Environmental Geology

, Volume 57, Issue 1, pp 91–98

Controlling factors and environmental implications of mercury contamination in urban and agricultural soils under a long-term influence of a chlor-alkali plant in the North–West Portugal

Authors

    • CESAM and Department of ChemistryUniversity of Aveiro
  • S. M. Rodrigues
    • CESAM and Department of ChemistryUniversity of Aveiro
  • C. Mieiro
    • CESAM and Department of ChemistryUniversity of Aveiro
  • E. Ferreira da Silva
    • ELMAS, Department of GeosciencesUniversity of Aveiro
  • E. Pereira
    • CESAM and Department of ChemistryUniversity of Aveiro
  • A. C. Duarte
    • CESAM and Department of ChemistryUniversity of Aveiro
Original Article

DOI: 10.1007/s00254-008-1284-2

Cite this article as:
Cachada, A., Rodrigues, S.M., Mieiro, C. et al. Environ Geol (2009) 57: 91. doi:10.1007/s00254-008-1284-2

Abstract

This study aims at assessing the extent of total mercury (Hg) contamination in urban and agricultural soils under long-term influence of a chlor-alkali plant, located at about 1 km away from a town centre. Moreover, it aims at identifying the main factors controlling Hg contents’ distribution and associated potential hazards to environment and human health. The median value of total Hg for soil surface layer (0–10 cm) was 0.20 mg/kg (data ranging from 0.050 to 4.5 mg/kg) and for subsurface layer (10–20 cm) 0.18 mg/kg (data ranging from 0.046 to 3.0 mg/kg). The agricultural area showed higher Hg concentrations (ranging from 0.86 to 4.5 mg/kg) than urban area (ranging from 0.05 to 0.61 mg/kg), with some results exceeding target values set by the Dutch guidelines. Mercury concentrations observed in the studied area are more likely to be associated with the influence of the chlor-alkali plant and with the use of historically contaminated sludges and water from a nearby lagoon in agriculture, than to the impacts of urban development. The statistical correlations between Hg concentrations and soil properties suggest that anthropogenic metal sources should influence the spatial distribution more than the geological properties. Although the Hg emissions were drastically reduced 10 years ago, the area under influence of the chlor-alkali plant is still facing potential health and environmental threats arising from soil contamination.

Keywords

MercuryUrban soilsAgricultural soilsChlor-alkali plant

Introduction

Mercury (Hg) and its compounds are widely recognized as highly toxic to humans, ecosystems and wildlife (EC 2005). In general, Hg deposited in soils derives mostly from wet and dry removal/deposition processes of Hg emitted from natural surfaces or geological sources and/or re-emission of Hg originally released by or discharged from anthropogenic sources (Schroeder and Munthe 1998). Once in the soil compartment, Hg can persist for a long time even after sources have been removed and it can undergo several chemical and biological transformation processes such as the following: adsorption to humic substances, Hg0 oxidation, Hg (II) reduction or methylation; depending on chemical parameters such as pH, temperature, potential redox and organic matter; and also on soil biological characteristics (Stein et al. 1996; UNEP 2002).

It is known that soil contamination may strongly affect the human health, for example through inhalation, skin contact or ingestion of contaminated particles (Abrahams 2002). In addition, Hg accumulated in soils may leach to groundwater or may be taken up by crops and eventually enter the food chain. However, the effects on human health of Hg contents in soils are still unclear, and therefore it is difficult to establish threshold concentrations above which toxic problems are likely to occur. The severity of soils pollution and its hazard will be dependent not only on total Hg concentration, but also on the proportion of their mobile and bioavailable forms, which are controlled by several physicochemical properties of soil (Stein et al. 1996; UNEP 2002).

