JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 9 January 2007
Published in J Environ Qual 36:70-79 (2007)
DOI: 10.2134/jeq2006.0254
© 2007 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Biasioli, M.
Right arrow Articles by Ajmone-Marsan, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Biasioli, M.
Right arrow Articles by Ajmone-Marsan, F.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Biasioli, M.
Right arrow Articles by Ajmone-Marsan, F.
Related Collections
Right arrow Toxic Trace Metals
Right arrow Urban Soils
Right arrow Heavy Metals
Right arrow Soil Pollution

TECHNICAL REPORTS

Heavy Metals in the Environment

Potentially Toxic Elements Contamination in Urban Soils

A Comparison of Three European Cities

M. Biasiolia,*, H. Grcmanb, T. Kraljb, F. Madridc, E. Díaz-Barrientosc and F. Ajmone-Marsana

a DI.VA.P.R.A., Chimica Agraria, Università di Torino, Via Leonardo da Vinci, 44, 10095 Grugliasco, Torino, Italy
b Univerza v Ljubljani, Biotehniska fakulteta, 1000 Ljubljana, Slovenia
c Instituto de Recursos Naturales y Agrobiología de Sevilla (CSIC), Apartado 1052, 41080, Sevilla, Spain

* Corresponding author (mattia.biasioli{at}unito.it)

Received for publication June 30, 2006.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Studies on several cities around the world confirm that urban soils are subject to heavy anthropogenic disturbance. However, these surveys are difficult to compare due to a lack of common sampling and analytical protocols. In this study the soils of Ljubljana (Slovenia), Sevilla (Spain), and Torino (Italy) were extensively sampled and analyzed using common procedures. Results highlighted similarities across the cities, despite their differences in geography, size, climate, etc. Potentially toxic elements (PTE) showed a wide range in concentration reflecting a diffuse contamination. Among the "urban" elements Pb exceeded the legislation threshold in 45% of Ljubljana, 43% of Torino, and 11% of Sevilla samples while Zn was above the limits in 20, 43, and 2% of the soils of Ljubljana, Torino, and Sevilla, respectively. The distribution of PTE showed no depth-dependant changes, while general soil properties seemed more responsive to anthropogenic influences. Multivariate statistics revealed similar associations between PTE in the three cities, with Cu, Pb, and Zn in a group, and Ni and Cr in another, suggesting an anthropogenic origin for the former group and natural one for the latter. Chromium and Ni were unaffected by land use, except for roadside soils, while Cu, Pb, and Zn distribution appeared to be more dependent on the distance from emission sources. Regardless of the location, climate, and size, the "urban" factor—integrating type and intensity of contaminant emission and anthropogenic disturbance—seems to prevail in determining trends of PTE contamination.

Abbreviations: PTE, potentially toxic elements • PO, parks and open spaces • RB, riverbanks • OG, ornamental gardens • RD, roadsides • CCE, calcium carbonate equivalent • OC, organic carbon • LJU, Ljubljana • SEV, Sevilla • TOR, Torino


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A GREAT proportion of the European population lives in cities. By the year 2010 the majority of the world's population will live in urban areas and approximately 40% of them will be children (United Nations, 2004). It has been recognized (Ravetz, 2000; Brown, 2003; van Kamp et al., 2003; Grimm and Redman, 2004) that environmental quality information is of crucial importance for appraising the sustainability of urban areas. Decision-makers and specialists that devise and carry out urban public policy are faced with the necessity to integrate the multidimensional aspect of urban environmental quality into the planning process but the role of soil often fails to be appreciated. The first step toward a better understanding of the importance of soils in the urban environment is to study and describe their unique characteristics, adapting the concepts and methods that are commonly used for agricultural or forestry soils to the urban environments.

