No abstract

In many tropical areas of the world, pristine forests have become rare. Nevertheless, due to shifts in the human population the area covered by secondary forests is increasing. These forests may harbor a rich flora and fauna and are considered to be main refuges for species of primary forests. However, this issue is far from clear. To assess the conservation value of secondary forests in the Atlantic Forest of Brazil, we compared the diversity of spiders in differently aged secondary forests with old-growth forests. Within a larger project treating several invertebrate taxa, we sampled spiders using a standard protocol in 24 sites of three successional stages (5-8, 15-20, 30-50 years old) and old-growth forests (>100 years untouched) in two nature reserves. We describe the diversity and structure of the assemblages using morphospecies and genera and analyze richness at the genus level. Generic richness and diversity showed no differences between successional stages; i.e., did not increase from the youngest to older forests, but guild diversity did increase. The youngest stage showed the highest variability in generic composition, and the turnover of genera and species was strong between the younger forests (5-20 years old) and forests older than 30 years. High alpha diversity, high turnover among sites and the lack of differences in richness between stages support the value of secondary forests for species conservation in the region studied.

No abstract

In the Atlantic Forest in Brazil, where no more primary forests exist, the value of secondary forests for biodiversity conservation is becoming more and more important. We studied the spiders in a relatively wellpreserved region of the Mata Atlântica, where the matrix of the landscape is still forest. We addressed the contribution of different spatial levels including forest stages to total diversity and analyzed the patterns by additive partitioning of beta diversity on genus and morphospecies level and for different sampling methods. Beta diversity was strongly based on turnover, not on gain/loss. All spatial levels (sample, stage, area, locality) contributed more to beta diversity than expected, without stronger influence of stage. Patterns were consistent for both identification levels and all methods. We conclude that in this landscape the protection of large areas encompassing all forest stages, without special attention to old-growth, is the best way to conserve the regional species richness.

The data presented here have been collected in the southern part of the Atlantic Forest (Mata Atlântica) in the state of Paraná, Brazil within a bilateral scientific project (SOLOBIOMA). The project aimed to assess the quality of secondary forests of different regeneration stages in comparison with old-growth forests with regard to diversity of soil animals and related functions. The Atlantic Forest is a hotspot of biological diversity with an exceptionally high degree of endemic species, extending over a range of 3,500 km along the coast of Brazil. The anthropogenic pressure in the region is very high with three of the biggest cities of Brazil (São Paulo, Rio de Janeiro, and Curitiba) lying in its extension. An evaluation of the value of secondary forests for biodiversity conservation is becoming more and more important due to the complete disappearance of primary forests. In 2005, we sampled spiders in 12 sites of three successional stages (5-8, 10-15, 35-50 yr old, three replicates of each forest stage) and old-growth forests (> 100 yr untouched, also three replicates). All sites were inside a private nature reserve (Rio Cachoeira Nature Reserve). We repeated the sampling design and procedure in 2007 in a second private reserve (Itaqui Nature Reserve). The two nature reserves are within about 25 km of each other within a well preserved region of the Mata Atlântica, where the matrix of the landscape mosaic is still forest. A widely accepted standard protocol was used in a replicated sampling design to apply statistical analyses to the resulting data set and allow for comparison with other studies in Brazil. Spiders were sorted to family level and counted; the adult spiders further identified to species if possible or classified as morphospecies with the help of several spider specialists.