Relationship between Spatial Variability Pattern of Wheat Yield and Soil Properties

Abstract

Aims: Determining effects of spatial variation of some soil properties on wheat quantity and quality variation in order that proper soil and inputs management can be applied for sustainable wheat production.
 Study Design: Analyzing data of a field with center pivot irrigation system and uniform management using the geostatistical method.
 Place and Duration of Study: Soil and Water Research Department, Fars Agricultural and Natural Resources Research and Education Center, Darab, Iran, from September 2013 to February 2014.
 Methodology: Wheat yield data harvested by class lexion 510 combine from 25 m2 plots (11340 locations) with the corresponding geographical location were used. Besides, soil properties and wheat yield were measured at 36 randomly selected points on the field. Interpolation of parameters was predicted with the best semi-variogram model using kriging, inverse distance weighted (IDW), and cokriging methods.
 Results: Results showed that wheat yield varied from 2 to 10.08 tons per hectare. Cokriging with cofactor of kernel weight interpolator had more accuracy compared to the combine default interpolator (kriging). A logical, linear correlation was found between different parameters. The best variogram model for pH, OC, and ρb was exponential, for EC, TNV, SP, soil silt and clay percentage was spherical, and for soil, percentage sand was Gaussian model. Data of soil sand, silt, and clay percentage, EC, TNV, and SP had strong spatial structure, and soil pH, OC, and ρb had moderate spatial structure. The best interpolation method for soil pH, EC, sand and silt percentage was kriging method; while, for TNV, SP, OC, ρb, and clay percentage was IDW.
 Conclusion: There was a close relationship between wheat yield variation and changes in the soil properties. Soil properties and wheat yield distribution maps provided valuable information which could be used for wheat yield improvement in precision agriculture.