Delineating soil nutrient management zones based on fuzzy clustering optimized by PSO

Delineation of soil nutrient management zones provides basis for variable fertilization technique and is the important link of variable fertilization management actualized in precision agriculture. After the spatial variability characteristics and structure of six soil nutrients were analyzed, they were taken as the variables to delineate soil nutrient management zones. The fuzzy clustering algorithm optimized by particle swarm optimization (PSO) was used to delineate management zones, and two indices were introduced to ascertain the reasonable number of management zones. According to the calculation, the reasonable number for the study area was 2. As single factor variance analysis was used to analyze soil nutrient data of the practical samples in each management zone, all soil nutrients but available phosphorus had great differences among management zones on the confidence level of 99%. Zone 1 had the higher soil fertility and zone 2 had the lower. The delineation result indicated that fuzzy clustering optimized by PSO had a good performance on delineating management zones and variable fertilization management was feasible in the study area.