Urban areas concentrate an enormous amount of energy and matter to ensure their functions and produce a correspondingly large amount of wastes. The formation of urban agglomerations alters the relationships between the components of the ecosystem, i.e., soil, air, water, and biota. The ecological functions of the soil can be severely limited. For example, surface sealing modifies gas and water circulation in the soil, and contaminants that are emitted by human activities reduce the environmental soil quality. In particular, heavy metal contamination of urban soils was found to influence soil microorganisms in terms of basal respiration rate, microbial biomass, dehydrogenase, phosphatase, sulphatase, and glucosidase activities (Castaldi et al., 2004; Yuangen et al., 2006). The proximity of urban soils to humans increases the probability that soil components, including pollutants, may be carried into the human body through inhalation, ingestion or dermal contact (Abrahams, 2002). In urban settings the soil performs other important social and economical functions that are also often neglected, such as supporting green areas and landscaping (Begonchea-Morancho, 2003; Chiesura, 2004).

Studies of urban soils are available for a number of cities. In Europe data are available for Athens (Chronopoulos et al., 1997), Warsaw (Pichtel et al., 1997), Madrid (De Miguel et al., 1998), Palermo (Manta et al., 2002), and Naples (Imperato et al., 2003), but the lack of a common methodological approach makes these surveys difficult to compare. Intensity of sampling, sampling depth, sample preparation procedures, the selection of size fractions, methods of sample digestion, and analytical techniques vary across the surveys.

In this study, the soils of three cities of southern Europe were extensively investigated and compared, using common sampling and analytical techniques to (i) highlight differences and similarities in soil properties, (ii) to investigate their degree of contamination with potentially toxic elements (PTE), and (iii) to outline a methodology that could be adopted at the continental scale.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The cities of Ljubljana (Slovenia), Sevilla (Spain), and Torino (Italy) were selected for this study. They are part of a consortium of cities that have been under study in the project URBSOIL (EVK4-CT2001–00053) funded by the European Commission within the fifth Framework Program of Research and Development. A summary of the general characteristics of the three cities is presented in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of the three cities.

 
The city of Torino is among the largest cities in Italy; it has approximately 900000 inhabitants and has a long industrial history, mainly car manufacturing factories and metallurgical industries. Torino has developed on an alluvial plain, formed by the rivers Po, Sangone, Stura di Lanzo, and Dora Baltea. The watershed of these rivers contains a mixture of very diverse rocks that have contributed to the chemical composition of the derived soils. In particular, serpentinites are present that might have increased Cr and Ni contents in the alluvial deposits (Lombini et al., 1998). Recent studies have reported high Cr and Ni levels in the alluvial soils of the region (Facchinelli et al., 2001; Abollino et al., 2002; Biasioli et al., 2006).

Sevilla is the largest city of the Autonomous Community of Andalucía (southern Spain), with the fourth largest population in Spain at just over 700000 inhabitants. Its industrial activity has been mainly related with fertilizers, pottery, building materials, ship-building, and processing of agricultural products. Most green areas were established less than one century ago, and many of them are not more than fifty years old. The city is located on the alluvial plain of the Guadalquivir River, which has often caused severe floods in the past.

Ljubljana is the capital, and with almost 260 000 inhabitants, the largest city in Slovenia. Many people from neighbor smaller municipalities work in Ljubljana so daily migration and, as a consequence, traffic are large and still increasing. Besides industry and central heating stations, traffic is the main source of pollution. The northern part of Ljubljana lies on alluvium (predominantly calcareous gravel) of the river Sava. Two hills (Sisenski hrib and Golovec) in the middle of the city consist of noncalcareous shaly claystones, siltstones, and sandstones. In the southern part the fine alluvium sediments of the river Ljubljanica prevail.

