No abstract

Nestedness is widely observed in natural metacommunities, but its underlying mechanisms are still poorly understood. The distribution of habitats in the landscape and differences in dispersal rates of distinct insect taxa can determine the nestedness of the metacommunity. Here, we evaluated how species habitat specialization contributes to metacommunity nestedness in insect groups with different dispersal capacities in a mountaintop landscape in south-eastern Brazil. We sampled ants, butterflies and dung beetles in two main habitats, naturally fragmented forest islands and a grassland matrix (campo rupestre), during both dry and rainy seasons. We classified species according to their degree of habitat specialization (generalists or specialists) based on the relative frequencies and abundances between these two contrasting habitats. Forty of 211 species were classified as habitat specialists, seven as habitat generalists. It was not possible to classify the remaining species. The metacommunity was nested in structure, with habitat generalist species contributing more to nestedness than habitat specialists. Nonetheless, habitat distribution in the landscape did not affect the nestedness of the metacommunity. Our findings reveal that species sorting (for habitat specialists) and mass effects (for habitat generalists) are concurrent processes in the mountaintop forest–grassland mosaic. Our study helps to advance our understanding of the differences in the distribution of generalist and specialist species in a tropical mountaintop landscape and improves our ability to predict and manage the increasingly adverse effects of changes in land use and climate on metacommunities and ecosystem functions.

Naturally fragmented landscapes are adequate systems for evaluating patterns and mechanisms that determine species distribution without confounding effects of anthropogenic fragmentation and habitat loss. We aimed to evaluate an ant metacommunity's spatiotemporal patterns in montane forest islands amid a grassland-dominated matrix. We assessed these patterns by deconstructing the ant metacommunity into forest-dependent and habitat generalist species. We sampled twice a year (summer and winter) over 2 years (2014 and 2015), using soil and arboreal pitfall traps, in fourteen forest islands (varying in size, shape, and connectivity) in the Espinhaço Range Biosphere Reserve, Brazil. We evaluated the relationship between ant species richness, composition (β-diversity), and predictor variables of forest island structure (canopy cover and understory density) and landscape structure (forest amount, number of forest islands, and shape). We sampled 99 ant species, 66.7% of which were classified as forest-dependent and 33.3% as habitat generalist species. We found that ant β-diversity was higher in space than in time, and that species composition variation in time (temporal β-diversity) differed between ant species groups. Both ant groups responded differently to forest island and landscape structure characteristics. Landscape structure seems to act as a spatial filter and the forest islands' local characteristics as an environmental filter, which jointly determine the local and regional diversity. We demonstrate the importance that forest archipelagos pose to ant metacommunity's structure and dynamics in montane tropical regions. Mountaintop conservation and management strategies must consider the forest island archipelago to maintain the biodiversity and the functioning of these systems.

Ants, an ecologically successful and numerically dominant group of animals, play key ecological roles as soil engineers, predators, nutrient recyclers, and regulators of plant growth and reproduction in most terrestrial ecosystems. Further, ants are widely used as bioindicators of the ecological impact of land use. We gathered information of ant species in the Atlantic Forest of South America. The ATLANTIC ANTS data set, which is part of the ATLANTIC SERIES data papers, is a compilation of ant records from collections (18,713 records), unpublished data (29,651 records), and published sources (106,910 records; 1,059 references), including papers, theses, dissertations, and book chapters published from 1886 to 2020. In total, the data set contains 153,818 ant records from 7,636 study locations in the Atlantic Forest, representing 10 subfamilies, 99 genera, 1,114 ant species identified with updated taxonomic certainty, and 2,235 morphospecies codes. Our data set reflects the heterogeneity in ant records, which include ants sampled at the beginning of the taxonomic history of myrmecology (the 19th and 20th centuries) and more recent ant surveys designed to address specific questions in ecology and biology. The data set can be used by researchers to develop strategies to deal with different macroecological and region‐wide questions, focusing on assemblages, species occurrences, and distribution patterns. Furthermore, the data can be used to assess the consequences of changes in land use in the Atlantic Forest on different ecological processes. No copyright restrictions apply to the use of this data set, but we request that authors cite this data paper when using these data in publications or teaching events.