Although there are important natural sources of Hg that include volcanoes, gaseous release from sea waters, lakes, topsoils and biomass burning (i.e., forest fires), anthropogenic activities are considered a major source of Hg and a significant contributor to soil contamination. The drainage from ancient mining areas (i.e. Almaden in Spain, Idrja in Slovenia and Monte Amiata in Italy) or nearby industries and maritime ports, atmospheric emissions from coal or fossil fuel combustion, cement production, waste incineration and metal smelting, refining and manufacturing are documented as the most relevant Hg sources (Schroeder and Munthe 1998; UNEP 2002; Bloom et al. 2004; EC 2005; Zagar et al. 2006). Moreover, Hg has been widely used in pigment paints, thermometers, vapour and fluorescence lamps, dental preparations, manufacturing of batteries and electrical devices (DEFRA 2002).

Agriculture is also an important source of Hg in soils due to the use of sewage sludge or the application of fertilisers, lime and manure containing Hg in minor amounts. Irrigation with water abstracted from polluted streams and former use of fungicides and seed disinfectants containing Hg were also important sources of Hg in the past (DEFRA 2002). However, emissions of Hg from chlor-alkali plants using metallic Hg for the electrolytical production of chlorine have been reported to have a significant impact on the environment (Biester et al. 2002) and they were considered the single largest source category of anthropogenic Hg in many industrialized countries, prior to the 1970s (Schroeder and Munthe 1998).

The studied town, Estarreja, is under the influence of a chemical complex, known to be one of the major sources of pollution in the region and located at about 1 km away from the town centre. A chlor-alkali plant, functioning since 1950s in this chemical complex, was an important source of Hg contamination in the environment for the past several decades. In spite of the extensive characterization of Hg contamination over a nearby lagoon (Pereira et al. 2005; Coelho et al. 2006), there are no reports on the impacts of the chemical complex on the urban area of Estarreja or on soil from the agricultural sites located in the surrounding areas.

This study aims at assessing whether soil contamination by Hg is of concern in the urban area and nearby agricultural sites and whether it is limited to the vicinity of the industrial area. In addition, it aims at identifying the most relevant factors controlling the spatial distribution of Hg contamination in the nearby areas of the chlor-alkali plant and potentially associated environmental impacts.

General characterization of the study area

Estarreja is a small coastal town in the Northwestern Portuguese coast as shown in Fig. 1, with an urban area of about 2.5 km2 and a population of about 7,000 inhabitants. Since the 1950s and until 1975, chlor-alkali effluents were discharged into open canals through several kilometres of agricultural fields until reaching the lagoon of Aveiro, an important ecosystem in the northwestern Portuguese coast. In the middle of the 1990s, the chlor-alkali plant replaced the mercury cathodes used in the electrolytic cells by a cleaner technology (membrane cells) and Hg emissions were drastically reduced, as observed by Pereira et al. (1997), in the sedimentary record of the lagoon.
https://static-content.springer.com/image/art%3A10.1007%2Fs00254-008-1284-2/MediaObjects/254_2008_1284_Fig1_HTML.gif
Fig. 1

Map showing the location of the urban areas of Estarreja and Aveiro, the coastal lagoon of Aveiro and the chemical complex including the chlor-alkali plant

The soil type of Estarreja area is mainly podzol, found in sand dunes and low terraces (5–8 m), although Cambisols can also be found (Inácio et al. 1998). Podzols are characterized by their low nutrient status, sandy texture, high C/N ratio, low cation exchange capacity, high permeability and low pH value (Inácio et al. 1998).

Material and methods

Soil sampling and sample pre-treatment procedures

Twenty-six sampling sites were chosen as representative for the study area: 9 from ornamental gardens/recreational parks (OG/PA), 6 from roadsides (RD) and 11 from agricultural (AG). In each sampling site composite samples were collected (about 1 kg in each point) at two depths: surface (SF) at 0–10 cm and subsurface (SB) at 10–20 cm. Samples were collected by using a plastic spade previously cleaned with distilled water and ethanol and transferred into plastic bags. Once at the laboratory the samples were dried (oven dried at 40°C until constant weight), sieved (<2 mm) and grounded (<150 μm, using an agate mill), according to ISO 11464 method (ISO 1994a).

