Tamanho da área amostral e fitossociologia da vegetação arbórea regenerante em fragmento de floresta atlântica em Pernambuco, Brasil

Abstract

This study aims to evaluate the methodology of the relative efficiency of the sampling in the definition of the area and shape of plots in surveys of regenerating tree in an urban fragment protected from Dense Ombrophylous Forest of the Low Lands, in Jaboatão dos Guararapes, Pernambuco, Brazil. For the execution of the research, were installed 40 plots of 5m x 5m; 5m x 10m; 10m x 10m and 10m x 25m, including 20 plots in each forest environment (edge and interior). In all plots, we cataloged, identified and measured the circumferences at height of the base at 30 cm from the soil of all individuals whose height were equal or greater than 1.0 meter and circumference at breast height at 1.30 m of soil, less than 15.0 cm. The phytosociological parameters were estimated for each plot size, and after the selection of the form and the area or the ideal plot, the floristic composition, vertical and horizontal structures between the border and interior environments of the fragment were described. It can be observed that as the sample size was increased in the studied area was demanded a greater average and total time of inventory measurement in the field. Consequently, a greater number of individuals, species richness, and diversity, variations in the number of individuals and species among the sample units, a frequency of species and concentration of individuals in the initial classes of regeneration were obtained. However, a lower coefficient of variation, sampling error, abundance, uniformity of the number of individuals per species and species dominance. Regardless of the shape and area of the plots, we observed that levels of diversity, equability as well as estimates of abundance and dominance were similar to those already registered in the researches with the regeneration of tree species in the State Atlantic Forest of Pernambuco. The sampling of the tree regeneration in the Mata de Manassu considering the different plots sizes proposed indicated that the plot of 10m x 10m was the highest efficiency, given the high sampling precision, high species richness and low sampling cost. The estimators Jackknife of first-order indicated that regardless of the plot dimension adopted, the samplings presented an accuracy superior to 77% of the observed wealth versus estimated wealth. The results indicate a decrease in the characterization of species richness when adopting very small plots, or even when including admissible sampling error criteria of more than 10% in environments of Dense Ombrophylous Lowland Forest, and provides guidance on the planning and direction of future works to be developed with the natural regeneration of tree species in the Atlantic Forest biome in Pernambuco.