Optimization of crop rotations based on geostatistical analysis of soil acidity and earth remote sensing data, taking into account financial costs

Abstract

The article presents the possibilities of detailed analysis of the spatial distribution of soil acidity to reduce liming costs and optimise land use in five working areas of the experimental farm “Gutko S.”. Based on variogram analysis, the patterns of acidity distribution in key areas are determined. Regression analysis showed a significant and high polynomial dependence between the NDVI index and soil acidity (the correlation ratio is 0.60–0.75 at sites No. 2–4) and a significant direct linear relationship at site No. 1. Geostatistical analysis revealed an average spatial dependence (residual variance of 29.9%) at site No. 3. Based on the strong relationship between the average values of the NDVI index for the summer months over 3 years (9 images) and soil acidity, it was proposed to use NDVI as a predictor for optimising the sampling grid using stochastic modelling. It was found that the relationship with NDVI is more pronounced in areas where the relief is less fragmented. Based on calculations of costs for liming, the advantage of detailed acidity accounting over the classical methods of agrochemical survey used in the Republic of Belarus was substantiated. The profit per rotation was about 1,200 US dollars from an area of 184.5 ha. Based on the results of the analysis of acidity distribution, the NDVI index and field history, a more detailed scheme of elementary plots with crop rotations that take into account soil acidity was proposed. The limitations of the available agricultural equipment prevented a more detailed division of the plot.