Lead and lead isotopes in agricultural soils of Europe – The continental perspective

Applied Geochemistry 27 (2012) 532–542

Contents lists available at SciVerse ScienceDirect

Applied Geochemistry

journal homepage: www.elsevier .com/ locate /apgeochem

Lead and lead isotopes in agricultural soils of Europe – The continental perspective

Clemens Reimann a,⇑, Belinda Flem a, Karl Fabian a, Manfred Birke b, Anna Ladenberger c, Philippe Négrel d,Alecos Demetriades e, Jurian Hoogewerff f, The GEMAS Project Team 1

a Geological Survey of Norway (NGU), P.O. Box 6315 Sluppen, N-7491 Trondheim, Norwayb Bundesanstalt für Geowissenschaften und Rohstoffe (BGR), Postfach 510153, D-30631 Hannover, Germanyc Geological Survey of Sweden (SGU), Box 670, S-751 28 Uppsala, Swedend BRGM, Service Métrologie Monitoring Analyse, 3 Avenue Claude Guillemin, BP 6009, 45060 Orléans Cedex 2, Francee Institute of Geology and Mineral Exploration, Entrance C, Olympic Village, Acharnae, Athens GR-13677, Greecef Oritain Global Ltd., 8 Pacific Street, Dunedin 9010, New Zealand

a r t i c l e i n f o a b s t r a c t

Article history:Received 8 November 2011Accepted 20 December 2011Available online 27 December 2011Editorial handling by R. Fuge

0883-2927/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.apgeochem.2011.12.012

⇑ Corresponding author.E-mail address: [email protected] (C. Reim

1 S. Albanese, M. Andersson, A. Arnoldussen, R. BaD. Cicchella, E. Dinelli, B. De Vivo, W. De Vos, M. DurisM. Eklund, V. Ernstsen, P. Filzmoser, T.E. Finne, D. FU. Fugedi, A. Gilucis, M. Gosar, V. Gregorauskiene, A. GP. Hayoz, G. Hobiger, R. Hoffmann, H. Hrvatovic, S. HG. Jordan, J. Kirby, J. Kivisilla, V. Klos, F. Krone, P. KweckP. Lucivjansky, D. Mackovych, B.I. Malyuk, R. MaquiN. Miosic, G. Mol, P. O’Connor, K. Oorts, R.T. OtteseS. Pfleiderer, M. Ponavic, C. Prazeres, U. Rauch, I. SaI. Schoeters, P. Sefcik, E. Sellersjö, F. Skopljak, I. SlT. Stafilov, T. Tarvainen, V. Trendavilov, P. Valera, VA.M. Zissimos, Z. Zomeni.

Lead isotopes are widely used for age dating, for tracking sources of melts, sediments, Pb products, foodand animals and for studying atmospheric Pb contamination. For the first time, a map of a Pb isotope land-scape at the continental-scale is presented. Agricultural soil samples (Ap-horizon, 0–20 cm) collected at anaverage density of 1 site/2500 km2 were analysed for Pb concentration and Pb isotopes (206Pb, 207Pb,208Pb). Lead concentrations vary from 1.6 to 1309 mg/kg, with a median of 16 mg/kg. Isotopic ratios of206Pb/207Pb range from 1.116 to 1.727 with a median of 1.202. The new data define the soil geochemicalPb background for European agricultural soil, providing crucial information for geological, environmentaland forensic sciences, public health, environmental policy and mineral exploration. The European conti-nental-scale patterns of Pb concentrations and Pb isotopes show a high variability dominated by geologyand influenced by climate. Lead concentration anomalies mark most of the known mineralised areasthroughout Europe. Some local Pb anomalies have a distinct anthropogenic origin.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Lead has been mined and used by humans for several thousandyears. The accumulated total world Pb production since ancienttimes is estimated to be 300 Mt (based on Nriagu, 1998 and up-dated to 2010 according to current world mine production figures).Estimates of the anthropogenic Pb fraction in the environmentrange from <10% (Strauss, 1978; Kownacka et al., 1990) to >90%(Nriagu, 1979). Detailed studies on environmental samples(e.g., peat bogs, ice cores, sediment cores) has suggested majorPb contamination of the northern hemisphere since ancient times

ll rights reserved.

