基于最小数据集的黄土高原矿区复垦土壤质量评价

矿产开发影响了土壤质量,特别是在黄土高原生态脆弱地区。通过植被恢复能够改善矿区复垦土壤质量。为了揭示矿区复垦土壤质量在植被恢复过程中的变化,该文以黑岱沟矿区排土场不同恢复年限不同植被类型、未复垦地和周边自然植被恢复区为研究对象,选取21项理化生指标作为总数据集(totaldataset,TDS),运用主成分分析(principal component analysis,PCA)结合Norm值构建评价指标最小数据集(minimum data set,MDS),通过非线性(non-Linear,NL)和线性(linear,L)两种评价方法对研究区土壤质量进行了评价。研究表明:黄土高原北部典型矿区复垦土壤质量评价指标MDS包括粉粒百分比、有机质、速效磷、钠吸附比和微生物碳;2种评价方法下,植被恢复均对复垦土壤质量指数(soil quality index,SQI)有了显著提升(P 0.05),复垦20 a灌木SQI高于复垦10 a灌木SQI,复垦12 a草本SQI高于复垦20 a草本SQI,然而所有复垦土壤SQI均未达到自然植被恢复土壤SQI;由于非线性土壤质量评价方法(SQI-NL)具有更大的土壤质量指数变化区间和变异系数,此外,在SQI-NL和线性土壤质量(SQI-L)评价两种方法下,MDS和TDS之间决定系数分别为0.911和0.866,因此,非线性土壤质量评价方法在该区域具有更好的适用性,并且最小数据集能够较准确地进行土壤质量评价。

Soil quality evaluation for reclamation of mining area on Loess Plateau based on minimum data set

Mineral exploitation affects soil quality especially in the ecologically fragile areas the Loess Plateau. The quality of soil can be improved by vegetation restoration. In order to reveal the changes of soil quality in the process of vegetation restoration in the reclaimed land of mining, we took Heidaigou mining area and the surrounding natural vegetation restoration area as the research area, and selected 21 physicochemical indicators as the total data set (TDS). The minimum data set (MDS) was constructed by using principal component analysis (PCA) and Norm values. Two evaluation methods, non-linear (NL) and Linear (L), were used to evaluate the soil quality in this study area. The results show that the minimum data set include percentage of silt (Silt), organic matter (OM), available phosphorus (AP), sodium adsorption ratio (SAR) and microbial carbon (MBC). It indicates that in addition to conventional nutrients and physical properties, soil quality in this region is restricted by salinization and microbial biomass. The order of weight of MDS was: MBC (0.241) > OM (0.235) > AP (0.193) > Silt (0.174) > SAR (0.157). The applicability of the two evaluation methods is verified and we find that the overall spatial pattern of soil quality obtained by the two methods is similar, but there are still differences in details. According to the analysis results of the two evaluation methods, soil quality index (SQI) were all significantly improved by vegetation restoration (P < 0.05).Under different vegetation restoration types, the average value of the SQI-NL based on the MDS was sorted as follows: natural vegetation restoration land (NL, 0.587) > shrubs reclamation for 20 years (RLS 20, 0.479) > herbs reclamation for 12 years (RLH 12, 0.455) > shrubs reclamation for 10 years (RLS 10, 0.453) > herbs reclamation for 20 years (RLH 20, 0.424) > arbors reclamation for 20 years (RLA 20, 0.364) > and unreclaimed land (UL, 0.262) (P < 0.05), the average value of the SQI-L based on the MDS was sorted as follows: NL (0.522) > RLS 20 (0.444), > RLS 10 (0.412), > RLH 12 (0.401) > RLH 20 (0.400) > RLA 20 (0.326) > UL (0.228) (P < 0.05). In summary, different vegetation restoration years and different vegetation restoration types significantly improved the SQI of reclaimed land of mine areas (P < 0.05); however, the SQI of all reclaimed land has not reached that of natural vegetation restoration land, which indicates that the soil quality restoration of reclaimed land of mine areas is a long process. Since the SQI - NL method has a wider value range of SQI and coefficient of variation, this method has better applicability in this region than SQI-L, this is because 1) in the case of a wide range of soil quality index, the identification and classification accuracy of soil quality can be improved, which is conducive to the accurate evaluation of soil quality status and the timely detection of problems; 2) the coefficient of variation represents the sensitivity of SQI to changes in different environmental conditions, the higher the sensitivity of soil quality index to changes in environmental conditions is, the clearer the influence factors of soil quality change can be reflected, thus more effectively guiding the scientific management of soil quality improvement. Under the SQI-NL and SQI-L methods, the determination coefficients between MDS and TDS are 0.911 and 0.866, respectively, indicating that MDS can accurately replace TDS for soil quality evaluation. This study summarizes the MDS and soil quality evaluation methods suitable for the reclamation soil quality evaluation in the mining areas of the Loess Plateau, and recommends the future research and application in the same area and soil conditions.