Espacialização da estimativa da produtividade em milho de segunda safra

Abstract

The corn productive potential is driven by genetic factors and inherent to the management of the crop, which may present spatial behavior in the crops. In this sense, yield estimates can be improved by taking into account the spatial variability of the stands. The objective of this work was to evaluate, in a spatial way, the contribution of the corn productive components in the crop yield estimation, as its relation with the real yield of the grain. The work was carried out at Fazenda Capim Branco, at the Federal University of Uberlândia-MG, in an area of 25 ha, cultivated on Red Latosol, with corn crop under no-tillage system, under rainfed. To collect the information, a sampling grid of two points per hectare (50 points) was defined, where a useful plot of 10m2 was demarcated in the center of each square. Real yield was determined by obtaining the grain weight of the ears contained in the useful plots. The estimated yield was determined from the evaluation of the crop productive components. The final stand was obtained for the useful plot. The number of ears, rows per ear, grains per row and the thousand seed weight (TSW), considered the removal of 5 plants in each of the 2 central rows of the useful plot. Data were analyzed descriptively, using Pearson's linear correlation and spatially, using geostatistics. The highest correlations were between the estimated yield and the TSW components (r=0.74) and number of grains per row (r=0.58), which was corroborated from the thematic maps for these attributes. Spatially, the TSW presented a similar range value when compared to the corn estimated yield. There is presence of spatial dependence for all attributes studied, with PMS having the greatest impact on the estimation of corn yield. There was an overestimation of corn yield in relation to real yield. The data, when evaluated in a spatial way, allow us to observe a similar pattern of behavior for both yield determinations.