ann).ritz, M.J. Batista, A. Bel-lan,, A. Dusza-Dobek, O.A. Eggen,light, S. Forrester, M. Fuchs,ulan, J. Halamic, E. Haslinger,usnjak, L. Janik, C.C. Johnson,o, L. Kuti, A. Lima, J. Locutura,

l, M. McLaughlin, R.G. Meuli,n, A. Pasieczna, V. Petersell,

lpeteur, A. Schedl, A. Scheib,aninka, A. Šorša, R. Srvkota,. Verougstraete, D. Vidojevic,

(Komárek et al., 2008; Bindler, 2011). The European Commissionhas identified diffuse contamination as one of the eight majorthreats to soil quality (European Commission, 2006).

Because Pb deposits have characteristic Pb isotope composi-tions (for data on the economically important deposits see:Sangster et al., 2000), Pb concentration and Pb-isotope ratios incombination may in some cases be used as a fingerprint to tracePb to its source. Natural Pb comprises four stable isotopes (naturalabundance in brackets): 204Pb (1.4%), 206Pb (24.1%), 207Pb (22.1%)and 208Pb (52.4%) in varying proportions, uniquely defined by thethree ratios 206Pb/207Pb, 208Pb/206Pb, 206Pb/204Pb. The reliable mea-surement of the isotope 204Pb needs much care and specialisedinstrumentation and is time consuming. The ratios between themore abundant isotopes 206Pb, 207Pb and 208Pb can be easily deter-mined by inductively coupled plasma mass spectrometry. In envi-ronmental sciences the 206Pb/207Pb isotope ratio is commonly usedto suggest Pb contamination of different compartments of theenvironment at the local to global scale (for a recent review onthe use of Pb isotopes in environmental sciences see Komáreket al., 2008). Data presented in Sangster et al. (2000) demonstratethat the 206Pb/207Pb isotope ratio of the most important Pb depositsvaries between 0.98 and 1.41. One major producer, Broken Hill inAustralia, is characterised by a very low 206Pb/207Pb isotope ratioof 1.04, while the largest Pb deposits in the world, the Mississippi

C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542 533

Valley type deposits in the United States, show a characteristicallyhigh 206Pb/207Pb isotope ratio of 1.4. It thus would be easy to deter-mine the origin of Pb in a sample, if one would have to decidewhether Pb from Broken Hill or the Mississippi Valley was the ori-ginal source.

In natural samples, determining the origin of Pb is more com-plicated and there are certain requirements for the successful useof Pb isotopes in environmental studies. For example, the isotoperatios of all sources of Pb that contribute to the Pb concentrationof a sample must be known. This will rarely be the case, whensoils or plant materials are collected in nature. Furthermore,and even more importantly, it is required that the isotopic com-position of Pb is not significantly affected by physico-chemicalfractionation processes in natural systems (Komárek et al.,2008), e.g., during soil formation and weathering. To be able toproceed with using Pb isotopes in forensic and environmentalstudies on a reliable foundation, it is important to know the Pbbackground concentration and isotopic compositions on a globalor at least continental scale.

However, in spite of its importance, not even the continental-scale distribution of Pb and Pb isotopes combined has ever beendocumented for any geochemical sample material. As the resultof a united effort of the European Geological Surveys, backed bythe European metals industry, here the first geochemical referencemaps of Pb concentrations combined with Pb isotope ratios at theEuropean scale are provided, against which the impact of contam-ination can be realistically assessed. The maps localise the majoranomalies of Pb in agricultural soils and help identify the processesthat generate the observed distribution patterns.

