Resumen de Site-specific management zones based on geostatistical and fuzzy clustering approach in a coastal reclaimed area of abandoned salt pan
Wenquan Liu, Fang Lu, Guanquan Chen, Xingyong Xu, Hongjun Yu
Agricultural measures for cotton (Gossypium spp.) cultivation generally adopt a traditional and uniform management method in the reclamation area of the abandoned salt pan on the southern coast of Laizhou Bay, without considering spatial variability of the soil’s physical environment. This method might lead to overfertilization in some high-nutrient plots and poor fertilization in other low-nutrient plots, which will cause high heterogeneity of crop yields and low profits.
The objectives of this study were to delineate and to assess the suitability of the resulting site-specific management zones by describing spatial variability in soil attributes and crop yields. A 4.8 ha field of reclamation area in the coast was selected as the research object. A total of 12 soil attributes representing the physio-chemical attributes of soil (0-20 cm) and aboveground biomass (AGB) of cotton on 48 sampling locations were chosen from the study area. Principal component analysis and the fuzzy k-means clustering method were used to determine optimal clustering number of the study area and identify management zones. Results showed that the number of reasonable management zones were two, and confidence levels of soil organic matter (SOM) and AGB were significantly different at a probability level p < 0.01, while total N, available N, available K, cation exchange capacity, and soil salinity showed significant differences at p < 0.05. Management zones determined based on the selected soil attributes and crop AGB provide farmers with the opportunity to perform specific site management in cotton fields to achieve precision agriculture and provide a reference for formulation of a reclamation strategy for the salt pan.
