Assessing Local Uncertainty of Soil Protection in an Olive Grove Area with Pruning Residues Cover: A Geostatistical Cosimulation Approach

Abstract

The spatial characterization of soil protection is important for preventing land degradation. Mulch covers of pruning residues are increasingly used by farmers, but little is known regarding their spatial variability. Thus, field research was conducted in a 14·7-ha traditional olive grove located in Córdoba (Spain), to evaluate the variability of the initial density of pruning (Dpo), tree crown volume (V) and N content in residues (Np). Dry pruning weight and Np were characterized in 49 and 47 samples, respectively, covering approximately half of the area. V was measured in 178 locations covering all of the study area. The spatial continuity of these variables was described by spherical variograms. Geostatistical simulation algorithms were employed. One hundred realizations were made for Np through direct sequential simulation. Because the association with V was large (rxy = 0·72, p-value < 0·001), we ran a geostatistical cosimulation algorithm to obtain 100 maps of Dpo. The final density of pruning (FDp) was determined by applying a residue decomposition model for a 2-year period to pairs of previously simulated maps of Np and Dpo. After setting soil protection cut-off values for FDp, the soil protection probability was calculated for pruning strips of 2 and 4 m width. Finally, we converted the probability maps into binary maps and classified the farm into protected, unprotected and uncertain zones. V provided valuable input to evaluate soil protection. The application of stochastic simulations coupled with a decomposition model was useful to identify critical areas. This methodology can be easily applied to other crops. Copyright © 2017 John Wiley & Sons, Ltd