Diversity and function of tree species in human-modified Atlantic Forest landscapes

Agroforestry has been suggested as a sustainable strategy to promote biodiversity conservation in agricultural land, and farmers' knowledge can be crucial to design more diverse agroforestry systems. Therefore, the main objective of this study was to assess the conservation value of agroforestry systems in relation to forest areas in different stages of natural succession as well as farmers' knowledge. The conservation value of agroforestry systems (coffee and pastures) was assessed through tree diversity indices and compared with values found in forest areas using rarefaction curve. In parallel, we assessed the tree functions according to farmers' knowledge using an ethnobotanical survey. Species richness in agroforestry systems was similar to the forest in the early intermediate (40 to 50 years) of succession, whereas the diversity, based on Shannon and Simpson index, was similar to the forest in the pioneer and late pioneer of succession stage (5 to 25 and 5 to 10 years, respectively). According to farmers, the main functions performed by trees in both types of agroforestry system were shade for crops, fauna attraction and firewood. The functions food production and soil cover are restricted to coffee systems, while aesthetics and wooden stake are restricted to pasture systems. Moreover, we found a high correlation between the diversity of functions, based on farmers' knowledge, and the diversity of species found in the field. Our results suggest that agroforestry systems used by family farmers are important to maintain the diversity of trees at farm and landscape level. In addition, farmer`s knowledge on the functions of tree species is crucial for the design of highly diverse agroforestry systems.

Automated summary

Agroforestry systems are important for maintaining tree diversity at the farm and landscape level. The conservation value of these systems is similar to forest areas in different stages of natural succession. Farmers' knowledge on tree functions is crucial for the design of highly diverse agroforestry systems.