Dai Liang-Liang, Xu Hong-Gen, Gong Hao, Peng Zhi-Gang, Xiao Kai-Qi, Wu Huan-Huan, Xu Qing-Yang, Guo Jun, Tang Yuan-Yuan, Zhang Jun
Changsha Natural Resources Comprehensive Survey Center of China Geological Survey, Changsha 410600, China.
Xi'an Mineral Resources Survey Center of China Geological Survey, Xi'an 710000, China.
Huan Jing Ke Xue. 2023 Apr 8;44(4):2243-2251. doi: 10.13227/j.hjkx.202205248.
The naturally high background value region of Cd derived from the weathering of carbonate has received wide attention. Due to the significant difference in soil physicochemical properties, soil Cd content, and bioavailability of different parent materials in the karst area, there are certain limitations in using the total soil Cd content to classify the environmental quality of cultivated land. In this study, surface soil and maize samples of eluvium and alluvial parent material in typical karst areas were collected systematically; the contents of maize Cd, soil Cd, pH, and oxides were analyzed, the Cd geochemical characteristics of different parent soils and the influencing factors of their bioavailability were revealed, and scientific and effective arable land use zoning suggestions based on the prediction model were suggested. The results showed that the physicochemical properties of different parent material soils in the karst area were obviously different. The alluvial parent material soil had low Cd content but high bioavailability, and the maize Cd exceeding rate was high. The maize Cd bioaccumulation factor was significantly negatively correlated with soil CaO, pH, Mn, and TC, and the correlation coefficients were -0.385, -0.620, -0.484, and -0.384, respectively. Compared with the multiple linear regression prediction model, using the random forest model to predict the maize Cd enrichment coefficient had higher accuracy and precision. Furthermore, a new scheme for the safe utilization of cultivated land at the plot scale based on soil Cd and predicted crop Cd content was proposed in this study, making full use of arable land resources to ensure crop safety.
源自碳酸盐风化的镉自然高背景值区域受到了广泛关注。由于喀斯特地区不同母质的土壤理化性质、土壤镉含量和生物有效性存在显著差异,利用土壤总镉含量来划分耕地环境质量存在一定局限性。本研究系统采集了典型喀斯特地区残积母质和冲积母质的表层土壤及玉米样品;分析了玉米镉、土壤镉、pH值和氧化物含量,揭示了不同母质土壤的镉地球化学特征及其生物有效性的影响因素,并基于预测模型提出了科学有效的耕地利用分区建议。结果表明,喀斯特地区不同母质土壤的理化性质差异明显。冲积母质土壤镉含量低但生物有效性高,玉米镉超标率高。玉米镉生物累积系数与土壤CaO、pH值、Mn和TC显著负相关,相关系数分别为-0.385、-0.620、-0.484和-0.384。与多元线性回归预测模型相比,利用随机森林模型预测玉米镉富集系数具有更高的准确性和精度。此外,本研究提出了基于土壤镉和预测作物镉含量的地块尺度耕地安全利用新方案,充分利用耕地资源以确保作物安全。