Elsevier

Geoderma

Volume 311, 1 February 2018, Pages 120-129
Geoderma

Geographically Weighted Principal Components Analysis to assess diffuse pollution sources of soil heavy metal: Application to rough mountain areas in Northwest Spain

https://doi.org/10.1016/j.geoderma.2016.10.012Get rights and content

Highlights

  • GWPCA method improves the explanation of the spatial distribution of soil heavy metal.

  • Winning variables from GWPCA analysis are related to geogenetic and atmospheric sources.

  • Ashes from wildfires could be a source of soil heavy metal pollution.

Abstract

Heavy metal pollution can result in the degradation of the soil, air and water bodies' quality affecting the health of all living organism. We analyze the spatial distribution of the concentrations of soil heavy metal and relationship with natural or anthropogenic origin. The analysis was performed in Principality of Asturias (mountain region of NW of Spain), where soil heavy metal pollution has become a severe problem. First, a standard Principal Components Analysis (PCA) was performed over a population of 334 soil samples to identify the sources of fourteen heavy metal and metalloids (Ag, As, Ba, Hg, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, Zn). Due to the high geological heterogeneity of the territory, the PCA analysis was improved using a variant of PCA known as Geographically Weighted Principal Components Analysis (GWPCA). The first six principal components in a standard PCA account for about 57% of soil heavy metal variability but when GWPCA is performed this figure increases to > 80% in some areas. We conclude that GWPC1 corresponds to a geogenetic component with changing winning variables adapted to the geological characteristics of the territory (lithology and mining), while GWPC2 corresponds to a factor related to atmospheric pollution including heavy metal released from vegetation cover via wildfires.

Introduction

Soil heavy metal pollution has become a severe problem in many parts of the world due to the fact that the metal pollution is covert, persistent and irreversible (Bini et al., 2011, Zhang et al., 2009, Zhiyuan et al., 2014). This kind of pollution not only degrades the quality of the atmosphere, water bodies and food crops, but also threatens the health and well-being of animals and human beings by way of the food chain (Dong et al., 2011, Nabulo et al., 2012, Wang et al., 2012). Heavy metals such as mercury (Hg), chromium (Cr), cadmium (Cd) or metalloids like arsenic (As) are present in the environment free or as part of different molecular forms (Chen et al., 1999). In natural soils they are present at a background level and usually occur as cations which strongly interact with the soil matrix (Alloway, 1995). Therefore, some physicochemical properties of soils such as pH and organic matter are important parameters that control the accumulation and the availability of heavy metals in the soil environment. Generally speaking, heavy metals are distributed heterogeneously in the Earth's crust as an effect of geological processes, and the elemental contents of non-polluted soils are incorporated by rock weathering processes. Among these, the main factor that dictates the elemental content of a soil is the composition of parent material but this can be increased due to anthropic causes (Alloway, 1995, Kabata-Pendias, 2004, Harmanescu et al., 2011).

In industrial areas, anthropogenic activities such as agriculture, urbanization, industrialization and mining increase the metal concentration baseline (Adriano, 1992, Sheppard et al., 2000, Facchinelli et al., 2001, Wei and Yang, 2010, Zhong et al., 2012). The geological features such as lithology or mineralized areas associated to faults or thrusts exert a strong control on heavy metal concentrations and their variability in soils (Alloway, 1995, Kabata-Pendias, 2004). Smelting from industrial activities and cities is recognized as the most important source of heavy metals in the environment but little is known about the role of wildfires, which are frequent in mountainous areas. Ash is a key component of the land affected by wildfires (Cerda and Doerr, 2007, Bodi et al., 2014, Pereira et al., 2013b). Furthermore, the ability of some natural plant species, named metalofitas, to take up, translocate and accumulate heavy metals in their shoots (Nanda et al., 1995, Chaney et al., 1997) is well known. The combustion of these plant species could produce smog, necromasse and ashes enriched in heavy metals that, when deposited in topsoil, contribute to raise the concentration of nutrients and pollutants such as heavy metals in soils. Nowadays, a research effort concerning the legacy of atmospherically-deposited elements (e.g. heavy metals) in burned soils is needed but during the last decade some authors have shown the role of ash in the Earth and Soil System (Bodi et al., 2014, Pereira et al., 2013a). Elemental composition studies carried out on ashes from California burned soils reported high concentrations of heavy metals such as Zn > Ba > Cu > Mn > Ag > As > Cd > Cr > Co (Plumlee et al., 2007, Hageman et al., 2008a, Hageman et al., 2008b). Recently, a review about the heavy metal composition in wildfire ashes from Australian soils was published (Santin et al., 2015). All these research studies seem to indicate that wildfires are an important source of heavy metals in soils.

