Soil attributes and weed seedbank spatial correlation

Abstract

The crop yield potential is affected by the crop-weed competition and their control create a dependence on herbicide use who brings, as consequence, soil impacts. Knowing the weed’s spatial distribution on the field is a feasible alternative for improving the crop yield. The goal of this paper is the identification of the spatial variability on physical and chemical attributes of soil as well as the weed’s seedbank so that, when correlated, may find standards to help on field management. The experiment was conducted on Uberlandia Federal University premises at soybean no-till area. Using georeferenced soil samples, were analyzed the physical and chemical attributes as well as the weed’s seedbank. The weed population on controlled environment was quantify, sorting out broadleaf and grassy weeds species. The obtained data were analyzed by descriptive statistic and geostatistics for a semivariogram modeling, interpolation by the kriging methodology and the spatial variability maps achievement. The average value, coefficient of variation (CV%), asymmetry, kurtosis coefficient and the significant linear correlations interfered on data spatial variability which we concluded by the spatial dependences on the attributes that had a linear correlation between them. The semivariograms presented varied range between 202 to 752 meters. Using the maps, verified two different regions for the broadleaf and grassy weeds seedbank. For both situations there was influence by the soil attributes on infestation level, which makes it possible to target the herbicide management reducing costs and the environmental impact. From the analyzed data we conclude that there is a spatial dependence for the physical and chemical soil attributes and their spatial distribution explains the weed seedbank spatial variability.

Keywords

• precision agriculture.
• soil fertility.
• spatial variability.