A soil sampling strategy for mapping trace element concentrations in a test area

A sampling strategy, based on previous knowledge of a test area, was compared with a Reference Sampling in the frame of the EU-funded CEEM project. The aim was a comparison of sampling strategies for the drafting of concentration maps for a set of trace elements (Cd, Pb, Cu and Zn), with three-dimension information that can be used for risk assessment. The main constraint was that only 15 soil samples were to be analysed. The test area (in Dornach, Switzerland) included different soil types and land uses (forest, arable land and permanent grasslands). The sampling strategy proposed defined strata on the basis of land use, then applying a systematic regular sampling within each stratum. A composite sample was formed from each individual sample. In forests, H and Ah horizons were taken separately. In arable lands, a depth within the ploughed layer (0-15 cm) was analysed. In grasslands, samples were taken at two depths (0-10 and 10-20 cm) to provide information on the vertical variability. The comparison of the results obtained with those derived from the Reference Sampling allowed us to estimate if the information obtained from the proposed sampling strategy was sufficiently representative for an assessment of the level of elements considered. It was concluded that recommendations on sampling strategy depend on the interaction pattern of the element. In this case, for mobile trace elements, such as Cd, the information derived from sampling of two soil layers was not as important as that for the rest of metals studied that accumulated in the upper layer. If the H horizons in forests was not considered this led to unreliable predictions of the concentration levels, since the concentrations in H and Ah horizons differed irrespective of whether they were expressed in mg kg(-1) (the H horizon was enriched in those trace elements, except for Cd) and in mg m(-2) (the low bulk density of the H horizon led to a lower trace element level than in Ah horizon). Finally, in undisturbed systems such as grasslands, sampling 0-10 cm instead of 0-20 cm is recommended to prevent dilution of the concentration of trace elements.     

Quantifying uncertainty of the reference sampling procedure used at Dornach under different soil conditions

The reference sampling procedure, used in the CEEM soil project on a single test area at Dornach, had been applied under different soil conditions in a sampling proficiency test and in the Swiss national soil-monitoring network related to pollution (NABO). Methods of analysis of variance (ANOVA) were used to quantify sampling and analytical uncertainty. The soil sampling procedure and the chemical analysis were considered as two parts of the same measurement process and the so-called 'top-down' approach was used to quantify their combined contribution to the uncertainty. By this approach a systematic error of one sampler or of one laboratory becomes a random error when assessed as part of a multi-sampler or multi-laboratory comparison.     

Description of the test area and reference sampling at Dornach

The selected test area for the comparative evaluation of European methods on sampling and sample preparation of soils (CEEM Soil) at Dornach near Basel (Switzerland) is located at approximately 400 m a.S.l. on the north-west rolling footslope of the Jura mountain chain that has a temperate climate. The area is known to be airborne polluted by emissions of a non-ferrous metal recycling plant. The geology is dominated by Jurassic limestone and (decarbonated) loess loam. The land use is deciduous forest (beech, oak) in the upper part and traditional cherry orchards with permanent grassland which have mostly been converted into arable land. The detailed soil survey distinguishes four different soil mapping units in the test area of 0.61 ha. The reference sampling was performed in a 190-m-long and 20-40 m large transect area following four devices: (1) composite samples (0-10 cm and 0-20 cm), each from 25 stratified single samples of 61 squares of 10 x 10 m; (2) hypotheses and soil horizon-based composite samples at nine locations; (3) horizon-related samples from the four described soil profiles representing the mapping units; and (4) three soil cores with 5-cm interval samples taken in the major land use units. Altogether 301 reference soil samples were taken.     

