Indicadores preliminares para avaliação da restauração em reflorestamentos de ambientes ciliares

Abstract

The restoration success in forest ecosystems can be adequately assessed by correct selection of indicators that represent the achievement of established goals. The discriminant analysis technique on indicators selection consists of separation and classification of new observations on pre-defined groups, reducing the number of variables that are discriminant functions linearly dependent of the original variables. This study aims to define an index composed by structural attributes (number of species and individuals planted, height, basal area, number of regenerant species and individuals) and chemical and pedological soil attributes to classify riparian reforested environments regarding to restoration taking as reference reforestation around the the Volta Grande reservoir, Minas Gerais State, Brazil. Eleven variables were used for previous classification of plots in partially restored or unrestored groups and also used for discriminant analysis. Variables selected by the discriminant function generated were: number of species and basal area of planted individuals, number of regenerant species and individuals , litter accumulation and soil cation exchange capacity. Compatibility of 98% from previous plot classifications and after index formation, show the representativeness of the selected variables on evaluation of restoration of riparian reforestations.