Sampling
The sampling strategy was based on preliminary data obtained in previous studies by Madrid et al. (2004) on the city of Sevilla, by Biasioli et al. (2006) on the city of Torino, and by Madrid et al. (2006) on the short range variability of urban soil properties. In those studies the indication had emerged that a judgmental sampling design had to be used to cover the entire city area. Various problems have to be faced in urban soil sampling mainly because of the fragmentation of the exposed surfaces and the variability of soil properties. In urban settings, land use changes relatively rapidly so in many cases the information obtained from the soil maps or aerial pictures can be outdated. Once the sampling sites have been located on a map, a visual survey is then indispensable to review the mere existence of a sampling point. Another crucial inspection regards the homogeneity of the area. A land use change can be very intrusive and mixings or additions can occur that dramatically alter the soil properties (Bullock and Gregory, 1991; Pichtel et al., 1997; Stroganova et al., 1998; Alexandrovskaya and Alexandrovskiy, 2000). Another point that is peculiar to urban soil sampling is the accessibility of the soil. Private gardens, industrial complexes, military, and other restricted areas are examples of soils that cannot be sampled or can be sampled only by specific consent of the relevant authority. In other cases it is forbidden to sample soils of zones that have a particular landscape or archaeological value and other problems may derive from the city itself as traffic limitations can impede the access to certain areas or heavy traffic can greatly slow down the operation. These difficulties often imply that systematic sampling cannot be adopted, except on large homogeneous areas. Rather, the sampling strategy must be based on expert evaluation and subsequently adapted at the moment of field campaign.

For the present study four land uses were identified and sampled in each city: parks and open spaces (PO), riverbanks (RB), ornamental gardens (OG), and roadsides (RD). A total of 803 samples were collected in the three cities. There were 123 sampling sites in Torino, 154 in Sevilla, and 130 in Ljubljana (Fig. 1).


Figure 1
View larger version (95K):
[in this window]
[in a new window]

  		Fig. 1. Sampling locations in the three cities.

 
Soil samples were taken at depths of 0 to 10 cm (SF) and 10 to 20 cm (SB) with spades. The spades were washed with deionized water and wiped dry with paper towels after each use. At each sampling site composite samples were obtained by mixing a minimum of four subsamples from the area surrounding the site after removal of the herbaceous cover. Plant roots and large debris were also discarded. About 3 kg of soil were taken at each location and stored in plastic bags. Coordinates of the sampling points were recorded by means of a portable GPS device.

Analytical Methods
Samples were air-dried, gently crushed, and sieved to <2 mm particle size with plastic sieves to avoid metal contamination (ISO, 2006). A portion of each sample was further ground to 0.15 mm for aqua regia (HCl/HNO3, 3:1 solution) digestion (ISO, 1995a). Particle size distribution was determined by sedimentation and sieving (ISO, 1998). The pH was determined in a CaCl2 solution, 1:5 soil/solution ratio (ISO, 2005), electrical conductivity in a H2O solution, 1:5 soil/solution ratio (ISO, 1994), organic carbon by elemental analyzer (ISO, 1995b) in Torino, and by K2Cr2O7 oxidation in Sevilla and Ljubljana. An internal reference material was used to ensure that the two methods gave comparable results. Cation exchange capacity (CEC) was determined with BaCl2 at pH 8.1 (ISO, 1995d) in Torino and by exchange with ammonium acetate at pH 7 in Ljubljana and Sevilla. The latter method has been reported to overestimate the exchange capacity of acid soils (Sumner and Miller, 1996) but this is seldom the case for the soils under study. Carbonates were measured using a volumetric method (ISO, 1995c) and expressed as calcium carbonate equivalent (CCE). Due to technical problems, it was not possible to measure CCE in Ljubljana SB samples. The aqua regia extracts were analyzed for Cu, Cr, Ni, Pb, and Zn by flame atomic absorption spectrometry (FAAS) (Torino and Ljubljana) and inductively coupled plasma-optical emission spectrometry (ICP–OES) (Sevilla). Duplicates were made for all samples and results accepted when the coefficient of variation was within 5%. A blank and the CRM (certified reference material) 141 R (Community Bureau of Reference, Geel, Belgium) were included in each batch of analyses for quality control of metal measurements. Results were considered satisfactory when within a range of ±10% from the certified value. An internal reference material was also included in each batch of analyses for quality control and inter-laboratory comparison of soil physical-chemical analyses.

Statistical Analysis and Data Treatment
Statistical analysis was performed using the software MINITAB 13 (Minitab, 2002) for Windows and SPSS 12 (SPSS, 2004) for Windows. The data were geographically managed and processed with the GIS software Arcview 3.2 (ESRI, 1998).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Soil Properties
The mean values, medians, standard deviations, minimum, maximum, and number of samples of soil data are reported in Table 2 for the two sampling depths SF (0 to 10 cm) and SB (10 to 20 cm) for Ljubljana (LJU), Sevilla (SEV), and Torino (TOR).