Quantitative evaluation of the CEEM soil sampling intercomparison

The aim of the CEEM soil project was to compare and to test the soil sampling and sample preparation guidelines used in the member states of the European Union and Switzerland for investigations of background and large-scale contamination of soils, soil monitoring and environmental risk assessments. The results of the comparative evaluation of the sampling guidelines demonstrated that, in soil contamination studies carried out with different sampling strategies and methods, comparable results can hardly be expected. Therefore, a reference database (RDB) was established by the organisers, which acted as a basis for the quantitative comparison of the participants' results. The detected deviations were related to the methodological details of the individual strategies. The comparative evaluation concept consisted of three steps: The first step was a comparison of the participants' samples (which were both centrally and individually analysed) between each other, as well as with the reference data base (RDB) and some given soil quality standards on the level of concentrations present. The comparison was made using the example of the metals cadmium, copper, lead and zinc. As a second step, the absolute and relative deviations between the reference database and the participants' results (both centrally analysed under repeatability conditions) were calculated. The comparability of the samples with the RDB was categorised on four levels. Methods of exploratory statistical analysis were applied to estimate the differential method bias among the participants. The levels of error caused by sampling and sample preparation were compared with those caused by the analytical procedures. As a third step, the methodological profiles of the participants were compiled to concisely describe the different procedures used. They were related to the results to find out the main factors leading to their incomparability. The outcome of this evaluation process was a list of strategies and methods, which are problematic with respect to comparability, and should be standardised and/or specified in order to arrive at representative and comparable results in soil contamination studies throughout Europe. Pre-normative recommendations for harmonising European soil sampling guidelines and standard operating procedures have been outlined in Wagner G, Desules A, Muntau H, Theocharopoulos S. Comparative Evaluation of European Methods for Sampling and Sample Preparation of Soils for Inorganic Analysis (CEEM Soil). Final Report of the Contract SMT4-CT96-2085, Sci Total Environ 2001;264:181-186. Wagner G, Desaules A, Munatu H. Theocharopolous S, Quevauvaller Ph. Suggestions for harmonising sampling and sample pre-treatment procedures and improving quality assurance in pre-analytical steps of soil contamination studies. Paper 1.7 Sci Total Environ 2001b;264:103-118.     

Analytical aspects of the CEEM soil project

In the past, exercises aiming at an assessment of data uncertainty in environmental analysis were usually restricted to the analysis step, while sampling and pre-analytical sample treatment was largely ignored. Collaborative studies on the quantification of sampling errors require, besides a suitable and well characterized test site, the availability of a reference laboratory for the analysis of all of the samples taken in the context of the study by all participants and also test methods which do not contribute large and variable uncertainties due to long and complex analytical methodologies. Here we summarize the major analytical aspects of a European project on the identification and quantification of sampling influences on the determination of lead, cadmium, copper and zinc in soil. The participant group included the leading soil analysis laboratories in Europe; the test site at Dornach (CH) was well suited for the purpose and showed high metal gradients and differentiated land use. The analytical methods (wavelength-dispersive X-ray fluorescence spectrometry and solid-state Zeeman AAS) used in the study showed stable performance characteristics within the confidence interval of the certified reference materials used for the measurement quality control over the entire project period. Additionally, double-blind tests on split samples showed agreement of data in very narrow limits thus demonstrating the reliability of the reference database.     

A flexible methodology for the characterisation of soils: a case study of the heavy metal status of a site at Dornach

The sampling scheme devised to study the site at Dornach was successful in identifying most of the features and element distributions in the site using only 15 composite samples. The scheme was based on dividing the area into six individual sampling units each of which represented a different combination of land use and soil type. The three-dimensional information was obtained by taking samples at different depths, either on a fixed depth basis for cultivated soils or according to soil horizons for undisturbed soils. In each sampling unit the composite samples were bulked from 12 individual samples. The sampling point for each individual sample was chosen according to a stratified random plan. Key features of the scheme were the flexibility which required the use of experienced personnel and the fact that the sole task of the sampling exercise was to characterise the site with no preconceptions of possible outcomes. The results showed that the high concentrations of Cd in the soil probably resulted from the high concentrations of Cd in the underlying limestone and not primarily from atmospheric inputs. Higher levels of copper were found in soils close to fruit trees probably as a result of using sprays on the trees. The concentrations of Pb and Zn were relatively uniform throughout the site although there was some evidence of increased Zn concentrations with depth analogous to the Cd distributions. The fact that these findings were in agreement with the known characteristics of the site indicates the successful application of the chosen methodology in this study.     

