Geostatistical study of soil pH using different sampling approaches

Abstract

Detailed studies of soil agrochemical properties is an important task to highlight the diversity of the area. Our study object is located in a weakly rolling relief area. Three different cartographic methods to determine soil pH were studied: 1) combined soil samples were taken following prevailing soil variety boundaries, consulting the available previous research material; 2) combined soil samples were taken accounting for the prevailing boundary of soil variety; 3) combined soil samples were taken following the regular grid. In order to clarify the optimal size of the plot single combined soil sample was taken from 2, 4 and 8 ha plots. Similar values of soil pH were achieved in a weakly rolling relief in context of distribution of soil acidity classes using different methods of soil sampling, except taking the samples for 2 ha plots on a regular grid, which outputs the shares of medium acid soils (pH 4.6-5.0) twice less than the other two methods. If 4 ha plots were used, the regular grid sampling resulted in a three times less share of relatively acid soils (pH<5.5) than mapped using other two methods. Characteristics of experimental variograms and Moran's I statistic were used to test the spatial continuity of soil pH. The main findings was that weak or moderate spatial dependence between the pH values was observed only for sampling schemes based on 2 ha plot size.