View this table:
[in this window]
[in a new window]

  		Table 2. Descriptive statistics of soil general properties in the three cities at the two sampling depths: 0 to 10 cm (SF) and 10 to 20 cm (SB).

 
The mean pH values are around neutrality in all the three cities and for both depths. Values range from 4.2 (TOR) to 7.9 (TOR) with 90% of the samples within the range 6.6–7.5.

Despite the common use of de-icing salts in the cities of Ljubljana and Torino and the proximity to the ocean and the Mediterranean climate of Sevilla that would suggest some accumulation of soluble salts, electrical conductivity is generally low.

The soils show a diffuse presence of carbonates, with mean values of 172 g kg–1 (LJU SF), 203 to 210 g kg–1 (SEV, SF-SB), and 48 to 50 g kg–1 (TOR, SF-SB). In the case of Sevilla (Clemente and Paneque, 1974) and Ljubljana this is probably due to the carbonatic parent material over which the city soils have developed but in Torino can be partly due to the inclusion of extraneous, carbonaceous materials such as debris of construction, concrete, and bricks (Bullock and Gregory, 1991; Jim, 1998; Alexandrovskaya and Alexandrovskiy, 2000; Biasioli et al., 2006).

The organic carbon (OC) content ranges from 1.0 (TOR) to 89.0 g kg–1 (LJU) with mean values of 49.7 to 37.0 g kg–1 (LJU, SF-SB), 19.1 to 11.9 g kg–1 (SEV), and 20.6 to 14.6 g kg–1 (TOR). The OC contents are relatively low in TOR and SEV and higher in LJU reflecting differences in climate (Table 1) and vegetation cover. Many soils in SEV and TOR were in fact bare at the time of sampling while most soils had a grass cover in LJU.

Total N has mean values of 3.7 to 2.8 g kg–1 (LJU, SF-SB), 1.8 to 1.2 g kg–1 (SEV), and 1.6 to 1.1 g kg–1 (TOR), ranging from a minimum of 0.2 g kg–1 (TOR) to a maximum of 9.6 g kg–1 (SEV), in line with the OC content. The C/N ratio is similar in the three cities, being on average 13 in LJU, 11 in SEV, and 14 in TOR. Values are very variable ranging from as low as 1 (SEV) to 37 (TOR) indicating that C cycling is often perturbed. Alterations in the C cycle are common in urban soils and were reported by Pouyat et al. (2006) in a study on carbon storage in six cities in the United States and by Norra et al. (2005), who analyzed 13C and 15N distribution in urban soils in Karlsruhe, Germany.

Cation exchange capacity ranges from a minimum of 1 cmol kg–1 (TOR) to a maximum of 41 cmol kg–1 (LJU), with mean values of 29 to 28 cmol kg–1 (LJU, SF-SB), 15 to 14 cmol kg–1 (SEV), and 11 to 10 cmol kg–1 (TOR), reflecting the differences in OC content and particle-size distribution of the soils from the three cities.

Figure 2 shows the particle size distribution of all the samples. Torino soils are grouped in the coarser fractions, while LJU soils have a higher silt content, and SEV samples have a more variable texture.


Figure 2
View larger version (41K):
[in this window]
[in a new window]

  		Fig. 2. Particle size distribution (PSD) for all the samples.

 
Potentially Toxic Elements
Table 3 reports the mean values, medians, standard deviations, minimum, maximum and number of samples for Cr, Cu, Ni, Pb, and Zn pseudototal content (aqua regia-extractable) in the three cities and at the two sampling depths.


View this table:
[in this window]
[in a new window]

  		Table 3. Descriptive statistics of potentially toxic elements (PTE) concentrations in the three cities. Results are presented for the SF (0 to 10 cm) and SB (10 to 20 cm) layers.

 
The boxplots of the five elements analyzed for the three cities are reported in Fig. 3.


