Edge effect evaluation in a forest fragment taking the tree diameter as indicator

Abstract

The objective of this paper was to determine whether differences exist in the tree structure between edge and interior of a fragment of Atlantic Forest in a rural area of Marechal Floriano – ES, with the hypothesis that there is predictability relationship between the diameter at breast height (DBH) of each individual with the distance to the edge. 187 trees were measured in 5 installments of 6 mx 6 m 3 m apart from each other. The DAP logaritimizados (LDAP) and distances were tested for normality by the Shapiro-Wilk test, Anderson-Darling and K ², showing no normal distribution. He became the Kruskal-Wallis test for differences between medians and rankings and Tuckey for nonparametric identification of the level of significance. To verify the correlation between LDAP and the distance to the edge we did a simple linear regression.Kruskal-Wallis test detected significant differences between the median LDAP between plots and nonparametric Tukey test showed no significant difference between plots 1 and 5. A simple linear regression showed no relationship between the LDAP predictability and distance to the edge. On the tiny fragments edge effect can enter the innermost portions, becoming entirely edge habitats. As the area has been cleared and goes through a regeneration process, with early successional species.