Often geogenetic and anthropogenic sources of heavy metals are superimposed in the territory and it is very difficult to separate the contribution of each one from the soil heavy metal backgrounds, making difficult the identification of the boundary between natural and contaminated soils. Thus, statistical methods such as Principal Components Analysis (PCA) and clustering have been extensively used to identify sources of heavy metals in the environment (Hu et al., 2013). PCA is a useful tool to discover diffuse pollution sources by analyzing metal associations in each principal component (Zhang et al., 2009, Wei and Yang, 2010). However, the PCA method can be complemented using a variant of the method called Geographically Weighted Principal Components Analysis (GWPCA) when spatial heterogeneity occurs in the data (Harris et al., 2011, Demšar et al., 2013). In fact, a statistical hypothesis test is normally performed in GWPCA in order to establish the existence of spatial heterogeneity. In essence, this method consists in performing a local PCA, that is, in the neighborhood of each observation, instead of a global standard PCA.

The main objective of this research is to find the natural soil heavy metal backgrounds in the Principality of Asturias, discovering possible sources of diffuse pollution using PCA over a soil population of 334 taken in the most pristine areas of the territory. The spatial heterogeneity of background levels of metals and its relationship with lithology and human activities was addressed with geographically weighted principal component analysis (GWPCA).

Section snippets

Study area

This research was performed in the Principality of Asturias which is a mountain region with 10,600 km2 located in the north of the Iberian Peninsula, Spain (Fig. 1). It dates from the Hercynian Orogenic cycle, but its relief was rejuvenated during the Alpine cycle and runs parallel to the coast of the Cantabrian Sea following an east–west direction.

The climate of the area is included within the type known as “oceanic cold–temperate domain”, with mild temperatures and abundant rainfall, being

Preliminary analysis

Descriptive statistics of the 334 soil samples show average values below normal values on the Earth's crust shown in the last row of the same table, only Mn, Zn and Hg surpass slightly these thresholds. Nevertheless, it is important to note the high heterogeneity of the studied soils. However, if attention is focused on the range and maximum values of the soil population (Maxi in Table 1), Mn, Zn, Cr, Pb, Co, Ni, Cu, Ba, Sb and Hg surpass the crust normal values.

Table 2 shows the correlation

Discussion

Analyzing Table 5 it is possible to conclude that when the most extensive bedrock is schist the most frequent winning variables in GWPC1 are Cu, Co and Ni. In quartzites the most frequent winning variables are Ba, Hg and Ag, while in the case of limestone, As and Pb are the winning variables. Similarly, Cr and Mo are present in the watersheds with coal beds. Then, all the metals with the highest loadings in GWPC1 are related to watershed bedrocks and could be interpreted as a geogenetic

Conclusions

The use of GWPCA presents clear advantages over standard PCA since the former provides information regarding the spatial distribution of the percentage of variance and the variables with most influence in each of the components, while this information is normally obscured using a global analysis. The analysis of the spatial variability of variance and loadings of the components allows a better comprehension of the relationships between the different variables under study across the study area.