Analytical procedures and QC/QA methodologies

Soil pH in CaCl2 and water (ISO 1994b), organic matter (LOI at 400°C) and cation exchange capacity (CEC) (ISO 1995) were determined in the <2 mm fraction. Elemental C, N and H analysis was performed (in the <150 μm fraction) using a microanalyser CNHS-932 (LECO). Sand, silt and clay fractions were quantified by using a Sedigraph 5100 (Micromeritics®).

Hg concentrations were determined by pyrolysis atomic absorption spectrometry (AAS) with gold amalgamation using AMA-254 (LECO). This method is based on the pyrolysis of the soil sample in a combustion tube at 750°C in an oxygen atmosphere, collection of elemental mercury vapour in a gold amalgamator and detection by AAS. The particular advantage of this technique is that determinations are directly performed on solid samples (<150 μm fraction) without further sample pre-treatment, and therefore less likely to be prone to matrix interferences. Sample detection limit (based on 500 mg sample size) was calculated to be 21 ± 6 ng/kg (defined as 3 × standard deviation of the blank; n = 15; 95% confidence interval). All determinations are reported on a dry matter basis.

Precision was evaluated by calculating variation between triplicate analyses, being the relative standard deviation (RSD) bellow 7% for all replicates. Two reference materials were used as an estimate of accuracy: BCR 141R, calcareous loam soil (Hg total content 0.25 ± 0.02 mg/kg mean half-width 95% confidence interval of the mean) and BCR 142R, light sandy soil (Hg total 0.067 ± 0.011 mg/kg). Recovery values were 96 ± 2% for BCR 141R and 96 ± 3% for BCR 142R.

Descriptive statistics and hypothesis testing

SPSS 15.0 for Windows was used for statistical analysis of data. Descriptive statistics was initially performed on the data, box-plots were obtained and normality was tested by running the Kolmogorov–Smirnov test. Some parameters did not follow a normal distribution, and in such cases non-parametric statistics was used: Mann–Whitney Rank Sum Test was used for the comparisons between agricultural and urban soils and between the two depth layers. ArcGis® 9.0 was used to plot maps for highlighting the spatial distribution of soil mercury concentrations.

Results

General characterization of soils

In what concerns the general parameters, Estarreja soils can be characterized as slightly acidic (pH ranging from 4.8 to 7.2, with a median of 6.2), with low CEC (median of 11.1 cmol/kg, ranging from 5.3 to 19.4), intermediate content in total carbon (median of 1.9%, ranging from 0.49 to 5.1) and low total nitrogen content (median of 0.22%, ranging from 0.05 to 0.5). As regards texture (according to the United States Department of Agriculture classes), they are classified as sandy silt or silty sand. General parameters characterized were in accordance with the expected results for the soil type, which are podzols.

Mercury concentrations in soils

The studied area can be divided into three categories: urban, agricultural and industrial. Nevertheless, no sampling was carried out inside the industrial area. The distribution of Hg as well as the location of industrial and urban areas can be observed in Fig. 2 (for SF layer) and 3 (for SB layer).
https://static-content.springer.com/image/art%3A10.1007%2Fs00254-008-1284-2/MediaObjects/254_2008_1284_Fig2_HTML.gif
Fig. 2

Spatial distribution of Hg in surface layer of Estarreja soils (©Instituto Geográfico Português and Direcção Geral dos Recursos Florestais)

https://static-content.springer.com/image/art%3A10.1007%2Fs00254-008-1284-2/MediaObjects/254_2008_1284_Fig3_HTML.gif
Fig. 3

Spatial distribution of Hg in subsurface layer of Estarreja soils (© Instituto Geográfico Português and Direcção Geral dos Recursos Florestais)