Figure 3
View larger version (20K):
[in this window]
[in a new window]

  		Fig. 3. Boxplots of the potentially toxic elements (PTE) contents (mg kg–1) in the three cities. Boxes represent interquartile ranges (IQR) and stars are outliers, calculated as the values below Q1–1.5·IQR and above Q3+1.5·IQR. Whiskers are range excluding outliers. Medians are marked in each box with a line, means with a dot.

 
The data in Table 3 and the boxplots highlight a diffuse contamination of these urban soils with PTE. The main sources of pollution for PTE in urban settlements are industrial emissions, traffic, burning of fossil fuels, and wastes from industrial and residential activities (Bergback et al., 2001; Möller et al., 2005). Urban soils, as a consequence, often show high PTE concentrations.

The soils of Torino show the highest average values for all the PTE, most probably because of the main industrial activity, car manufacturing, which emits metallic contaminants to the environment.

All three cities show high values of Pb and Zn whose most usual source is traffic. The high Cr and Ni concentrations found in Torino soils can be partly attributed to the alteration of Cr-Ni-serpentinites that are common in the alluvial deposits over which the city has developed (Alloway, 1995; Lombini et al., 1998; Biasioli et al., 2006). This is substantiated by the Pearson correlation coefficients obtained for the five metals in soils from the three cities (Table 4). Copper and zinc and copper and lead are significantly correlated. The association of copper, lead and zinc in urban soils has already been observed and these elements have been identified as typical "urban" metals by some authors (De Miguel et al., 1998; Madrid et al., 2002; Manta et al., 2002; Biasioli et al., 2006). Nickel and chromium are also correlated indicating their possible common lithogenic origin. The association between Ni and Cr, and their common, refractory nature in the soils of Torino has also been confirmed by sequential extraction procedures (Davidson et al., 2006).


View this table:
[in this window]
[in a new window]

  		Table 4. Correlation matrix for potentially toxic elements (PTE) in all urban samples; cells show the Pearson correlation coefficient and the corresponding p value (in parentheses).

 
In an attempt to corroborate the hypothesis of a common origin of some PTE, a cluster analysis was applied to the standardized PTE concentrations and soil properties using Ward's method, with Euclidean distances as the criterion for forming factors. This type of analysis was efficiently applied in the investigation of PTE sources in urban dusts (Yongming et al., 2006) and in urban soils (Lee et al., 2005). The results of the cluster analysis reported in Fig. 4 group zinc, copper, and lead in a cluster and nickel and chromium in another, confirming the correlation results. When other soil parameters are added, sand appears to be associated with the latter elements, further substantiating their probable lithogenic origin, while other soil parameters are regrouped in clusters at low similarity values.


Figure 4
View larger version (10K):
[in this window]
[in a new window]

  		Fig. 4. Dendrogram for potentially toxic elements (PTE), pH, cation exchange capacity (CEC), organic carbon (OC), clay, and sand in all samples.

 
Table 5 presents the national or regional threshold concentrations for remediation of soils. The comparison of the limits adopted or proposed in the three countries highlights the lack of common criteria for the definition of thresholds for soil contaminants (e.g., some limits are a function of different land uses, some are a function of the pH, some are ranges, and some unique values.).


View this table:
[in this window]
[in a new window]

  		Table 5. Maximum acceptable concentrations established in the three countries of the study for the potentially toxic elements (PTE) in soils.

 
The mean PTE content in Sevilla is always below the regional limits, with 7% of the samples above the limit for Cu, none for Cr and Ni, 11% for Pb, and 2% for Zn. In Ljubljana the limit is exceeded by 18% of the samples for Cu, 6% for Cr, 4% for Ni, 45% for Pb, and 20% for Zn. Average PTE concentrations in Torino soils always exceed national limits except for Cu; 19% of the soils exceed the limit for this element, 38% for Cr, 87% for Ni, and 43% for Pb and Zn.

Apart from the particular case of Cr and Ni in Torino, Pb and Zn seem to be the PTE that most often exceed the limits, despite the differences in their values.

The soils of the urban areas of Ljubljana, Sevilla, and Torino appear to be spatially very variable within each city due to the continuous direct and indirect anthropic disturbance.

