma, alpha, beta diversity, and habitat-use guilds react to a forest cover gradient in Assis Brasil, AC, Brazil. We carried out the ant sampling in 12 circular areas of 785.000 m2) with different forest cover percentage (13 – 100%). In each circular area, we sample ants along transects at the four radials and in each transect, we installed four pitfall traps at ground level (100 m apart). Gamma diversity (ɣ) was the general number of ant species sampled in each circular area, alpha diversity (α) as the average number of species per transect within each sampling area, and beta diversity (β) as a proxy for species composition change among the four transects within each sampling area, which was calculated as Bray-Curtis dissimilarity index (βBray) and additionally partitioned into its turnover (βBray.bal) and nestedness (βBray.gra) components, the first referring to species replacement and the last to gain or loss of species. Additionally, we accounted the number of species of each ant groups of habitat affinity, forest specialist, open-habitat specialist and generalist. Gamma and beta diversity, but not alpha, increased along the forest cover gradient. Turnover was the main component of beta diversity and increased with forest cover. The number of species of ant forest specialist increased, open-habitat decreased, and generalist did not changed. Thus, areas with few forest cover harbor a small number of ant species, which species have a high distribution leading to a weak difference on species composition within the same landscape. Moreover, forest cover decreasing play a selective effect on ant habitat-use guilds, keeping only the generalist that together with open-habitat specialist probably not play the same ecological roles of forest specialist ant species. Therefore, the reduction of forest cover, plays a strong simplification on ant assemblages, which became less speciose and dominated by generalist species.