Modeling spatial distribution patterns to delineate irrigation and nutrient management zones for high-density olive orchards

Delineation of multivariate management zones is an effective approach to implement sustainable agriculture. Super-high density (SHD) is an effective growth strategy that is widely used in all olive-producing countries, particularly in Iran. The present field-scale study was carried out in the 120 hectare SHD planting system of ‘Arbequina’ olive orchards to provide a full picture of the management zones model automated in the GIS model builder. Geostatistical techniques were used to characterize the spatial variability patterns and distribution maps of three different groups of soil physical and chemical parameters and leaf nutrient contents. Principal component analysis (PCA) and weighted overlay analysis revealed a homogeneous management class for the above three groups. Finally, irrigation and nutrient management zones were determined by a fuzzy K-means (FKM) algorithm. Results showed that soil and leaf properties had a moderate to strong degree of spatial correlation; the spherical model, with a lower mean RSS and a higher R2, offered a good fit; also, Co-Kriging (CoK) produced better estimates than other methods. The variables showing the highest variances in the PCA were used as inputs for the FKM algorithm. The optimum number of irrigation and nutrient management zones was 4 and 5, respectively. The study area, in terms of percentage, was divided into MZ3 > MZ2 > MZ4 > MZ1 for irrigation management and MZ2 > MZ3 > MZ1 > MZ4 > MZ5 for the nutrient management zones. GIS model builder acted successfully under a large number of input layers. These findings, thus, suggest developing this workflow of the GIS model builder by the desired modifications in similar studies on the regional scale for other orchards or crops.