Variation with Depth
In general pollutants coming from atmospheric deposition are expected to accumulate in the surface layers and their concentrations to decrease with depth.

A paired t test was conducted on the differences between the two sampling depths (0 to 10 and 10 to 20 cm), assuming that the two layers belong to the same population as the null hypothesis. The test was conducted under the assumption that variances were unequal as shown by a Levene's test for homoscedasticity, using the 95% confidence interval.

Results of the test for the soil properties and the PTE in the two layers of the three cities are presented in Table 6.


View this table:
[in this window]
[in a new window]

  		Table 6. Paired t test of soil properties and potentially toxic elements (PTE) contents in surface against deeper layer.

 
For all the five PTE H0 could not be rejected, i.e., metal concentrations in SF and SB samples can be considered as belonging to the same populations or, in other terms, metals do not seem to accumulate in the surface layer.

Other soil parameters such as pH (TOR and LJU), sand (SEV), silt (TOR and LJU), clay, OC, total N, and CEC (LJU) were found to be different in the two layers. In particular, pH increases with depth in Torino and Ljubljana. Clay increases with depth, while organic C and CEC were found to be higher in the surface layer. The higher CEC in the surface layer can be attributed to the higher OC content.

The absence of trends of contaminants with depth might reflect the mixing and additions that most of the soils of these cities undergo. On the other hand, some general soil properties seem not to be affected by these same disturbances. Organic carbon, for example, shows a certain degree of resilience and readjusts relatively rapidly. Metal contamination therefore becomes a useful indicator of human interference with urban soil functions.

Land Use Influence
As previously described, four different land uses were sampled in the three cities, parks and open spaces (PO), riverbanks (RB), ornamental gardens (OG), and roadsides (RD). The selection of these land uses was based on the assumption that they might be differently influenced by urban activities (e.g., maintenance practices for OG, traffic for RD etc.). It is to be mentioned that the field identification of a unique land use for a sampling site was not always possible as uses often overlap. For example, ornamental gardens can be found along roads and parks are often located close to rivers.

The boxplots of PTE concentrations in the three cities according to land use are presented in Fig. 5.


Figure 5
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 5. Boxplots of the potentially toxic elements (PTE) contents (mg kg–1) in the three cities according to land uses. Boxes represent interquartile ranges (IQR) and stars are outliers, calculated as the values below Q1–1.5·IQR and above Q3+1.5·IQR. Whiskers are range excluding outliers. Medians are marked in each box with a line, means with a dot. Numbers are median values.

 
The overall distribution of the samples into the different land uses was 460 for the PO class, 214 for RD, 80 for OG, and 60 for RB. Uneven distribution in the classes and overlapping interquartile ranges prevent consistent comparisons but some trends can be observed.

Copper, Pb, and Zn appear to be similarly influenced by land use, as their median values always follow the same order RB < PO < RD < OG. The low PTE contents observed in RB soils might be due to the general young age of these surfaces while soils of parks (PO) are usually located further from the emission sources and generally less subjected to anthropogenic disturbance such as inclusion of extraneous materials, mixing, etc. Higher values in RD and OG can be partly explained by the proximity to sources for RD and by maintenance practices in OG.

Chromium and Ni appear to be particularly concentrated in the RD soils. This suggests that although the lithogenic origin is prevalent, anthropogenic contributions cannot be ruled out. This is further confirmed by the observation that Cr and Ni are not prevalent in PO soils where the proportion of TOR samples is similar to RD.

Principal Component Analysis
The associations found within metal contents and between metals and soil properties (Table 4 and Fig. 4) were further investigated by principal component analysis (PCA). This has proved to be a useful tool to study the origin of PTE in soils (Boruvka et al., 2005; Chen et al., 2005; Ruiz-Cortés et al., 2005). Table 7 reports the results of a Varimax-rotated PCA conducted on all the samples for the variables OC, sand, Cu, Cr, Ni, Pb, and Zn. Inclusion of other variables did not improve significantly the proportion of variance accounted for. The table reports the correlation coefficients between each variable and the first two components (those with eigenvalues greater than unity), accounting for about 74% of the total variance. The first component (49.0% of the variance) shows the association between Cr, Ni, and the sand fraction. The second component (24.8% of the variance) includes Cu, Pb, and Zn. These two components are in line with the outcome of the cluster analysis, confirming the common origin of Cr and Ni on the one hand and of Cu, Pb, and Zn on the other.


