A Study of Soil Nutrients Characteristics Based on Principal Component and Cluster Analysis

The objective of the study was to analyze and determine soil characteristics and apply fertilizer rationally by the results of available nitrogen, available phosphorus, available potassium, soil pH, organic matter, trace elements content characteristics in Jilin Province. Principal component analysis and cluster analysis methods were applied to evaluate fertility characteristics in the soil research. Soil characteristics of available nitrogen, available phosphorus and available potassium and organic matter, pH value, soil trace element content index were explored based on 55 soil samples (0-30 cm). The results showed that the optimal partition number was 6 nutrient clusters: Ⅰ) The content of Slow-release potassium, organic matter, manganese and boron was very low； Ⅱ) The content of total phosphorus was low, the content of Fe and boron was high； Ⅲ) The content of organic matter was high； Ⅳ) The content of available potassium was high, the content of boron was low； Ⅴ) The content of available potassium and Fe was low； Ⅵ) The content of available potassium, total phosphorus, pH were high, the content of manganese was low. The ratios of these 6 various clusters were 40%， 5%， 24%， 13%， 13% and 5%. In order to simplify the fertilization process, the soil categories could be divided into fewer classes.