Uma nova abordagem para definição da suficiência amostral em fragmentos florestais nativos

Abstract

The sampling sufficiency is an important point to guarantee the knowledge of the variable of interest in a population. However, a common problem is the subjectivity of the methods used to determine the sampling sufficiency, leading to equivocated interpretations of the studied physiognomy. Thus, the objective of this study was to determine the sampling sufficiency applying a new approach combining the plateau response linear regression with the central limit theorem and to compare it with the traditional adjustment method, in 20 native fragments along the São Francisco River Watershed - MG. The traditional adjustment method presented some dubious results such as a false plateau point in fragments: 22, 27, 77, 100 and 112. Based on the new approach the plateau did not occur and clearly the plateau point left no doubt about the sampling sufficiency.