Avaliação da área de unidades amostrais para o cerrado: Aplicação prática em projetos de intervenção ambiental

Abstract

The Cerrado is the second largest Brazilian biome and has been exploited for various anthropogenic uses, such as agriculture, mining, urbanization, pasture, and forestry. In order to control this exploitation and mitigate its negative consequences, several federal, state, and municipal laws have been proposed, aiming at the environmental preservation of native areas or areas with restricted use in the country. To comply with this legislation, it is important to request the Environmental Intervention Authorization Document (DAIA), which necessitates the completion of quantitative and qualitative forest inventory. Most forest formations have a very large population, necessitating sampling. This process assesses the forest as a whole based on data collected from a fraction of it. Although the execution of forest inventory is mandated by law, there is no mention of the minimum size of sample units. Therefore, it is crucial to evaluate different sizes of sample units to obtain the required information in surveys that are part of a DAIA. Hence, this work aims to define the minimum size of rectangular sample units capable of estimating variables required by legislation to compose a DAIA. The study was conducted in the Cerrado Sensu Stricto physiognomy, in legal reserve areas of three farms located in the region of Monte Carmelo, Minas Gerais. Based on the characteristics of the areas, in defining the sampling design, we opted for the process of entirely random sampling and the fixed-area sampling method, with rectangular sample units initially with dimensions of 20 x 50 m (1,000 m²). Subsequently, similar sample units will be launched starting from the same initial vertex and with internal allocation, so that larger sample units encompass smaller ones. Since no size of sample unit was able to achieve sample sufficiency and the parameters of an environmental intervention project, it will be necessary to increase the number of plots for all sample sizes. However, as the sample of 800 m² indicated 32 plots as the ideal size, being the smallest result presented, and having statistically similar diversity indices to the other dimensions, it is suggested to launch 800 m² plots to complement the results of this study, in order to obtain sample sufficiency for all analyses, with lower sampling error and variance.