Testing different monitoring protocols for restoration sites in highlands of southeast Atlantic Forest biome

Assessment and monitoring of restored forests are essential to correct and improve restoration techniques, especially in tropical ecosystems. Our objective was to monitor direct sowing restoration sites in highland Atlantic forest in southeast Brazil and compare different methods to assess recovery success. Monitoring protocols use different methodologies. The study sites were three rural areas located in the buffer zone of the Serra do Mar State Park, in the mountainous region of Cunha. We monitored direct seeding restoration sites using five different protocols (large plot, subplot, small plot, transect line and touch) after six months of planting and and until one year after that. We expected to observe a progressive restoration path in the one year observation, which we found, and that different sampling methods would produce different results and, on the contrary, we observed similarities in monitoring protocols in relation to the total richness and abundance of plant species. Despite this, we found that the large plot method was the best when compared to the others in terms of monitoring species abundance and that the touch method proved to be more efficient for monitoring species richness. Studies testing different monitoring methods are scarce and should be sought out, especially in the United Nations Decade for Ecosystem Restoration.