Study of the influence of soil use on the relationships between physical-chemical variables through network analysis

Abstract

The human being when using the soil gradually modifies its properties which can cause the loss or reduction of the health of the soil. The objective of this work was to analyse whether different land uses produce differences in the relationships of a set of 24 physical-chemical variables. The study area was located southeast of Madrid (Spain) in which 10 points were sampled where the three main soil uses (cereal, olive grove and natural) were adjacent. The relationships between the physical-chemical variables for each use were determined by Spearman's bivariate correlations (P<0.05). The study of these relationships was carried out using network analysis techniques. Soil use influenced the relationships between the variables analysed. Agricultural soils, especially cereal soils, had fewer and worse number of correlations. Agricultural soils had a greater tendency to form modules independent of each other, which would indicate a lower resilience of these soils. The most important physical-chemical variables changed between uses: related to organic matter in the network of natural use and related to different inorganic colloids in agricultural uses.