Purpose

Source apportionment is a crucial step toward reducing heavy metal (HM) pollution within soils. Although comparison of receptor models to apportion sources is well-established in air and water pollution, it has been poorly implemented in evaluations of soil pollution. This study aimed (1) to assess the accumulation of As, Cr, Cu, Ni, Pb, Zn, Cd and Hg in soils of fluvial islands of the lower Yangtze River and (2) to apportion sources in a comparative evaluation of two receptor models, thus providing basic information for rational management of soil pollution in the study area.

Materials and methods

Soil samples were collected from representative fluvial islands in the Anhui section of the lower Yangtze River, China. A total of 207 topsoil samples (0–20 cm) were collected at a density of 1 sample per 1 km² and 43 subsoil samples (150–180 cm) were collected at a density of 1 sample per 4 km². The extent of pollution was estimated by determining total soil HM concentrations and calculating the geoaccumulation index (I_geo). The Local Moran’s index (LMI) was applied to detect the spatial distribution of HMs. Absolute principal component scores-multiple linear regression (APCS–MLR) and positive matrix factorization (PMF) were adopted to apportion the sources of HMs.

Results and discussion

The mean degree of contamination of the HMs can be ranked as follows: Cd > Hg > Cu > Ni > Cr > Zn > As > Pb. Average HM concentrations in topsoil were consistently higher than those in subsoil, with total HM concentrations varying between fluvial islands. Source apportionments suggested that the particle size effect was the dominant factor (59.3%) in the APCS–MLR model, whereas in the PMF model, calcium minerals played the most important role (50.2%). Concentrations of Cr and Ni were mainly related to the particle size effect, whereas concentrations of As, Cu, Pb and Zn were mainly related to the particle size effect, fixation by calcium minerals and atmospheric deposition from mining activities. The primary source for Cd anomalies was associated with calcium minerals and atmospheric deposition, whereas the accumulation of soil Hg was mainly driven by coal combustion.

Conclusions

This study indicated there was no extensive HM contamination in the study area except for Cd. Compared with APCS–MLR, PMF provided an optimal reconstruction of HM concentrations in soils. Thus, the PMF model could be applied to soils for HM pollution management in areas with multiple emission sources.

中文翻译：

---

长江下游安徽河滩区土壤重金属的源解析：APCS–MLR和PMF的比较

目的

源分配是减少土壤中重金属（HM）污染的关键步骤。尽管在空气和水污染中已经建立了将受体模型与分配源进行比较的方法，但是在土壤污染评估中却没有很好地实现。这项研究的目的是（1）评估长江下游河流岛土壤中As，Cr，Cu，Ni，Pb，Zn，Cd和Hg的积累，以及（2）比较两种受体的分配来源模型，从而为研究区土壤污染的合理管理提供基础信息。

材料和方法

土壤样品是从长江下游安徽段的代表性河滩上采集的。总共以每1 km ² 1个样品的密度收集了207个表土样品（0–20 cm），并且以每4 km ² 1个样品的密度收集了43个地下土壤样品（150–180 cm）。通过确定土壤总HM浓度并计算_地质累积指数（I _geo）估算污染程度。应用本地莫兰指数（LMI）来检测HM的空间分布。采用绝对主成分评分-多元线性回归（APCS–MLR）和正矩阵分解（PMF）来分摊HM的来源。

结果和讨论

HMs的平均污染程度可按以下等级排序：Cd> Hg> Cu> Ni> Cr> Zn> As> Pb。表层土壤中的平均HM浓度始终高于地下土壤中的总HM浓度，河流总岛之间的总HM浓度也有所不同。来源分配表明，颗粒尺寸效应是APCS中的主要因素（59.3％）–MLR模型，而在PMF模型中，钙矿物质起着最重要的作用（50.2％）。Cr和Ni的浓度主要与粒径效应有关，而As，Cu，Pb和Zn的浓度主要与粒径效应，钙矿物固着和采矿活动中的大气沉积有关。Cd异常的主要来源与钙矿物和大气沉积有关，而土壤Hg的积累主要由煤燃烧驱动。

结论

这项研究表明，除镉外，研究区域内没有大量的重金属污染。与APCS – MLR相比，PMF提供了土壤中HM浓度的最佳重建方法。因此，PMF模型可应用于土壤中多排放源地区的重金属污染管理。