When considering the entire sampling area, the median concentration of Hg in surface layer (SF) was 0.20 mg/kg, ranging from 0.05 to 4.5 mg/kg. For subsurface layer (SB) the median concentration was 0.18 mg/kg ranging from 0.05 to 3.0 mg/kg. Hg levels did not show a normal distribution in any of the two layers. Besides, the variability of results—high standard deviation (0.94 mg/kg in SF and 0.72 mg/kg in SB), positive skewness (3.7 mg/kg in SF and 3.1 mg/kg in SB) and median values much lower than the mean values (0.48 mg/kg in SF and 0.40 mg/kg in SB)—can be considered a qualitative indicator of an anthropogenic origin of this element. No significant differences were found between the two layers (Mann–Whitney Rank Sum Test, p = 0.712). Figure 4 shows the box plots of Hg levels for both layers highlighting the dispersion of results (with the outliers and extreme values).
https://static-content.springer.com/image/art%3A10.1007%2Fs00254-008-1284-2/MediaObjects/254_2008_1284_Fig4_HTML.gif
Fig. 4

Box-plot showing the variation of Hg concentrations for both layers in all sampling area and within agricultural (AG) and urban (UR) area. Boxes define the interquartile range and the line is the median. Outliers (values between 1.5 and 3 box lengths) and extreme values (more than 3 box lengths) are also shown

Separating the study area into two different areas, urban and agricultural (which include all sampling sites outside the urban and industrial areas shown in Figs. 2 and 3), the highest median concentration was found in the agricultural sites, both in SF and SB layers. Figure 2 (SF samples) and Fig. 3 (SB samples) show that concentrations of Hg are clearly lower in the town centre than in its outskirts. Hg concentrations in the urban area show a median of 0.13 mg/kg ranging from 0.05 to 0.61 mg/kg in SF layer and a median of 0.11 mg/kg ranging from 0.05 to 0.57 mg/kg in SB layer. The median value for the agricultural area was 0.21 mg/kg (ranging from 0.12 to 4.5 mg/kg) in SF layer and 0.23 mg/kg (ranging from 0.86 to 3.0 mg/kg) in SB layer. These differences are highlighted in Fig. 4. Differences between the two areas were found to be statistically significant (Mann–Whitney Rank Sum Test) for both layers p = 0.018 for SF and p = 0.009 for SB).

A significant (p = 0.05) negative correlation (ρ = 0.453) was observed between Hg and carbon/nitrogen (C/N) ratio in the SF layer, while in the SB layer significant correlations were found between Hg and \( {\rm p}{{\text{H}}}_{{{\text{CaCl}}_{2} }} \) (ρ = −0.569), Total Nitrogen (ρ = 0.603), Total Carbon (ρ = 0.447) and C/N (ρ = −0.423).

Discussion

Main drivers of mercury contamination in soils: the influence of human activities

Industrial activity, urban development and agricultural practices can be important drivers of Hg contamination. The spatial distribution of Hg in surface layer of the studied area, as shown in Figs. 2 and 3, shows that some of the highest Hg concentrations can be found in the samples collected near the chemical complex (e.g. sample 15.AG and 10.RD). High values were also found in sample 24.AG, located near a canal, which transported the contaminated effluents from the chemical complex into a water stream (close to sample 25.AG) that ends in the lagoon. Nevertheless, in spite of some high values of Hg found, soils from this study are not as heavily contaminated as expected due to proximity and long term influence of the chlor-alkali plant, and considering a previous study conducted in the industrial area (within an area of 10 × 6 km) that reported a median value of 0.59 mg/kg ranging from 0.12 to 49 mg/kg (Inácio et al. 1998). The study by Inácio et al. (1998) was performed within the chemical complex area, and therefore, only the sites located in the immediate vicinity the chemical complex are coincident with those from present study. Nevertheless, in such cases, Hg concentrations were found to be higher in the past. This can be due to analytical methods used, as analysis were performed in the <63 μm fraction, which results in an Hg enrichment. Moreover, Inácio et al. (1998) carried out their study when the chlor-alkali plant was still using a few mercury-cells. Nowadays, the emissions are considerably reduced due to changes in the production technology and therefore the levels observed now might be a legacy of the effects of industrial contamination in soils surrounding the chemical complex. The type of soil (mainly acidic sandy soils with low retention capacity) and also re-emissions of Hg to the atmosphere may have contributed to reduction of Hg contents accumulated in soils in the past.