Fig. 1. Generalised geological map of Europe. For a detailed geological map of Euro

2. The survey area

Fig. 1 shows a simplified geological map of Europe, includingthe main features discussed in this paper. Further maps coveringtopography and land use of Europe can be found in almost anyworld atlas. A number of maps covering different themes at aboutthe scale of the GEMAS project (topography, geology, tectonics,fault and fracture zones, distribution of different rock types, dis-tribution of the main sedimentary basins, precipitation and pop-ulation density) can be found in Reimann and Birke (2010). ForEurope, an excellent source of land use information is the CORINEland use map of Europe (GLC 2000 database, 2003). A detailedgeological map of Europe is provided by Asch (2003), concisedescriptions of the geology of Europe can be found in Blundellet al. (1992) and McCann (2008). The soil atlas of Europe providesa wealth of information about the soils of Europe, but also con-tains maps of average precipitation, temperature, land use, popu-lation density, extent of the last glaciation and soil texture (Joneset al., 2005).

3. Material and methods

3.1. Sampling and sample preparation

Sampling took place from summer 2008 to early spring 2009. Intotal, 2211 samples of agricultural soil (Ap-horizon, 0–20 cm) werecollected at an average sampling density of 1 site per 2500 km2

(grid of 50 � 50 km) across Europe. A field handbook describes

pe see Asch (2003) – <http://www.bgr.de/karten/IGME5000/IGME5000.htm>).

Table 1Statistical parameters of Pb concentration (in mg/kg) and isotope ratios in Europeanagricultural soil (Ap horizon, 0–20 cm).

534 C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542

the sampling procedure in detail (EGS, 2008). All samples wereshipped to a central laboratory for preparation (air-drying andsieving to <2 mm using nylon screens).

Minimum Median Maximum

Pb 1.6 15.7 1309206Pb/207Pb 1.116 1.202 1.727207Pb/208Pb 0.287 0.403 0.417208Pb/206Pb 1.477 2.067 2.702

3.2. Analyses

Lead concentrations were determined following an aqua regiaextraction using an inductively coupled plasma mass spectrometer(ICP-MS) using tight external quality control procedures (Reimannet al., 2009a). All Pb isotope ratio measurements were carried outon a sector field inductively coupled plasma mass spectrometer(SF-ICP-MS; ELEMENT 1, Finnigan MAT, Bremen, Germany). A sam-ple weight of 0.5 g was carefully stirred with 8 mL 7 N HNO3 inPTFE vessels on a Vortex Genie shaker before the samples were ex-tracted under N2 pressure at 250 �C in an ultraclave (Milestone).During heating the pressure increases from the start pressure of50 bar to approximately 120 bar. After cooling the acid extractswere filtered through folded Whatman filters. All samples were di-luted to Pb concentrations below 100 lg/L and a HNO3-concentra-tion of 5% (v/v) prior to SF-ICP-MS analysis. The common Pbisotope standard NIST SRM 981 was used to correct for instrumen-tal mass discrimination. The instrumental uncertainty in each

Fig. 2. Map of the Pb concentration in European agricultural soil (Ap horizon, 0–20 cm,

measurement was between 0.003 and 0.0002 on the ratio. The cer-tified reference material Mess-3 (NRC, National Research CouncilCanada) was prepared in the same way as the samples and usedas a day-to-day control standard of both the SF-ICP-MS analysisand the digestion. From these data, the reproducibility of the Pbisotope ratio was estimated at 0.11% for 208Pb/207Pb and 0.10%for 206Pb/207Pb.

3.3. Data analysis and mapping

Geochemical data are compositional (closed) data, element con-centrations reported in wt.% or mg/kg sum up to a constant and are

<2 mm fraction, aqua regia analyses). Numbered anomalies as detailed in Table 2.

C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542 535

thus not free to vary. The information value of such data generallylies in the ratios between the variables (Aitchinson, 1986; Filzmo-ser et al., 2009). Filzmoser et al. (2009) discuss problems and pos-sibilities of univariate data analyses for compositional data. The

Table 2Explanation of the Pb concentration anomalies as indicated in Fig. 2.

