Ethnobotany and ecological perspectives on the management and use of plant species for a traditional fishing trap, southern coast of São Paulo, Brazil.

The cerco-fixo is an artisanal fishing trap widely used by traditional communities in the estuarine region of the southern coast of the state of São Paulo, Brazil. The primary goal of the study was to investigate, through ethnobotanical and ecological approaches, the use of plant species by traditional fishermen to build the cerco-fixo at Cardoso Island State Park and Cananéia Island. Ethnobotanical data were collected through interviews, direct observation, plant collection and identification, and document analysis. An ecological evaluation was also done comparing five 20 × 20 m plots in a managed area to five 20 × 20 m plots in an unmanaged area, both within arboreal sandy soil vegetation called restinga arbórea, found within the Brazilian Atlantic Forest domain. This study involved 34 fishermen living at Cardoso and Cananéia Islands. The fishermen know more than 90 Atlantic Forest plant species that can be used to build the cerco-fixo. Tree species from the family Myrtaceae were the most quoted in the interviews. With respect to the ecological evaluation, the cluster analyses showed greater heterogeneity in terms of floristic composition (i.e. greater floristic dissimilarity) within the plots of the managed area. The analyses of diversity showed a slightly higher species richness and slightly lower values for Shannon, Simpson, Hurlbert's PIE and Evenness indices in the managed area (59 species; H' = 3.28; 1/D = 10.77; E = 0.80; Hurlbert's PIE = 0.91) compared to the unmanaged area (54 species; H' = 3.39; 1/D = 20.21; E = 0.85; Hurlbert's PIE = 0.95). The Hutcheson's t test showed no significant difference between both areas' Shannon diversity indices (t: -1.04; p: 0.30). These results are attributed to the greater dominance of the palm species Euterpe edulis Mart. in the managed area (28.2% of the trees sampled at this area; n = 118), which equals twice the percentage of individuals of the same species found for the unmanaged area (14.6% of the sampled trees; n = 48). We discuss the impact of the fishermen's harvesting practices in the managed area with an emphasis on three main points: (1) the harvesting practices are likely not contributing to a decrease in diversity in the managed area; (2) the greater heterogeneity in terms of floristic composition found for the managed area may reflect a mosaic pattern created by the opening of small tree-gaps distributed across this area over the course of more than 50 years; (3) the disturbance promoted by the fishermen's harvesting practices can be compared to natural disturbances of low impact that create mosaic patterns in tropical forests. This study emphasizes the prominence of the human dimension in ecological processes and the importance of considering the perspectives of local people when discussing the conservation of the natural environments in which these people live.