Livestock grazing impact differently on the functional diversity of dung beetles depending on the regional context in subtropical forests.

The replacement of native forest by cattle pastures reduces functional diversity; however, little is known about whether the changes depend on regional variation. Dung beetles are one of the most diverse and functionally important taxa; through organic matter burial, dung beetles improve soil quality. We collected dung beetles in native forests and cattle ranching areas in subtropical forests with contrasting climatic conditions: the Atlantic Forest, the Humid Chaco, and the Dry Chaco. We measured 11 traits related to the ecology and the physiology of species. Irrespectively of the region, functional richness was higher in forests (native and with cattle) when compared to open pastures. Humid forests (Atlantic Forest and Humid Chaco) showed higher functional richness than Dry Chaco. Functional dispersion in humid forests was similar between native forest and livestock systems, however, functional dispersion in the Dry Chaco was higher in open pastures compared to native forest. According to our results, native forests and forests with cattle maintain functional diversity in all regions. However, in the case of open pastures, the response depends on the regional context; the replacement of native forest by open pastures strongly affected functional diversity in humid forests and showed less impact on dry forest.

Changes in taxonomic and functional diversity of plants in a chronosequence of Eucalyptus grandis plantations.

Tree plantations have become one of the fastest-growing land uses and their impact on biodiversity was evaluated mainly at the taxonomic level. The aim of this study was to analyze environmental changes after the Eucalyptus plantation in an area originally covered by natural grasslands, taking into account the alpha and beta (taxonomic and functional) diversity of plant communities. We selected nine plantation ages, along a 12 years chronosequence, with three replicates per age and three protected grasslands as the original situation. At each replicate, we established three plots to measure plant species cover, diversity and environmental variables. Results showed that species richness, and all diversity indices, significantly declined with increasing plantation age. Canopy cover, soil pH, and leaf litter were the environmental drivers that drove the decrease in taxonomic and functional diversity of plants through the forest chronosequence. Based on the path analyses results, canopy cover had an indirect effect on plant functional diversity, mediated by leaf litter depth, soil pH, and plant species richness. The high dispersal potential, annual, barochorous, and zoochorous plant species were the functional traits more affected by the eucalypt plantations. We recommend two management practices: reducing forest densities to allow higher light input to the understory and, due to the fact that leaf litter was negatively associated with all diversity facets, we recommend reducing their accumulation or generate heterogeneity in its distribution to enhance biodiversity.

Different land-use types equally impoverish but differentially preserve grassland species and functional traits of spider assemblages.

Land-use change is one of the major drivers of biodiversity loss by introducing environmental modifications, which excludes native species unable to adapt to the novel conditions. Grasslands are among the most threatened biomes; understanding the influence of different land-use types on native species is crucial to achieving sustainable management policies. We hypothesized that land-use types that partially conserve the original vegetation cover would show higher taxonomic and functional diversity and similarity with native assemblages than land-use types that replace the original vegetation cover. We compared the taxonomic and functional alpha and beta diversity of spider assemblages between soybean crops, eucalypt plantations, and cattle fields with seminatural grasslands. Through null models, we assessed the standardized effect sizes to test differences in the strength of environmental filtering among land-use types. Environmental changes introduced by different land-use types resulted in assemblages differentiated in species and trait composition, taxonomically and functionally impoverished with respect to seminatural grasslands. All land-use types drove species replacement and trait loss and replacement of grassland spiders. Each land-use showed a characteristic species and trait composition. Most of the grassland traits were not lost but were under or over-represented according to the land-use type. Only in soybean crops the formation of spider communities would be mainly driven by environmental filtering. Changes in land-use decreased species diversity and modified the composition of spider species and functional traits leading to differentiated spider assemblages. As spider species and traits varied among land-uses, a mitigation measure against grasslands biodiversity loss could be the development of productive landscapes with a mosaic of land-use types, as each of them would provide microhabitats for species with different requirements. Because land-use types mainly led to the rearrangement of grassland functional trait values, most of spider functions might be conserved in mosaics of land-use types.

Dung beetles response to livestock management in three different regional contexts.

The response of biological communities to human disturbances depends on factors acting at local and regional scale and on the interaction between them. We compared the response of native forest dung beetle communities to cattle grazing under regional contexts differing on precipitation patterns (Atlantic forest and humid and dry Chaco). Through multivariate and GLMM analyses we contrasted richness and composition across regions and land uses and explored the role of local and regional variables accounting for those changes. We captured a total of 44101 individuals of 109 species. The interaction between local and regional variables influenced the response to livestock management. In the two wet regions (humid Chaco and Atlantic forest) diversity was similar in the native forest regardless of cattle presence but differs strongly in open pastures. In contrast, in the dry Chaco, differences between native forest and land use were not evident. Vegetation structure was a major determinant of species richness, whereas regional climate determined differences in species composition. We concluded that the response of dung beetles to livestock management cannot be generalized for all biomes. In dry ecosystems, dung beetles are probably pre-adapted to environmental conditions imposed by cattle ranching whereas in wet ecosystems the impact of cattle ranching is more significant.

Species-fragmented area relationship.

The species-area relationship (SAR) gives a quantitative description of the increasing number of species in a community with increasing area of habitat. In conservation, SARs have been used to predict the number of extinctions when the area of habitat is reduced. Such predictions are most needed for landscapes rather than for individual habitat fragments, but SAR-based predictions of extinctions for landscapes with highly fragmented habitat are likely to be biased because SAR assumes contiguous habitat. In reality, habitat loss is typically accompanied by habitat fragmentation. To quantify the effect of fragmentation in addition to the effect of habitat loss on the number of species, we extend the power-law SAR to the species-fragmented area relationship. This model unites the single-species metapopulation theory with the multispecies SAR for communities. We demonstrate with a realistic simulation model and with empirical data for forest-inhabiting subtropical birds that the species-fragmented area relationship gives a far superior prediction than SAR of the number of species in fragmented landscapes. The results demonstrate that for communities of species that are not well adapted to live in fragmented landscapes, the conventional SAR underestimates the number of extinctions for landscapes in which little habitat remains and it is highly fragmented.