We aimed to compare the soil ant diversity in different land use systems from Atlantic Forest area, in Southern Bahia state, Brazil. The ants were sampled in 16 sites: two primary forest sites (un-logged forest); three young secondary forests (<8 years old); three intermediate secondary forests (8-20 years old); three old secondary forests (>20 years old); three Eucalyptus grandis plantations (3-7 years old), and two introduced pastures. Each site was sampled in three sampling points 15 m apart, and distant over 50 m from the site edge. In each sampling point we gathered the litter from a 1 m² and extracted the ants with Winkler extractors during 48h. We found 103 ant species from 29 genera and eight subfamilies. The five richest genera were Pheidole (19 species), Solenopsis (8), Apterostigma (10), Hypoponera (7) e Paratrechina (5). The highest ant richness density was found in the primary forest (7.4 species/sample; S = 37; n = 5); followed by the old secondary forest (5.33 species/sample; S = 48; n = 9); young secondary forest (5.25 species/sample; S = 42, n = 8); eucalyptus plantation (4.22 species/sample; S = 38, n = 9), intermediate secondary forest (3.5 species/sample; S = 35, n = 10, and introduced pasture (2.67 species/sample; S = 16, n = 6). The ecosystems with higher structural complexity showed the highest ant richness density by sample. Therefore, in the Atlantic Forest region, the eucalyptus plantation is a better alternative of land use to conserve the ant biodiversity than pastures, and quite similar to native secondary forests in ant community characteristics.

As mudanças espaciais da composição de espécies estão ligadas a aspectos ambientais da variação dos habitats. As paisagens naturais de Minas Gerais são organizadas em mosaicos criados pela ação antrópica ou por fatores naturais. O estudo realizou-se com escarabeíneos na Chapada das Perdizes, município de Carrancas, MG. A região apresenta quatro fitofisionomias principais: campo de cerrado, campo limpo, floresta e campo rupestre. O objetivo do estudo foi detectar diferenças na distribuição dos escarabeíneos e avaliar a diversidade beta (dissimilaridade de espécies) entre as fitofisionomias. Os escarabeíneos foram coletados com armadilhas do tipo alçapão, com isca de baço de boi ou fezes humanas, em quatro áreas de cada fitofisionomia. Foram coletados 2.363 indivíduos de 52 espécies: 1.247 indivíduos de 29 espécies na floresta, sendo 20 exclusivas dessa fitofisionomia; 352 indivíduos de 17 espécies no campo limpo, com quatro espécies exclusivas; 386 indivíduos de 19 espécies no campo de cerrado, com cinco espécies exclusivas; e 378 indivíduos de 20 espécies no campo rupestre, com cinco espécies exclusivas. A armadilha com fezes foi a mais eficiente na captura e a floresta foi a fitofisionomia com maior diversidade e abundância de escarabeíneos. Isso sugere que a cobertura vegetal influenciou a distribuição da comunidade desses besouros. A diversidade beta alta mostra que as fitofisionomias são complementares entre si. Portanto, todas são importantes para a conservação, já que o conjunto de espécies que ocorre em cada ambiente é essencial para a manutenção da diversidade regional.

The spatial changes in species composition are closely related to the environmental aspects associated to habitat variation. The natural landscapes of Minas Gerais are organized in mosaics, associated to anthropic action or natural factors. Here we study dung beetles in the Perdizes Plateau, Carrancas (Minas Gerais, Southeastern Brazil). The area of study has a landscape composed mainly by four phytophysiognomies: Brazilian savannah, altitudinal grasslands, rupestrian fields and forests. The objective of our study was to detect the spatial distribution in composition of dung beetles and estimate beta diversity between phytophysiognomies. The dung beetles were collected in four areas of each phytophysiognomy, using pitfall traps baited with carrion or human feces. A total of 2,363 individuals of 52 species were collected: 1,247 individuals of 29 species at the forest, 20 exclusive to this phytophysiognomy; 352 individuals of 17 species at the grassland, with four exclusive species; 386 individuals of 19 species at the Brazilian savannah, with five exclusive species; and 378 individuals of 20 species at the rupestrian field, with five exclusive species. The traps baited with feces were the most efficient in capturing dung beetles, and the forest was the phytophysiognomy with the highest diversity and abundance. This pattern indicates that plant-covering influences the structure of the dung beetle community. High beta diversity indicates that phytophysiognomies are complementary. However, the other phytophysiognomies are also important for conservation, once species composition occurring at each environment is essential for the maintenance of the regional diversity.

As comunidades de Formicidae (Insecta, Hymenoptera) foram estudadas em fragmentos de floresta semidecídua inseridos no bioma Mata Atlântica medindo de 2,99 a 45,5 ha na região do Alto do Rio Grande, Minas Gerais, Brasil. As formigas foram coletadas em 15 amostras de serapilheira de 1 m² em cada fragmento, usando o método de extrator de Winkler. Cada amostra teve distância mínima de 50 m uma da outra. Um total de 142 espécies de formigas foi distribuído entre 40 gêneros, 23 tribos e 10 subfamílias. As comunidades amostradas mostraram uma alta riqueza e altos valores no índice de diversidade. O tamanho da área dos fragmentos não influenciou na riqueza de espécies de formigas.

The community of Formicidae (Insecta, Hymenoptera) in semideciduous seasonal forest of the Atlantic rain forest biome was studied in remnants ranging from 2.99 to 45.45 ha, in the region of the Alto do Rio Grande, Minas Gerais State, Brazil. The ants collected in 15 samples of litter of one square meter each, using Winkler extractors method, in each forest fragment. Each sample were taken at intervals of 50 meters. A total of 142 species of ants, were distributed into 40 genera, 23 tribes and 10 subfamilies. The communities showed a high richness and diversity indices values. The size of the fragments did not influence the ant richness.