基于主成分分析的黄土沟壑区土壤肥力质量评价——以长武县耕地土壤为例

Evaluation of soil quality on gully region of loess plateau based on principal component analysis

以陕西省长武县所处的黄土高原沟壑区土壤为研究对象,通过主成分分析并结合Norm值的方法,筛选出活性有机质、全氮、速效磷、速效钾、粘粒、CEC、过氧化氢酶、磷酸酶和转化酶等12项指标,建立了黄土沟壑区土壤综合质量评价的最小数据集(MDS)。采用客观方法(主成分分析法)以及主客观相结合的方法确定了两种类型的权重系数,分别计算了土壤综合质量指数,发现两种方法计算所得的土壤综合质量指数之间具有很好的一致性。另外,把由主成分分析法计算的土壤综合质量指数与利用偏相关系数法计算出的土壤综合质量指数进行了线性回归分析,两者之间达到极显著相关,说明用主成分分析法在该地区进行土壤质量综合评价是客观可行的。各因子的隶属度表明,活性有机质是该地区主要的限制性因子,全氮次之。几种不同方法计算的土壤质量表明,果园土壤质量优于农田土壤的,不同地形部位土壤质量比较发现,塬面的土壤质量相对最高,坡地(梯田)次之,河滩地相对最低。

In this paper,we made the Loess Plateau of Changwu County in Shaanxi Province as the research object.The principal component analysis（PCA） combined with Norm values were employed to select the indexes and establish the minimum data set（MDS）,in which 12 indexes such as labile organic matter,total nitrogen,available phosphorus,potassium,clay,CEC,catalase and phosphatase enzyme were selected as representatives for calculating soil quality index.Two kinds of weight coefficients were determined by using objective method from PCA and a combination of subjective and objective methods respectively,and soil quality index was calculated.The results showed that the soil quality index calculated by two kinds of weight coefficients had a good consistency.In addition,the soil quality index calculated by the PCA method was correlated to that calculated by the integrated partial correlation coefficient method,indicating that PCA is suitable for soil quality assessment in this region.Values of membership functions of the index revealed that labile organic matter is a major limited factor in this region,and the total nitrogen is the second.Soil quality index calculated by different methods all showed that soil quality of the orchard is better than that of farmland,and soil quality in different landform positions is in the order of table-land〉slope-land〉bottom-land.