Biodiversity conservation in fragmented landscapes at the Atlantic Plateau of São Paulo (Brazil)

Abstract

Habitat fragmentation is presently one of the most serious ecological problems at present. In the tropics, a huge number of species is being lost every day even before becoming known by science. Since fragmented habitats will be the pattern in the future, environmental management actions must be taken to avoid the ruin of biological diversity and all benefits derived of it. This projects intends to study some vital effects of fragmentation in a landscape formed by secondary tropical forest fragments of the Atlantic Forest biome (23035'S, 23050'S; and 46°45'W, 47015'W) and to provide ecological basis for regional management, in order to promote the persistence of the highest number of species in the landscape. In this context, the main objective of this project is to verify what spatial arrangement and which landscape elements are needed to maintain the biological diversity, by: i) describing the history of fragmentation and regeneration of the landscape; ii) relating fragment size and matrix type to species diversity of some taxonomic groups, presence abundance of forest (meta)populations, according to their sensitivity to fragmentation; iv) investigating the ecological processes which determine the main tenance of (meta)populations, particularly those related with the influence of the surrounding matrix and forest connectivity. These objectives match the request of the BIOTA FAPESP Program. The project design includes eight sub-projects, dealing with the same landscapes and groups of forest fragments, with complementary objectives. We propose an approach that involves a multiple scale analysis, where either the whole landscape and the fragments will be focused 00, detailing on the matrix and corridor effects; studies will be carried out at the population and community levels, where a wide range of taxa will be examined (woody plants, primates, birds); umbrella-species and key-processes which may maintain high biodiversity will be the sought. Different theories on landscape ecology will be tested and parameters of the spatial landscape structure and of habitat quality will be integrated to metapopulation models, particularly the incidence function model. To understand the present landscape structure and its history of fragmentation and regeneration, two types of landscapes will be distinguished: one formed by a forest matrix, and another by an agricultural matrix. They will be characterized considering: i) landscape connectivity and heterogeneity, forest fragmentation and isolation (sub-project 1); ii) geomorphological and pedological features, and main physical geo-indicators for measuring the magnitude and velocity of key abiotic processes (sub-project 2). Each fragment will be characterized by: i) size; ii) heterogeneity of the forest eco-units; iii) internal environmental quality and heterogeneity of the eco-units; iv) shape and edge effects; v) border complexity; vi) degree of isolation and connectivity, and the potential for (re)colonization; vii) age and history of regeneration, as obtained by the analysis of aerial photographs from1962, 1973 and 1988 (sub-projects 1 and 3). To relate biodiversity patterns with forest fragment size and matrix type, twelve fragments will be studied in detail, concerning: i) species composition and diversity of adult trees (sub-project 3); ii) species composition and diversity of tree seedlings and saplings (sub-project 7); iii) the abundance of six bird species (sub-project 6); iv) the biomass and density of primates (sub-project 5). The influence of matrix and corridors on seed flows (sub-project 8), forest regeneration (sub-project 7), movement of selected bird species (sub-project 6) will be analyzed in more detail, dealing with a reduced number of fragments. Fragment size and connectivity will also be related with patch-occupancy patterns potentials for species (re)colonization, By working with high- and low-demanding landscape, as well as the landscape types which support only low demanding species. Here, 50 to 60 fragments will be considered, from 4-5 ha to 90 ha, being around half of with agricultural matrix. Initially, the following organisms will have their patch-occupancy patterns considered: i) six bird species (sub project 6); ii) one endemic palm-tree species (sub-project 4); iii) several tree species (sub-project 3). The comparison of these two groups of fragments in terms of size, environmental quality (sub-project 3) and spatial attributes (isolation, connectivity, matrix type) (sub-project 1) must indicate the factors which dictate the permanence of a given species in the landscape. Incidence function models will be applied to analyse patch-occupancy patterns and simulate the species occurrence in different scenarios of the landscape evolution. This approach will allow us to identify the best fragment conditions for maintaining the species in the landscape in a long term basis, assuming they behave as metapopulation, providing direct ecological basis for the conservation management of tropical forest fragments. One of the main results of the project proposed will be the generation of a data base with spatialized information on the regional physiography, structure and dynamics of the regional flora and landscape, besides publications in specialized journals, diffusion articles, multimidia material, such as CD-ROM and on-line data base, and a guide for the regional woody species, directed to non-specialists, particularly children of public regional schools. (AU)