Acknowledgements

This work has been partially funded by the Government of the Principality of Asturias (reference SV-PA-10-16. CN-10-089).

References (64)

  • J. Moreno et al.

    Abiotic ecotypes in south-central Spanish rivers: reference conditions and pollution

    Environ. Pollut.

    (2006)
  • G. Nabulo et al.

    Does consumption of leafy vegetables grown in pen-urban agriculture pose a risk to human health?

    Environ. Pollut.

    (2012)
  • C. Santin et al.

    Quantity, composition and water contamination potential of ash produced under different wildfire severities

    Environ. Res.

    (2015)
  • C. Santin et al.

    Wildfires influence on soil organic matter in an Atlantic mountainous region (NW of Spain)

    Catena

    (2008)
  • D. Sheppard et al.

    Metal contamination of soils at Scott Base, Antarctica

    Appl. Geochem.

    (2000)
  • E. Steinnes et al.

    Large scale multi-element survey of atmospheric deposition using naturally growing moss as biomonitor

    Chemosphere

    (1992)
  • B. Wei et al.

    A review of heavy metal contaminations in urban soils, urban road dusts and agricultural soils from China

    Microchem. J.

    (2010)
  • D.C. Adriano

    Biogeochemistry of Trace Metals. Advances in Trace Substances Research

    (1992)
  • B.J. Alloway

    Heavy Metals in Soils

    (1995)
  • M.J. Baxter

    Standardization and transformation in principal component analysis, with applications to archaeometry

    J. R. Stat. Soc.: Ser. C: Appl. Stat.

    (1995)
  • A. Cerda et al.

    Soil wettability, runoff and erodibility of major dry-Mediterranean land use types on calcareous soils

    Hydrol. Process.

    (2007)
  • H. Chen et al.

    Heavy metal pollution in soils in China: status and countermeasures

    Ambio

    (1999)
  • R.D. Dallmeyer et al.

    Pre-Mesozoic Geology of Iberia: (Papers from an International Conference Hosted by IGCP Project 233 (Terranes in the Circum-Atlantic Paleozoic Orogens) in Sept. 1986 in Oviedo, Spain). Project/IGCP

    (1990)
  • U. Demšar et al.

    Principal component analysis on spatial data: an overview

    Ann. Assoc. Am. Geogr.

    (2013)
  • A. de la Campa et al.

    Geochemistry and origin of PM10 in the Huelva region, southwestern Spain

    Environ. Res.

    (2007)
  • J. Dong et al.

    Assessing the concentration and potential dietary risk of heavy metals in vegetables at a Pb/Zn mine site, China

    Environ. Earth Sci.

    (2011)
  • S. Fernández et al.

    Spatial modelling of organic carbon in burned mountain soils using hyperspectral images, field datasets, and NIR spectroscopy (Cantabrian Range; NW Spain)

    Land Degrad. Dev.

    (2016)
  • A.S. Fotheringham et al.

    Geographically Weighted Regression: The Analysis of Spatially Varying Relationships/cA. Stewart Fotherington, Chris Brunsdon, and Martin Charlton

    (2002)
  • A.S. Fotheringham et al.

    Geographically Weighted Regression: The Analysis of Spatially Varying Relationships

    (2002)
  • H. Ghrefat et al.

    Application of geoaccumulation index and enrichment factor for assessing metal contamination in the sediments of Kafrain Dam, Jordan

    Environ. Monit. Assess.

    (2011)
  • I. Gollini et al.

    GWmodel: an R package for exploring spatial heterogeneity using geographically weighted models

    J. Stat. Softw.

    (2015)
  • F. Guitián-Ojea et al.

    Técnicas de análisis de suelos

    (1976)
  • Cited by (77)

    • Threats to mountainous soils: conservation and management strategies

      2023, Understanding Soils of Mountainous Landscapes: Sustainable Use of Soil Ecosystem Services and Management
    View all citing articles on Scopus
    View full text