It is not an easy task to make a distinction between the influence of the industrial Hg emissions and other anthropogenic impacts on the Hg contents observed in this study. There are three complicating factors that should be taken into account when considering the land uses: (a) the existence of small patches of agricultural soil within the main urban area; (b) patches of urban and industrial areas in the main agricultural area; and, (c) disturbed sites hardly classified into any of the categories mentioned above.

Values obtained for the urban area of this study appear to be only slightly higher than those obtained in a similar study conducted in a non-industrialized urban area, Aveiro (located 15 Km South from Estarreja, Fig. 1), where a median value of 0.09 mg/kg and a range of 0.01–0.59 mg/kg were observed within the 0–20 cm depth (Rodrigues et al. 2006).

Given the proximity of the two towns, the similarity of socio-economic and geographical features, the fact that the sampling regimes, sample pre-treatment and analytical methods were the same in both studies and that general soil properties (organic carbon content, pH, texture) of Aveiro urban soils are similar to those from Estarreja soils, the impacts of the urbanisation process (mainly the high heterogeneity found) appear to be similar in these two urban areas and may explain the values obtained within the town. However, the two towns considerably differ in terms of industrial background and size, and therefore, the influence of the industry on the levels obtained for Estarreja should not be disregarded. The comparison of the results from this study with those of other urban studies conducted worldwide can be misleading mainly due to differences among cities such as geological characteristics, population, living habits and industry, or due to different methodological approaches. Nevertheless, in general terms, values obtained in the entire study area can be considered comparable to those of Jacobstad, Finland (Peltola and Åström 2003), Antalya, Turkey (Guvenç et al. 2003), Oslo, Norway (Tijhuis et al. 2002), Poznań, Poland (Boszke and Kowalski 2006) and smaller than those from Palermo (Manta et al. 2002) or the Berlin urban area (Birke and Rauch 2000), reported in Table 1.
Table 1

Median, minimum and maximum concentration of Hg (mg/kg) in Estarreja soils and in other cities, and the number of inhabitants in each location

 

Inhabitants

Median value

Min

Max

Reference

Estarreja, Portugal

7,000

0.198

0.050

4.5

Present study

Aveiro, Portugal

36,000

0.091

0.015

0.50

Rodrigues et al. 2006

Jakobstad, Finlanda, b

20,000

0.093

0.010

2.31

Peltola and Åström. 2003

Berlin, Germanyb

3,000,000

0.190

0.420

71.2

Birke and Rauch 2000

Antalya, Turkey

200,000

0.090

Guvenç et al. 2003

Uppsalla, Swedena

136,500

0.139

n.dc

3.66

Ljung et al. 2006

Oslo, Norwaya, b

700,000

0.060

<0.010

2.30

Tijhuis et al. 2002

Poznań, Poland

578,900

0.146 (mean value)

0.017

0.746

Boszke and Kowalski 2006

Palermo, Italyb

1,000,000

0.680

0.040

6.96

Manta et al. 2002

a Aqua regia

b <2 mm

c Not detected

As Hg contents in urban sites can be considered low and are comparable to a small urban area nearby and other cities worldwide, the considerably higher Hg concentrations observed in sites outside the town centre are more likely to derive from the impact of industry rather than urbanisation.

