1. Ecosystem restoration is a global priority, currently promoted by several ambitious commitments. Most of the research to guide restoration practices was so far put on the recovery of above-ground functions and services, although soil health is increasingly recognized as a fundamental condition to restoration success. Soil restoration is particularly needed in mining areas, in which the surface soil layers are removed and the left over, highly deprived substrate does not support vegetation recovery.2. Here, we evaluated the potential of active restoration to recover soil attributes after bauxite mining in the Brazil's Atlantic Forest. We specifically studied how physical, chemical and microbiological soil attributes down to 40 cm depth respond to restoration (i.e. topsoil spreading and tree planting) by comparing newly mined areas, areas undergoing restoration for ~7 years and conserved forests. We further explored the associations between microbiological and physico-chemical soil attributes.3. Restoration interventions resulted in a 10-fold increase in the soil organic matter content compared to mined areas, as well as in the recovery of P content.Restoration was also shown to be associated with the recovery of the activity and diversity of soil bacterial communities, reaching similar levels to those observed in conserved forest. Restored and conserved forests were characterized by a higher proportion of Acidobacteria, Planctomycetes and Verrucomicrobia.All bacterial attributes were correlated with soil physical attributes, except for microporosity, and with organic matter and cation exchange capacity.

Soil phosphorus (P) is a major driver of forest development and a critically limited nutrient in tropical soils, especially when topsoil is removed by mining. This nutrient can be present in soils in the form of different fractions, which have direct consequences for P availability to plants and, consequently, for restoration success. Therefore, understanding how the stocks of different soil P fractions change over the restoration process can be essential for guiding restoration interventions, monitoring, and adaptive management. Here, we investigated the recovery of soil P fractions by forest restoration interventions on bauxite mine sites in the Brazilian Atlantic Forest. We assessed the concentration of different fractions of soil organic and inorganic P at (1) a bauxite mine prepared for restoration; (2) two former bauxite mines undergoing forest restoration for 6 and 24 years; and (3) an old-growth forest remnant. Overall, restored areas recovered levels of labile organic P (P o-NaHCO 3) at 5-40 cm and of moderately labile organic P (P o-NaOH) at different depths, exhibiting concentrations similar to those found in a conserved forest. The use of P-rich fertilizers and forest topsoil may have greatly contributed to this outcome. Some other fractions, however, recovered only after 24 years of restoration. Other inorganic P fractions did not differ among mined, restored, and conserved sites: nonlabile P i (residual P and P-HCl), labile P i (P i-NaHCO 3), and moderately labile P i (P i-NaOH). Forest restoration was able to promote efficient recovery of important soil P fractions, highlighting the value of restoration efforts to mitigate soil degradation by mining.

No abstract