The search for sustainability in all activities and the increase in the stringency of environmental law culminate in the adoption of actions aimed at the restoration of degraded areas. The restored are a shave to be evaluated and monitored in order to confirm whether the aims proposed at the beginning of the project have been achieved. For this purpose, indicators for evaluation and monitoring are used, which portray the ecological characteristics of the restoration site. Therefore, this paper aimed to evaluate an area restored through planting seedlings, approximately 7 years ago, using natural regeneration and the soil seed bank as indicators for evaluation and monitoring. In the study area it was delimited an area of 0.5 hectare, in which 40 plots of 2x2 m were allocated, with an interspace of 10 meters between them, forming 8 strata consisting of 5 plots, parallel to the waterway, being the stratum 1 the nearest to the creek and number 8 the furthest one. In order to evaluate the regeneration layer, all woody shrub individuals in the plots with height ≥ 30 cm and CAP <15 cm were recorded. To evaluate the seed bank, a soil sample (0.23 x 0.36 x 0.05 m) was collected from the center of each plot. The regenerating individuals and the ones recruited from the seed bank were identified and classified according to successional class and seed dispersal syndrome, and by analyzing their phytosociological parameters. In the regenerating stratum it was observed 162 individuals, 13 families and 22 species, with a number of 10125 individuals/ha. Psidium guajava and Myrtaceae were the species and the family, respectively, with the highest VI. The diversity was low (H' = 2.691), as well as the ecological dominance (J' = 0.870). It was observed a higher percentage of successional class of pioneers at the species level (50%) and at the level of individuals (66%). Zoochory was the predominant seed dispersal syndrome among species (81.8%) and individuals (84.6%). In the soil seed bank there were recruited 7,519 individuals, 27 families and 82 species, totaling 2270 seeds/m2. The specie Melinis minutiflora and family Poaceae showed the highest VI. Herbaceous plants were the predominant form of life among the species (68.3%), and grasses among the individuals (75.4%). The diversity was considered low (H'= 1.89), as well as the ecological dominance (J'= 0.429). Among the woody species, the sucessional class of pioneers was the predominant among individuals (77%) and species (41%). Regarding seed dispersal syndrome, anemochory prevailed at the level of individuals (85%) and zoochory at the species level (50%). The watercourse in the area of study, as well as the riparian vegetation, did not influence significantly (p > 0.05) the regenerating layer and the soil seed bank’s compositions. The high number of M. minutiflora individuals has the potential to interfere negatively in the restoration, suggesting that management actions are adopted aimed at its control. The forest restoration presents floristic and phytosociological parameters similar to other areas in the restoration process, however, lower than those observed in mature semideciduous seasonal forests. Finally, natural regeneration and the soil seed bank have worked as good indicators, as it allowed us to evaluate the shortcomings of the restored area and to propose management actions to accelerate the restoration process.