Objectives, concept and design of the CEEM soil project

Within the member countries of the European Union, different strategies and guidelines for the sampling and the preparation of soils are used to investigate soil contamination. As a consequence, there is reasonable doubt that the results gained by applying these different strategies and methods are comparable. In general, the relative contribution of sampling and sample preparation to the total uncertainty in soil contamination studies is a subject of speculation than of knowledge. For this purpose, the first European soil sampling inter-comparison test was organised as a project under the Standards, Measurement and Testing Programme of the European Commission. The aim of this project was to analyse and to compare the existing guidelines, the degree and the state of standardisation in this field in the European countries, and to compare their results when applied on the same test area under comparable conditions by representative sampling teams. The results of the project in general demonstrate that sampling and sample preparation errors reach about the same order of magnitude as errors caused in the chemical analysis. However, there are great differences depending on the elements investigated and their concentration ranges, as well as on pedological and land-use conditions in the test area. Consequently, sampling and sample preparation require the same attentiveness and equivalent measures for quality assurance (QA) and quality control (QC) as the chemical analysis, which has not been the case up to now. Insufficient comparability of the results is illustrated by the fact that different participants arrive at different conclusions for up to 61% of tested soil quality criteria (e.g. threshold levels). The methods and results of the project are presented in detail in the six following papers within this volume. The following 13 papers present some key studies and specific contributions of the participants, as well as synoptical papers of the invited experts in standardisation and accreditation.     

Identifying Land Contaminated by Chlorinated Hydrocarbon Solvents

This paper provides a method of identifying land contaminated with chlorinated hydrocarbon solvents by sampling and analysing soil gas. Relationships are presented which enable levels of soil gas contamination to be related to levels of soil and groundwater pollution. Two case studies are presented. Soil gas surveying of an industrial site with severe groundwater pollution demonstrated that standards for contaminated land are insufficient to protect groundwater. Solvent levels in groundwater were one to two orders of magnitude higher than UK permitted concentrations in drinking water, whereas soil contamination values did not reach polluted levels. Surveying of a site occupied by a disused fire station showed significant differences in patterns of soil gas and groundwater pollution. This is explained by direct pollution of groundwater though the well at the site and/or possible flow of polluted groundwater to the site from adjacent areas.     

European soil sampling guidelines for soil pollution studies

The soil sampling guidelines used in European countries (ESSG), as kindly provided by the national institutions which participated in the project, have been recorded, studied, evaluated and presented in this paper. The aim has been to ascertain what soil sampling guidelines exist in Europe; to detect similarities and differences (comparable results), advantages and deficiencies; to identify incompatible strategies and evaluate how methodologies might affect data quality; to investigate sources of deviations or uncertainties; to improve comparability and representativeness of soil sampling; to investigate the need for harmonised sampling guidelines; and to develop suggestions for standard operating procedures (SOP). Soil sampling guidelines throughout Europe differ as to whether they are applied by law, or used throughout the country. In some countries these are ISO/DIS related or based (ISO 10381-1, 1995; ISO 10381-2, 1995), or are produced by a scientific society or a standardisation body. As far as sampling strategy is concerned, not all sampling guidelines clearly describe the sampling scale, the specifications for contamination risk precautions, the sampling plan and protocol structure and the pre-analysis treatment of the soil samples. The purpose for sampling, in descending order of frequency, is soil pollution, soil fertilisation, general soil monitoring, background risk assessment, or else it is not specified. The majority of countries do not sample the top organic matter separately. Sampling depth is either related to the morphogenetic horizon or to ad hoc sampling depth, which is not specified in all cases. They suggest mass- and volume-related soil sampling, while the sampling pattern is not presented in all national guidelines. The criteria for area, site, unit, sub-unit, and point selection are mainly based on pedology and land use, following the history and pre-screening information or geology, or is site related. Some guidelines suggest the division of sampling units into sub-units. The sampling pattern is mainly grid sampling, grid and random sampling, or not mentioned. Sampling density inside the sampling unit either varies greatly or it is not mentioned, while the size of the sampling unit varies widely. Most guidelines require the collection of composite instead of simple samples, while some prefer sampling soil profiles. In the European SSG many technical details and steps are either not defined or vary, while in the pre-analysis treatment quality assurance (QA) and quality control (QC) approaches are used either both in the lab and in the field, or only in the field, or are not mentioned. The common points and the points in which harmonisation could be started or achieved are discussed.     

