Variabilidade espacial de colletotrichum truncatum em campo de soja sob três níveis de sanidade de sementes

Abstract

This study aimed to characterize the spatial variability of Colletotrichum truncatum (Schwein) incidence among harvested soybean seeds and to determine the best semivariogram model and method to represent C. truncatum incidence among these harvested seeds. The experiment was conducted under field conditions in 2009/10 crop cycle, in 3 plots of 9.9 x 10 m, using seeds inoculated with C. truncatum. The inoculum corresponded to 0.8, 1.6 and 2.4% of the total sown seeds. Three meshes were marked with GNSS receivers, totaling 112 points in each plot spaced at 1.3 m on the row. At the end of soybean cycle, seeds were harvested, dried and analyzed for their health according to the blotter test, considering the 336 demarked points. Four semivariogram models were adjusted to the collected data by using the methods ordinary least squares (OLS), weighted least squares (WLS), maximum likelihood (ML) and restricted maximum likelihood (REML). Cross-validation was used for the final selection of the semivariogram model and method. Then, kriging and kriging standard deviation were performed. The kriging maps illustrated the seed-to-seed transmission and its variance. Spatial dependence was found for C. truncatum transmission through the seeds. The best semivariogram model was spherical and the best fit was REML. There was a range of 0.95, 4.03 and 7.05 m for plots with 0.8, 1.6 and 2.4% inoculated seeds, respectively. The higher the primary inoculum of the plot, the higher the transmission to seeds near the inoculum source.