| 马雪璞,石伟伟,李尤,廖晓勇,王亮,罗凌志,曹红英.中国南方典型低山丘陵区不同地形下土壤镉累积的影响因素[J].农业环境科学学报,2026,45(7):1654-1664. |
| 中国南方典型低山丘陵区不同地形下土壤镉累积的影响因素 |
| Influencing factors of soil cadmium accumulation under different topographies in typical low mountain and hilly areas of southern China |
| 投稿时间：2025-07-29 |
| DOI：10.11654/jaes.2025-0730 |
| 中文关键词: 南方低山丘陵区 土壤镉累积 地形梯度 影响因素 随机森林 |
| 英文关键词: southern low mountain and hilly area soil cadmium accumulation terrain gradient influencing factor random forest |
| 基金项目:国家自然科学基金项目（42225707） |
| 作者 | 单位 | E-mail | | 马雪璞 | 中国科学院地理科学与资源研究所, 北京 100101
中国科学院大学, 北京 101408 | | | 石伟伟 | 中国科学院地理科学与资源研究所, 北京 100101
中国科学院大学, 北京 101408 | | | 李尤 | 中国科学院地理科学与资源研究所, 北京 100101 | | | 廖晓勇 | 中国科学院地理科学与资源研究所, 北京 100101 | | | 王亮 | 中国科学院地理科学与资源研究所, 北京 100101 | | | 罗凌志 | 中国科学院地理科学与资源研究所, 北京 100101
中国科学院大学, 北京 101408 | | | 曹红英 | 中国科学院地理科学与资源研究所, 北京 100101 | caohy@igsnrr.ac.cn |
|
| 摘要点击次数: 421 |
| 全文下载次数: 32 |
| 中文摘要: |
| 为了解低山丘陵区地形对土壤中Cd累积的影响机制，根据地形梯度高低采集了湖南某典型小流域251个表层土壤样品，以地形生态位指数（综合高程和坡度计算得出）反映地形梯度，分析土壤Cd在地形梯度上的累积特征，结合空间插值与莫兰指数探究地形梯度对Cd空间分布的影响，并利用随机森林算法识别出6种因子在不同地形梯度下对Cd累积的影响差异。结果表明：土壤Cd累积量随地形生态位指数的增加呈显著下降趋势（P<0.001），Cd的分布具有显著的空间聚类性，Cd累积热点集中在小流域中部平缓地形区，地形生态位指数<0.45，累积冷点集中在小流域南部和北部的丘陵区。随机森林模型在低、中、高地形下的R2值分别为0.75、0.65、0.73。基于随机森林模型的变量重要性分析结果显示，不同地形下对Cd累积分异的解释程度最大的因素分别是低地形为土壤有机碳（54.82%），中地形为土壤有机碳（32.67%），高地形为土壤电导率（43.43%）。当地形生态位指数>0.3时，土壤黏粒含量和距河流距离的相对贡献显著下降。各因子受地形梯度的影响程度（变异系数）依次为土壤电导率（0.74）>土壤黏粒含量（0.64）>距河流距离（0.48）>土地利用类型（0.36）>土壤有机碳（0.26）>土壤pH（0.24）。研究表明，不同地形梯度土壤Cd累积的影响因子存在差异。 |
| 英文摘要: |
| To understand the mechanism of terrain influence on Cd accumulation in the soil of low mountain and hilly areas, 251 surface soil samples were collected according to terrain gradient in a typical small watershed in Hunan. The terrain ecological index（calculated based on elevation and slope）was used to reflect the terrain gradient, and the accumulation characteristics of soil Cd on the terrain gradient were analyzed. The effects of terrain gradient on the spatial distribution of Cd were explored through spatial interpolation and Moran index, and by employing the random forest algorithm, this study identified variations in the influence of six factors on Cd accumulation across different terrain gradients. The results showed that the accumulation of soil Cd significantly decreased with the increase of terrain ecological index（P<0.001）. The distribution of Cd had significant spatial clustering, with Cd accumulation hotspots concentrated in the middle gentle terrain area of the small watershed（terrain ecological index <0.45）and accumulation cold spots concentrated in the hilly area in the south and north of the small watershed. The R2 values of the random forest model under low, medium, and high terrains were 0.75, 0.65, and 0.73, respectively. Based on the variable importance analysis derived from the random forest model, the dominant factors explaining the differentiation of Cd accumulation under different topographic conditions were as follows：in low-lying areas, soil organic carbon（54.82%）; in mid-elevation areas, soil organic carbon（32.67%）; and in high-elevation areas, soil electrical conductivity（43.43%）. When the terrain niche index exceeded 0.3, the relative contributions of soil clay content and distance from the river declined markedly. The degree of influence of each factor along the topographic gradient, as indicated by the coefficient of variation, followed the order：soil electrical conductivity（0.74）>soil clay content （0.64）>distance from the river （0.48）>land-use type（0.36）>soil organic carbon （0.26）>soil pH（0.24）. This study demonstrates that the factors influencing soil cadmium accumulation differ across various terrain. |
| HTML
查看全文
查看/发表评论 下载PDF阅读器 |
|
|
|