Archaeological recording and chemical stratigraphy applied to contaminated land studies

The method used by archaeologists for excavation and recording of the stratigraphic evidence, within trenches with or without archaeological remains, can potentially be useful to contaminated land consultants (CLCs). The implementation of archaeological practice in contaminated land assessments (CLAs) is not meant to be an exercise in data overkill; neither should it increase costs. Rather, we suggest, that if the excavation and recording, by a trained archaeologist, of the stratigraphy is followed by in-situ chemical characterisation then it is possible that much uncertainty associated with current field sampling practices, may be removed. This is because built into the chemical stratigraphy is the temporal and spatial relationship between different parts of the site reflecting the logic behind the distribution of contamination.An archaeological recording with chemical stratigraphy approach to sampling may possibly provide ‘one method fits all’ for potentially contaminated land sites (CLSs), just as archaeological characterisation of the stratigraphic record provides ‘one method fits all’ for all archaeological sites irrespective of period (prehistoric to modern) or type (rural, urban or industrial). We also suggest that there may be practical and financial benefits to be gained by pulling together expertise and resources stemming from different disciplines, not simply at the assessment phase, but also subsequent phases, in contaminated land improvement.     

Suitability of the OCDE tests to estimate contamination with 2,4-dichlorophenol of soils from Galicia (NW Spain)

The objective of the present study was to verify whether the generic reference levels (GRL) for soils contaminated with 2,4-dichlorophenol (2,4 DCP), established by Spanish legislation and published in the Real Decreto 9/2005, are accurate for Galician soils. For this, the surface horizons of seven soils under different types of land use were experimentally contaminated with different doses (between 0 and 10,000 times the GRL) of 2,4 DCP, and were then were subject to OECD toxicity test numbers 208 (root emergence and elongation) and 216 (soil nitrogen mineralization). The results obtained for the nitrogen mineralization test were difficult to interpret because they varied among soils, whereas the results of the root germination and elongation test were more coherent -- the values decreased with increasing doses of contaminant added to the soil. The results suggest that the root elongation test reflects the effect of this contaminant more clearly than the soil nitrogen mineralization test. Nevertheless, considering that in the lowest quality soils (i.e. agricultural soils containing a very low level of organic matter) contaminant doses of up to 1000 times the GRL did not affect root germination and elongation, it is clear that the GRL indicated in the relevant legislation are very low for the soils under study and that the threshold should be established taking into account the soil characteristics.     

Assessment of a remediation technique using the replacement of contaminated soils in kitchen gardens nearby a former lead smelter in Northern France

Vegetables cultivated in kitchen gardens that are strongly contaminated by heavy metals (Pb, Cd) may represent to consumers a means of exposure to these metals. This exposure is more problematic for those families that include a large quantity of home-grown vegetables in their diet. Researchers have shown that the majority of vegetables produced in kitchen gardens in the vicinity of the Metaleurop Nord smelter (Northern France) do not conform to European regulations. This study was carried out in three of these kitchen gardens. The concentrations of Cd and Pb in the topsoils were up to 24 and 3300 mg kg(-1) respectively. The method consisted of delineating a surface area of about 50 to 100 m(2) for each garden, then removing the contaminated soil and replacing it with a clean one. Seven species of vegetables were cultivated from 2003 to 2005 in the original contaminated soils and the remediated ones. The data showed a clear improvement of the quality of the vegetables cultivated in remediated soils, although 17% of them were still over the European legislative limits for foodstuffs. This suggested that there was a foliar contamination due to contaminated dust fallout coming from the closed smelter site and the adjacent polluted soils. In addition, the measurement of the Cd and Pb concentrations in the dust fallout showed that the substantial rise in metal concentrations in the remediated soil was not only due to atmospheric fallout. These results raise questions about possible technical, economic and sociological problems associated with this kind of remediation.     

