Forest restoration is an urgent need, given the fastness of degradation processes in tropical rainforests. However, many efforts and resources are invested in the rehabilitation without appropriate methodologies for evaluation of these activities and monitoring of the restored areas. The necessity of monitoring and evaluation has become consensus among those involved in restoration, since it is an onerous activity to take the risk of failure. Most monitoring is based on analysis of vegetation, although, it is important to make use of other parameters. Due to the reciprocal influence between soil and vegetation, this study proposed to select indicators of soil quality, allowing the evaluation and monitoring of forest restoration projects in an efficient, practical and low cost way. Thus, the aim is to select good indicators, mainly edaphic, for assessment of forest restoration projects, and assess forest restoration plantings installed in borrow areas near by the Balbina Hydroelectric Dam (Presidente Figueiredo - AM). Four areas were selected for sampling: a degraded area (AC), two restoration plantings (RF1, RF2) and a forest area (AR). Because there is no replica of each condition the present work is a case study. In each area, eight plots of 5 m x 2 m were allocated, where we performed a survey of natural trees and shrubs regeneration with HT > 30 cm and DBH < 5 cm, and soil macro fauna, by "pitfall" and soil monolith methods. At the same plots, it was measured canopy cover, the soil infiltration rate and soil resistance to root penetration and soil samples were collected for determination of bulk density, moisture, texture, acidity and fertility (Al, Ca, Mg, K, SOM, N, P, Fe and Zn) in the depth: 0-5 cm, 5-10 cm and 10-20 cm. The physical-chemical parameters of soil and canopy cover were opposed to a categorically gradient of restoration by ANOVA and Tukey test at 5%. To oppose the continuous gradients of natural regeneration and macrofauna, were conducted simple regressions of the physical-chemical soil attributes and canopy cover with respect to the richness of these two biological groups. The restoration plantings have evolved structurally, proving that the restoration was a right decision, because the degraded area did not improve their features with the abandon. RF2 is on a more advanced stage than the RF1, presenting conditions closer to the forest. RF1, for be closer to the degraded area (AC), and need further silvicultural assistance. The parameters that follow the categorical gradient were soil density, penetration resistance, infiltration rate, potassium, soil organic matter, nitrogen and canopy cover. Those which responded to soil macro fauna and natural regeneration gradients were soil bulk density, soil organic matter, canopy cover, potassium and effective CEC, while penetration resistance and pH in water evidenced a weak relationship. According to this study, the good indicators to evaluate the success of forest restoration projects near by the Balbina Hydroelectric Dam are: bulk density, soil organic matter and canopy cover. The methodology proved effective for evaluating forest restoration projects; however, it needs to be tested elsewhere, ideally with a larger number of replicas to be validated.
