Multiple regression analysis to assess the contamination with metals and metalloids in surface sediments (Aveiro Lagoon, Portugal)

https://doi.org/10.1016/j.marpolbul.2020.111470Get rights and content

Highlights

  • Heavy metals/metalloids studied in sediments of contaminated lagoon

  • Multiple regression model demonstrated different behavior for metals/metalloids.

  • Effects of geographical distances and geochemistry statistically separated

  • Anthropogenic point contamination source proven for As, Zn, Cu, Pb

  • Calculation of enrichment factors is complementary approach to multiple regression.

Abstract

An innovative multiple regression analysis was used to evaluate metal/metalloid contamination in the surface sediments of a coastal lagoon. The concentrations of metals/metalloids were represented as a function of geochemical characteristics of the sediments (fine fraction, concentrations of organic carbon, Ca, Al, Mn) and distances between sampling points. The effect of distances on the concentrations were negligible for Li, Co, Ni, Ba, V, Cr, and only geochemical variables specific for each element explained its spatial variation. The concentrations of As, Cu, Zn and Pb were influenced by both geochemical and geographical distance variables, the latter representing the anthropogenic influence and the extent of transport of contaminants away from the upstream source. Enrichment of the sediment with Ba, As, Co, Cr and V was determined mainly by enrichment with Mn. The proposed approach is supplementary to the traditional utilization of enrichment factors, and is better suited for systems with anthropogenic influence.

Introduction

Estuarine catchments have been strategic settings throughout human history, either as places of navigation, agricultural abundance or as locations of the biggest cities in the World (Kennish, 1996; McLusky and Elliot, 2004). Usually used as repositories of industrial and domestic effluents, estuaries are the end-point of numerous contaminants, the vast majority of which tend to settle and are thus stored in estuarine and marine sediments (Laurier et al., 2003; Prego and Cobelo-Garcia, 2003; Kim et al., 2004). Due to their toxicity to organisms and persistence in the environment, growing international consciousness has developed with regards to assessing the contamination status and protecting these ecosystems.

Traditionally, the dynamic of metals/metalloids in sediments is evaluated through the utilization of enrichment factors (EFs) (Abrahim and Parker, 2008), but recently, novel methodologies have been developed to evaluate the composition and dynamics of sediment characteristics. For example, artificial neural networks combined with residual kriging have been used to predict the spatial distribution of Cr in soil (Tarasov et al., 2018).

Multiple regression analysis allows for the verification of simple and higher order effects of several explanatory variables and their interactions. Upon simplification of the starting models, only a few significant explanatory variables will remain. It has been applied recently to evaluate biogeochemical processes in estuarine water (Stoichev et al., 2016, Stoichev et al., 2020). Multiple regression was used to quantify major components of lake sediments by near infrared spectra (Russell et al., 2019) or to find out the relative importance of hydrous iron and manganese oxides on the retention of trace metals in estuarine sediments (Turner, 2000). Multiple regression analysis has been successfully applied to study the behavior of Hg species in surface sediments from the Aveiro Lagoon (Stoichev et al., 2019). A relatively small area of the lagoon, the Laranjo Bay, suffered from chlor-alkali mercury pollution coming from a single upstream source (Pereira et al., 2009; Stoichev et al., 2019). However, sediment contamination in the system is not restricted to mercury, with reports of significant concentrations of other metals and metalloids, such as As, Pb, Zn (Costa and Jesus-Rydin, 2001).

The aim of this work was to evaluate the effectiveness of multiple regression analysis to study spatial variations in heavy metal/metalloid concentrations in surface sediments from contaminated shallow tidal environments. One type of explanatory variables depended on geographical distances. Another type of variables was related to sediment geochemistry, which, in coastal environments, should include indicators of both terrestrial (e.g. Al or Fe) and oceanic (e.g. Ca) influence (Perez et al., 2016; Gredilla et al., 2015a; Dias et al., 2007). An attempt was made to separate and quantitatively evaluate the effects of contaminant dispersion from a point source from those of geochemical processes as a potential influence for spatial distribution of contamination. This is possible if additive statistical effects of some explanatory variables (responsible for metal/metalloid dispersion and for the geochemistry of the sediments) on the dependent variables (e.g. concentration of contaminant) are found. For this purpose, equations were developed modelling metal/metalloid concentrations depending on different distance variables or geochemical explanatory variables.

Section snippets

Study area

Aveiro Lagoon is a coastal lagoon in the North of Portugal (Fig. 1) comprised of a network of channels, opening into the Atlantic Ocean by way of a single narrow channel. The lagoon covers an area of 83 km2 at high tide (HT) and 66 km2 at low tide (LT) and has an average depth of 1 m. The tidal range is minimum 0.6 m during neap tide; maximum 3.2 m during spring tide. The water residence time in the lagoon is approximately 2 days, however, it is more than two weeks in the contaminated area

Potential existence of seasonal trends

Analysis of covariance ANCOVA (season, geochemical variables) for surface sediments from the Aveiro Lagoon (nine samples taken in February and August 2012) demonstrated that seasonal trends for YE (metal/metalloid concentrations) were not observed (results not shown). When, for some sampling sites, winter/summer differences existed, they followed concentration changes observed for geochemical variables Xi. For example, the concentrations of As could be expressed by a combination of TC and the