Uncertainty of measurement or of mean value for the reliable classification of contaminated land

Classification of contaminated land is important for risk assessment and so it is vital to understand and quantify all of the uncertainties that are involved in the assessment of contaminated land. This paper uses a case study to compare two methods for assessing the uncertainty in site investigations (uncertainty of individual measurements, including that from sampling, and uncertainty of the mean value of all measurements within an area) and how the different methods affect the decisions made about a site. Using the ‘uncertainty of the mean value’ there is shown to be no significant possibility of ‘significant harm’ under UK guidance at one particular test site, but if you consider the ‘uncertainty of the measurements’ a significant proportion (50%) of the site is shown to be possibly contaminated. This raises doubts as to whether the current method using ‘uncertainty of the mean’ is sufficiently robust, and suggests that ‘uncertainty of measurement’ information may be preferable, or at least beneficial when used in conjunction.     

Deriving ecological risk-based soil quality values in the Basque Country

Most industrialised countries are currently drawing up their procedures to deal with contaminated land. In the Basque Country generic soil quality values applicable to different land uses have been drawn up for many common soil inorganic and organic pollutants, to assist in the site assessment process. In this paper we present those values aimed at protecting ecosystems as the main target at risk. Ecological risk was defined on the basis of effects on the structure of the soil invertebrate community. Ecotoxicological data from laboratory tests were used in combination with a theoretical extrapolation method. The following groups of compounds were assayed: heavy metals, cyanides, aromatic compounds, polycyclic aromatic hydrocarbons, HCH isomers, chlorinated organic compounds, and mineral oil. A specific toxicity database was created. The reliability of the values obtained as well as the limitations of the derivation process are discussed.     

Introducing a method of human health risk evaluation for planning and soil quality management of heavy metal-polluted soils—An example from Grugliasco (Italy)

European governmental institutions, as well as local communities, have recently become aware of the threats arising from soil pollution for the welfare of the population. Humans contact with soils is more likely in urban than in rural areas, and is strongly dependent on land use. Spatial planning and land management may have important impacts on the potential transfer of pollutants from contaminated soils to humans. In the present study, we propose a land use-based method for the evaluation of human health risks arising from heavy metal-contaminated urban soils, addressing in particular the influence of planning measures and land use changes on such risks. The method accounts for the role of the bioavailability of soil metals as a key factor in health risk assessment. In order to increase method applicability, a step-by-step procedure and a calculation tool were elaborated. The method can be used to identify areas in which the current or planned land use is associated with unacceptable health risks and to optimise the allocation of a certain land use to areas that are well suited and where the risks are minimal. A risk index is calculated for the area, taking the land use into consideration, as the sum of the risks from different exposure pathways and different heavy metals. For those areas where risk is identified as unacceptable, alternative planning or management options should be defined to achieve a maximal risk reduction in a cost-effective way. The method is illustrated using the Italian municipality, Grugliasco.     

