Local and landscape drivers of tropical forest regeneration in agricultural landscapes of the Atlantic Forest of Brazil

• Ano de defesa: 2018
• Autor(a) principal: Cesar, Ricardo Gomes
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Ecologia de paisagens; Florestas secundárias; Forest restoration; Landscape ecology; Natural regeneration; Regeneração natural; Restauração florestal; Second-growth forests
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/11/11150/tde-16072018-173459/

Forests established through native seedling planting (PL) and the establishment of secondary forests through natural regeneration (SF) are the main outcomes of large scale forest restoration. The decision making process of these approaches is conditioned by resilience. But the different outcomes of these approaches are as important as the decision making. SF are heterogeneous and - although there is a growing literature of the drivers of forest establishment - few works analyzed drivers of attributes of these recently established forests. In this context, our work aims to identify the differences between PL and SF and the local and landscape variables that affect SF attributes. To do so, we sampled the tree community in seasonal semideciduous forests of the Atlantic Forest established naturally (SF) and PL in agricultural landscapes in the Corumbataí Watershed, São Paulo State, Brazil. We observed that PL has similar biomass to SF and higher species richness. However, PL also showed lower abundance of young trees, animal-dispersed trees and lianas. Species composition between PL and SF also differs. SF established in abandoned eucalypt plantings showed species richness and biomass of native species similar to other SF forests. However, SF attributes vary greatly. In this context, SF show a large potential for providing food for fauna and storing carbon in a cost-efficient way. While PL can also provide these benefits, it may have its long-term sustainability compromised by the lack of regenerating trees. We then proceeded to investigate drivers of the heterogeneity observed in SF using generalized linear mixed models to estimate the effect of local and landscape variables on the biomass, species density and basal area of animal-dispersed trees of the SF sampled. SF surrounded by sugarcane plantations had lower biomass and basal area of animal-dispersed trees, while native forest cover in the landscape increased species density of SF. Forest age showed little or no importance in predicting SF attributes. These results highlight the importance of low impact agricultural practices and of strategies that increase native forest cover in agricultural landscapes, in order to increase the potential of SF to provide ecosystem services and conserve taxonomic diversity. The choice between establishing PL or fomenting SF for forest restoration is conditioned to the local and landscape context where restoration actions will be carried out. Although both approaches can potentially fulfill the objectives of restoration projects, special attention must be given to the recruitment of new individuals to maintain PL sustainability, while less impacting agricultural practices and more forested agricultural landscapes may increase the SF potential to provide ecosystem services and conserve biodiversity.