The vegetation of the upper part of the Mayo river watershed varies from wet grassy scrubland (≤ 3,365 m.a.s.l.) to alluvial forests (≥ 872 m.a.s.l.) which share characteristics with those in the amazon floodplain. These ecosystems have been exposed to high pressures linked to the expansion of monocultures, such as maize, rice and coffee by migrant farmers from the Andean region, which have resulted in the loss of biodiversity and ecosystem services. The substitution for rice of communities of Virola surinamensis and Hura crepitans associated to Mauritia flexuosa, Ficus trigona and Coussapoa trinervia has caused the fragmentation and loss of forests. Consequently, there is a need to establish models of community-based restoration to recover landscape connectivity through agroforestry systems.
The proposed model included a characterization of the composition and structure of the alluvial forest as reference ecosystem, recording richness, abundance, DBH, total height, trunk height, canopy height and cover. Diversity values were obtained using descriptive indexes of Shannon-Wiener, Simpson, Equitability and Fisher-α.
Community-based restoration ensures participation in the identification of plots and seed trees, phenological tracking, nursery construction, seedlings production, and restoration activities. The project includes capacity building on orchard management and native beekeeping to complement the installation of agroforestry systems.
The species proposed for the restoration were selected based on density analysis, considering those with a greater resilience and regeneration potential. A total of 346 species were recorded, grouped in 219 genera and 75 families. Magnoliopsida was the class with the highest number of species (94,80%). Leguminosae (16,05%) and Rubiaceae (9,46%) were the most important families. The Simpson dominance index shows that 90% of the species are dominant in primeval forest and 89% in secondary forest. The values of Shannon index are low in comparison with other tropical forest, which surpass 2,5 bits/ind.
Based on this characterization we prescribed Assisted Natural Regeneration (ARN) for the ecological restoration in seasonally flooded alluvial forest, enriching the plots with totally or partially esciophytic species, mainly in the areas with advanced regeneration, and with heliophytic species in open areas. The functionality of ecological restoration will be assessed based on studies of the use of the ecosystem under restoration by key seed dispersal species, linking the results with the design of connectivity networks.