Pathways of contamination and factors controlling mercury distribution in soils

As no gradient in Hg contents was observed with increasing/decreasing distances from the chemical complex, the deposition of airborne Hg, originally emitted from the chlor-alkali plant is not expected to be the only factor controlling the observed Hg concentrations in soil samples from the study area. The variability of Hg concentrations across the study area is expected to be related to the following pressures:
  • Atmospheric deposition and accumulation in soils of Hg emitted from the chlor-alkali plant in the past (historical legacy), more significant in the immediate vicinity of the chemical complex: as there are no historical data of Hg atmospheric concentrations, the effective contribution of this source is difficult to be determined.

  • Atmospheric deposition of Hg re-emitted from the industrial area (chemical complex) and the lagoon where Hg has been accumulated in the past as suggested by Biester et al. (2002). Some samples (10, 15, 16, 19 and 26) seem to be affected by atmospheric deposition as they show a topsoil enrichment higher than 1.5.

  • The application of sludges collected from lagoon canals in agricultural soils has been an historical practice in this region. This practice, associated with the irrigation of the soil with waters contaminated with Hg in the past, flooding occurrences, as well as possible use of certain Hg-containing fertilizers and pesticides in the past, may have been very important factor contributing to the accumulation of Hg in agricultural sites.

The physical and chemical properties of soils may also be important factors controlling the distribution of Hg contents in soils, by affecting metal mobilization-immobilization processes. In the present study, no significant correlations were found between Hg and general parameters such as organic matter (LOI), total C, N and H, CEC, exchangeable bases (Ca, Mg, Na and K), sand, silt or clay in the SF layer. None of these parameters explains the Hg variability, suggesting that anthropogenic sources influence soil content more than geological properties. Nevertheless, the negative correlations found between Hg, C/N and pH in SB layer, may be due to the association of Hg with the humus content (as low pH and low C/N ratio are indicative of humification). For the SF layer, the negative correlation between C/N ratio and Hg was also observed, although no correlation was found between Hg and the pH. Steinnes (1995) reported that in acidic soils with high humic content, Hg volatilization is inhibited, which may result in a higher retention of Hg in soils. In addition, the low C/N ratios (<11) and low pH values (5–5.5) found in the study area favour the presence of high methylmercury/total mercury ratios, particularly in frequently flooded areas, as reported by Remy et al. (2006). As only total Hg concentrations were determined in this study, it is not possible to evaluate the extent of the methylation processes occurring in the area. The analysis of methylmercury concentrations in these soils should be considered in future studies.

Potential environmental implications of mercury contamination

Soil quality guidelines and/or target values established in several countries for residential, recreational and agricultural areas are shown in Table 2. Considering the target value of 0.3 mg/kg of Hg in soils as set by the Dutch guidelines (VROM 2000), Hg concentration in soil samples (seven surface and five subsurface samples) from eight sites (9.SB, 10.SF, 14.SF/SB, 15.SF/SB, 16.SF, 24.SF/SB, 25.SF/SB and 26.SF—six of these sites within the agricultural area) was found to exceed this target value. Samples 15, 24 and 25 from agricultural sites showed Hg levels that are above those stipulated by Portuguese legislation for agricultural sites (1.5 mg/kg). All values observed were below the Dutch intervention value (VROM 2000) of 10 mg/kg and the Canadian guidelines for residential areas (6.6 mg/kg) (CCME 2002).
Table 2

Quality guidelines or target values established in some countries for residential, recreational and agricultural areas (mg/kg)

Country

Land use

Guideline/target

Limit/intervention

Portugala

Agriculture

1/1.5/2b

Canadac

Residential

6.6

Agriculture

6.6

Netherlandsd

0.3

10

Finlande

0.2

5

a Decreto-Lei 118/2006 de 21 de Junho

b pH ≤ 5.5/5.5–7 > 7.0

c CCME 2002

d VROM 2000

e Peltola and Åström 2003

The eight sites with values exceeding the Dutch guideline values, with the exception of sites 14, 16 and 26, are under the direct influence of the chemical complex. Site 16.OG is in the town centre of Estarreja, very close to a school. Sites 14.PA and 26.AG are located on the southern part of the study area. Site 14.PA is of special concern as it is a park where children normally engage in recreational activities. This site is very close to (and often inundated by) the river Antuã, which is severely contaminated by effluent discharges from small electroplating industries, textile industries and untreated domestic sewage. The main reason for these results should be the higher retention capacity of the soils in these sites due to their specific properties, mainly humification. In addition, sample 14 shows the highest percentage of fine fraction and organic matter that can result in a higher retention capacity of Hg.