Conclusions

The metal/metalloids, coming from a single upstream source, were suspected of contaminating surface sediments from a shallow coastal lagoon. Multiple regression analysis demonstrated that, from the studied chemical elements, only As, Cu, Zn and Pb were influenced both by geochemical and geographical distance variables, the latter representing the anthropogenic influence and the extent of transport of contaminants away from the upstream source. It is possible that for Pb, an additional source

CRediT authorship contribution statement

Teodor Stoichev:Conceptualization, Methodology, Writing - original draft.João Pedro Coelho:Formal analysis, Funding acquisition.Alberto De Diego:Conceptualization.Maria Gabriela Lobos Valenzuela:Formal analysis, Funding acquisition.Maria Eduarda Pereira:Funding acquisition.Aubin Thibault de Chanvalon:Conceptualization.David Amouroux:Conceptualization, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This research was supported by national funds through FCT (Foundation for Science and Technology) (Portugal) within the scope of UIDB/04423/2020 and UIDP/04423/2020. The financial support of European SUDOE Interreg IVB Program through the Orque-Sudoe project and of FONDECYT project (1150855) is acknowledged. João Pedro Coelho is funded by CESAM (UIDP/50017/2020+UIDB/50017/2020) and the Integrated Program of SR&TD ‘Smart Valorization of Endogenous Marine Biological Resources Under a Changing

References (45)

  • V.A. Martins et al.

    Assessment of the health quality of Ria de Aveiro (Portugal): heavy metals and benthic foraminifera

    Mar. Pollut. Bull.

    (2013)
  • R. Prego et al.

    Twentieth century overview of heavy metals in the Galician Rias (NW Iberian Peninsula)

    Environ. Pollut.

    (2003)
  • S.M. Rimmer

    Geochemical paleoredox indicators in Devonian–Mississippian black shales, Central Appalachian Basin (USA)

    Chem. Geol.

    (2004)
  • R.L. Rudnick et al.

    Composition of the continental crust

  • T. Stoichev et al.

    Multiple regression analysis to assess the role of plankton on the distribution and speciation of mercury in water of a contaminated lagoon

    J. Hazard. Mater.

    (2016)
  • T. Stoichev et al.

    Multiple regression analysis to assess the spatial distribution and speciation of mercury in surface sediments of a contaminated lagoon

    J. Hazard. Mater.

    (2019)
  • D.A. Tarasov et al.

    High variation topsoil pollution forecasting in the Russian Subarctic: using artificial neural networks combined with residual kriging

    Appl. Geochem.

    (2018)
  • K. Telfeyan et al.

    Arsenic, vanadium, iron, and manganese biogeochemistry in a deltaic wetland, southern Louisiana, USA

    Mar. Chem.

    (2017)
  • A. Thibault de Chanvalon et al.

    Particles transformation in estuaries: Fe, Mn and REE signatures through the Loire Estuary

    J. Sea Res.

    (2016)
  • N. Tribovillard et al.

    Trace metals as paleoredox and paleoproductivity proxies: an update

    Chem. Geol.

    (2006)
  • A. Turner

    Trace metals contamination in sediments from UK estuaries: an empirical evaluation on the role of hydrous iron and manganese oxides

    Estuar. Coast. Shelf Sci.

    (2000)
  • F. Yücesoy et al.

    Heavy metal geochemistry of surface sediments from the southern Black Sea shelf and upper slope

    Chem. Geol.

    (1992)
  • Cited by (15)

    • The effect of ocean warming on accumulation and cellular responsiveness to cobalt in Mytilus galloprovincialis

      2022, Marine Pollution Bulletin
      Citation Excerpt :

      Mussel specimens (Mytilus galloprovincialis) were collected from a non-contaminated area in the Ria de Aveiro, northwest Atlantic coast of Portugal. Recent publications have stated that Co concentrations in sediments from this coastal lagoon, ranging between 4.42 and 7.07 mg kg−1 (Stoichev et al., 2020), are lower compared to sediments from other systems, namely from the Galicia estuary, northwest of Spain (varying from 11.79 to 12.96 mg kg−1) and values of 24 and 54 mg kg−1 in Turkey and Australia, respectively (Birch et al., 2013; Stoichev et al., 2020). Negligible water Co concentrations were also reported in the sampling area, < 2 μg L−1 (Henriques et al., 2019).

    • Metal(oid)s accumulation (Hg and As) and their biochemical effects in Halimione portulacoides (Ria de Aveiro, Portugal)

      2022, Marine Pollution Bulletin
      Citation Excerpt :

      For the determination of total Hg, the already sieved sediment samples and the lyophilized plants were analysed directly in the LECO equipment, without the need for any pre-treatment. Regarding the quantification of total As in sediments and plants, these were performed according to US EPA 3051 protocol using microwave digestion with a mixture (3/1, v/v) of hydrochloric (37%, Sigma-Aldrich, p.a.) and nitric acids (69%, Sigma- Aldrich, p.a.) (Stoichev et al., 2020), followed by analysis in ICP-MS (Thermo Elemental, X-Series) equipped with a Peltier cooled impact bead spray chamber and a concentric Meinhard nebulizer. For metal quantifications in the soluble and insoluble fractions of plant tissues, the extraction protocol adapted by Castro et al. (2009) was applied.

    View all citing articles on Scopus
    View full text