Introducing a method of human health risk evaluation for planning and soil quality management of heavy metal-polluted soils—An example from Grugliasco (Italy)

European governmental institutions, as well as local communities, have recently become aware of the threats arising from soil pollution for the welfare of the population. Humans contact with soils is more likely in urban than in rural areas, and is strongly dependent on land use. Spatial planning and land management may have important impacts on the potential transfer of pollutants from contaminated soils to humans. In the present study, we propose a land use-based method for the evaluation of human health risks arising from heavy metal-contaminated urban soils, addressing in particular the influence of planning measures and land use changes on such risks. The method accounts for the role of the bioavailability of soil metals as a key factor in health risk assessment. In order to increase method applicability, a step-by-step procedure and a calculation tool were elaborated. The method can be used to identify areas in which the current or planned land use is associated with unacceptable health risks and to optimise the allocation of a certain land use to areas that are well suited and where the risks are minimal. A risk index is calculated for the area, taking the land use into consideration, as the sum of the risks from different exposure pathways and different heavy metals. For those areas where risk is identified as unacceptable, alternative planning or management options should be defined to achieve a maximal risk reduction in a cost-effective way. The method is illustrated using the Italian municipality, Grugliasco.     

Development of agro-environmental scenarios to support pesticide risk assessment in Europe

This paper describes work carried out within the EU-funded FOOTPRINT project to characterize the diversity of European agricultural and environmental conditions with respect to parameters which most influence the environmental fate of pesticides. Pan-European datasets for soils, climate, land cover and cropping were intersected, using GIS, to identify the full range of unique combinations of climate, soil and crop types which characterize European agriculture. The resulting FOOTPRINT European agro-environmental dataset constitutes a large number of polygons (approximately 1,700,000) with attribute data files for i) area fractions of annual crops related to each arable-type polygon (as an indicator of its probability of occurrence); and, ii) area fractions of each soil type in each polygon (as an indicator of its probability of occurrence). A total of 25,044 unique combinations of climate zones, agricultural land cover classes, administrative units and soil map units were identified. The same soil/crop combinations occur in many polygons which have the same climate while the fractions of the soils and arable crops are different. The number of unique combinations of climate, soil and agricultural land cover class is therefore only 7961. 26-year daily meteorological data, soil profile characteristics and crop management features were associated with each unique combination. The agro-environmental scenarios developed can be used to underpin the parameterization of environmental fate models for pesticides and should also have relevance for other agricultural pollutants. The implications for the improvement and further development of risk assessment procedures for pesticides are discussed.     

Tunnel safety,risk assessment and decision-making

This article gives a brief account of a project which was commissioned by the European Parliament and which has resulted in a report which has been published and is available on the web-site of the European Parliament [Beard, A.N., Cope, D., 2008. Assessment of the Safety of Tunnels. Commissioned by the European Parliament; Report IP/A/STOA/FWC/2005-28/SC22/29. Published in February 2008 on the European Parliament web-site under the rubric ‘Science and Technology Options Assessment’ (STOA)]. The project was funded by a grant from the European Parliament. The author was requested to carry out a study of tunnel safety and make recommendations to be considered for possible application within the European Union. The background to the project was the large number of catastrophic tunnel fires which have taken place in Europe since 1995. Twenty five recommendations are made within the Report the purpose of which is to help to increase tunnel safety in the European Union and, primarily, to help to move towards a common system of tunnel safety decision-making and risk assessment. This article focuses on some aspects of the content. However, it should not be assumed that aspects which are not included here are of lesser importance.     

Overview of Clean-Up Methods for Contaminated Sites

The selection of clean-up methods for contaminated land depends upon the nature and degree of contamination and on standards for clean up. There have been two approaches to clean-up standards, namely 'multifunctionality', where the site is cleaned up to a standard suitable for any use, and the 'end use'approach, as used in the UK, where clean up is to a standard suitable for the immediate use proposed for the site.
Clean-up options fall into four categories: (i) excavation and off-site removal; (ii) encapsulation and cover; (iii) retention of contaminated soil on site after treatment; and (iv) removal of contaminants from soil and reuse of cleaned soil on site.
In the UK, clean up has been dominated by off-site disposal and encapsulation/cover, while soil treatment techniques have been used mainly in the US and some European countries such as the Netherlands. This is largely as a result of the availability and relatively low cost of landfill in the UK and the lesser importance attached to date to groundwater considerations.
Soil treatment processes include biological, physicochemical, and thermal processes, all generally of much higher cost.