As concentrations of Hg will be dependent on pedo-lithological conditions, it is difficult to compare with other locals. In the particular case of the studied area, where the underlying rocks are mainly sandy rocks, very low levels of trace elements are expected as result of dissolution and chemical weathering. Therefore, information on regional background is useful, especially when establishing permissible pollution levels and comparing with other guidelines. An upper limit of 1.5 mg/kg is suggested by Inácio et al. (1998) as the acceptable levels of Hg for the chemical complex area, which is above the target guidelines of Netherlands, but similar to the Portuguese guidelines. Nevertheless, the median background value for the A horizon of Portuguese soils is estimated to be 0.05 mg/kg in Hg and for the O horizon 0.06 mg/kg (Ferreira 2004). A range between 0.04 and 0.05 mg/kg was also reported in the geochemical Atlas of Europe for the Northwest Portugal (Salminen 2005). The average Hg concentration in soils over the world (Manta et al. 2002) is reported to be from 0.05 to 0.10 mg/kg. In general terms, Estarreja showed a median concentration two times greater than the maximum mean value of soils over the world and four times greater than the Portuguese background.

Conclusions

Although nowadays the emissions of Hg from a chlor-alkali plant located at a chemical complex in Estarreja (Portugal) are considerably reduced, the legacy of decades of historical contamination cannot be disregarded. The Hg concentrations found in the present study are lower than those found in previous studies in soils from the vicinity of the chemical complex. However, a few sites can be considered contaminated when taking into account guidelines for soil protection. Historical Hg contamination is still affecting the soil quality in the study area, the chlor-alkali plant being the main anthropogenic source of Hg both directly (atmospheric deposition in the past) and indirectly (re-emission of Hg from contaminated sites, and application of contaminated sediments and water from the nearby lagoon). Agricultural sites, still under pressure by the contaminated channels, may constitute thereby a potential threat to human health. Avoiding the use of sludges from the nearby channels of “Ria de Aveiro” in agriculture is strongly recommended. In fact, this practice has been identified as an important factor affecting Hg concentrations in agricultural soils. Further studies on the uptake and bioaccumulation of Hg by plants in agricultural sites from the region are also recommended. Since soils in the area under study are mainly acidic sandy soils with low retention capacity, there is also potential risk of contamination for groundwater and this should be further analysed.

The effects of urbanisation on soils from the town centre are not very significant, and Estarreja can be compared to low-polluted cities such as Aveiro (Portugal), showing lower contamination levels than other more populated and strongly developed cities. The highest Hg concentration levels are more likely to be attributed to the industrial activities rather than urbanisation.

The statistical correlations between Hg concentrations and soil properties (organic matter, pH, texture, cation exchange capacity, total C, N and H) were assessed and such data suggested that anthropogenic sources should influence the spatial distribution more than the geological properties, especially in the SF layer. Hg content in soils is correlated with humus, this being the only soil property influencing Hg distribution in topsoils in the study area.

These data can be useful for further detailed studies and to complement other geochemical databases from urban areas. Potential hazards were identified and further works should encompass health studies and population interventions aimed at reducing exposures to Hg.

Acknowledgments

This work was part of the Portuguese project POCTI/CTA/44851/2002: SOLURB (“Towards a methodology for the assessment of environmental quality in urban soils”) funded by the Portuguese Foundation for Science and Technology (FCT).

Copyright information

© Springer-Verlag 2008