View this table:
[in this window]
[in a new window]

  		Table 7. Correlation coefficients between potentially toxic elements, organic C (OC), sand, and the first two principal components after Varimax rotation. Eigenvalues and percentage of variance these account for are given. Components with absolute values ≥0.600 are printed in italic. Communalities show the proportion of each variable accounted for by the factors.

 
Figure 6 displays the soils of the three cities represented on the plane of the first two components. The city of Torino appears to be clearly separated from Ljubljana and Sevilla along component 1 (Cr, Ni, and sand), confirming the peculiarities of the substrate of this city for these parameters. Ljubljana and Sevilla cannot be distinguished for component 1 and none of the three cities can be identified along component 2 (Cu, Pb, Zn, and OC). Results tend to confirm the hypothesis that, apart from Ni and Cr in Torino soils, the three cities cannot be distinguished from one another when considering PTE contents.


Figure 6
View larger version (26K):
[in this window]
[in a new window]

  		Fig. 6. Scatter of the sampling points on the plane of the first two principal components of the principal component analysis (PCA) after Varimax rotation. LJU, Ljubljana; SEV, Sevilla; TOR, Torino.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The three city ecosystems have been demonstrated to strongly influence the quality of their soils.

The soils show a diffuse contamination with PTE, often above the legislation limits. Copper, Pb, and Zn were confirmed to be common tracers of urban soil contamination, showing similar trends in the three cities. On the other hand, especially in Torino, Cr and Ni were found to be associated, reflecting their probable natural origin.

No depth-dependant trends for contaminants were observed in any city, probably due to the continuous mixing of urban soils. The anthropogenic influence on other soil properties seems to be less pronounced, as many soil parameters appear to be more resilient.

The fact that no vertical trends were observed for PTE concentrations shows the importance of carefully considering depth when sampling urban soils. Data suggest that, for the investigation of PTE surface contamination in urban soils, the sampling procedures could be limited to the upper layer (0 to 10 cm), or to a generic 0- to 20-cm layer.

Land uses appear to influence the distribution of typically "urban" PTE, such as Pb, Zn, and Cu, in relation to the distance of the site from contaminant sources and the influence of maintenance practices. Soils on RD show an addition of Cr and Ni on top of the lithogenic contribution.

Although a common sampling strategy is essential for this kind of study, the high spatial variability of soil properties of Ljubljana, Sevilla, and Torino confirmed the necessity of a city-specific sampling scheme.

The use of certified reference materials and of an internal reference material taken from an urban soil would be strongly recommended in view of the differences in the composition of the soil matrix.

The comparison of three southern European cities revealed that, regardless of the geographical location, climate, and size of the urban areas, the "urban" factor (regrouping the influence of the urban environment in terms of type and intensity of emission and human influence on soils) is dominant in determining urban soil contamination with PTE, which is common to, and similarly distributed within, the three cities studied.


   ACKNOWLEDGMENTS
 
This work was conducted with the financial support of the European project URBSOIL, under contract no. EVK4-CT-2001-00053. The authors are indebted to Dr. C.M. Davidson for her valuable comments on an earlier version of this manuscript. The help of C. Martini, R. Reinoso, E. Ruiz-Cortés, T. Pacnik, and S. Gogic with the sampling and analytical work is gratefully acknowledged.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Biasioli, M.
Right arrow Articles by Ajmone-Marsan, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Biasioli, M.
Right arrow Articles by Ajmone-Marsan, F.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Biasioli, M.
Right arrow Articles by Ajmone-Marsan, F.
Related Collections
Right arrow Toxic Trace Metals
Right arrow Urban Soils
Right arrow Heavy Metals
Right arrow Soil Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome