Georreferenced statistics: approach and practical application in Brazilian animal science
Mariah Tenório de Carvalho Souza and Enmelly Rayane Azevedo da Rocha
ABSTRACT The objective of this review was to improve knowledge about data geoprocessing as well as an application of georeferenced statistics for use of animal science. The advancement of knowledge in the various areas that make up the agricultural sciences has evidenced the limitations of traditional methods of statistics, in the practical application of spatial variability and in the analysis of variables of the soil-plant-water - atmosphere system in large areas territories. The national and international literature reports several methods of forage mass determination, however, the greatest difficulty is found in accurate methods for determining the availability of forage in critical or drought periods in Caatinga areas of Brazilian semiarid regions. The methodology proposed by geostatistics attempts to extract, even with randomness of the data collected, the structural and probabilistic characteristics of the regionalized phenomenon. The interpolation method is called kriging and is based on the sample data the regionalized variable and the structural properties of the semivariogram obtained from these data. Isoline maps have level curves representing a set of regionalized points and may represent an isovalue. The studies with the litter of dead material on the soil (burlap), cacti and cycads are already represented in areas of Caatinga, however, greater importance should be given to the practical application of georeferenced statistics and the evaluation of spatial distribution applied to animal science for a better understanding of environmental characterization.