Interpolators in predicting the estimated population density of Oebalus spp.

Abstract

The objective of this study was to compare interpolation methods of ordinary kriging and inverse distance weighted applied to the spatial distribution of Oebalus poecilus and Oebalus ypsinolgriseus population densities in irrigated rice. In two crops, it was generated a grid of 30 x 30 m sampling, termed Area 1 and Area 2 with 39 and 192 sampling units (1 m2, corresponding to 200 rice plants), respectively. Seven evaluations were carried out from sowing to harvesting. In these areas, O. poecilus and O. ypsilongriseus adults were quantified and the sum of each species was used for analysis. Values were submitted to the ordinary kriging interpolation and inverse distance weighted for different semivariogram models, in which the cross-validation technique was used for defining the best interpolator. The ordinary kriging interpolation method showed a better performance than the method of inverse distance weighted applied in the spatial distribution of population density of O. poecilus in rice cultivation. O. ypsilongriseus could not generate estimates for comparison.