Nr. GEMAS Pb-concentration anomalies – Supposed Primary Reason Possible o

Cities2 Bergen Sample ta

10 Amsterdam Industry, c11 London Industry23 Paris Industry40 Rome Volcanic r41 Naples Industry, V49 Lisbon Hardly vis

Ore deposits and districts1 Vassbo, Sweden Mining3 Bergslagen district, Sweden Mining7 Ireland, Zn–Pb–Cd deposits in Carboniferous rocks Undevelop8 Peak district/S. Peninne, UK Mining, st9 Central Wales mining field, Conway valley mineralisation Mining, sm

12 SW England mining field Mining, sm13 Rhenish slate mountains, Aachen–Stolberg and Benzberg ore

district in Germany; La Calamine, Moresnet district in BelgiumMining, sm

14 Northern anomaly: southern contact of Ramberg granite; southernanomaly: Thüringer Wald, Germany

Pb minera

15 Ore Mountains (Erzgebirge), Germany Mining, sm16 Kutna Hora Mining18 Silesian-Cracowian Zn–Pb-district Mining, in19 Slovakia, Banska Stiavnica Mining, in21 Freihung ore district, Germany Mining24 France Ba, F, Pb, Z26 Innsbruck area, Austria Mining Sc

anomaly –Salzburg–

27 Gurktaler Alps Pb–Zn occurences Small scal31 France, Poitou ridge Small Pb–

Jurassic ar33 Cantabrian basin (small Pb occurences) Industry?34 Massif Central, Cevenoles district, France Three maj

Pb extract35 S. Croatia, Pb mineralisation in Triassic carbonates Undevelop36 Sumadija, central Serbia Mining (R37 Lece-Haldiki, Serbia border to FYROM Mining (L38 Italy, Colline Metaliferre Mining39 Italy, northern Latium Volcanic r42 Sardinia, Iglesiente-Sulcis District Mining44 Guajaraz Pb deposit, Toledo, Spain Mining, gr47 Iberian pyrite belt Mining53 Linares – La Carolina district, Spain Mining (m

Contamination6 Midland valley, Scotland Mineralisa

17 Ostrava, Czech Republic Coal mine30 Lombardy, Northern Italy Downstre

industriali32 S of Oviedo, Spain – highway? Coal mine50 Sines, Portugal Local anom51 Nordenham lead smelter near Bremen Hardly vis52 Pribram, Pb smelter Ore depos

Geology5 Island of Mull, Scotland Evolved ac

20 Bohemian Massif, Czech Republic Granites, s25 Vosges, France Granites, s28 Hungary – border to Austria Tertiary v29 Southern Slovenia Karst, elev43 Catalonian coastal range granites Leucogran46 Central Spain, Pedochres Batholith Contains P48 Northern Portugal Granites, m

Unexplained4 NW Scotland Mineralisa

22 France No minera45 N of Ciudad Real, Spain Mn miner

solutions they suggested for the univariate case (bivariate or mul-tivariate results are not provided here) are used throughout thispaper. In terms of calculating statistical parameters, such as meanor standard deviation, it is important to note that compositional

ther sources at site/remarks

ken within settled areaontaminated floodplain sediments?

ocks with elevated Pb-background, industryesuvius and volcanic rocks

ible at scale of map, industry

edretching into industrial and current urban areas

elting, streches into urban areas and Cardiff coalfieldelting, stretches into urban areaselting, metals industry

lisation following Hercynian faults; mining

elting, metals industry

dustry, urban areasdustry

n veins in cover rocks near the Permo-Triassic unconformityhwaz/Brixlegg, Tyrol, Karwendelgebirge (Lafatsch, Tyrol); in southern part of themining in Obernberg/Brenner (Tyrol) and Schneeberg area (Italy); coincides with

Innsbruck–Brenner highwaye miningZn mine but also smelters (Melle, Alloue, Cherchonnes). Disseminated Pb inkosic sandstones.

or abandoned mines (Les Malines, Largentière, St Sebastine d’Aigrefeuille). Totaled > 500 Kt metal.ed, elevated natural background

udnik, Babe)ece, Trepca)

ocks with elevated Pb-background, small scale mining

anites

ore than 100 mines), smelting

tion, granitess, industrialised areaam Gorno and other Pb–Zn deposits but also a heavily urbanised andsed areas, industrialised area

alyible at scale of mapits, mining

id igneous rocksome miningmall Pb–Zn district in Southern Vosges (Lembach)

olcanics(?)ated Pb background in soilsites, Pb–F–Ba vein-type mineralisation; granites on the French siteb–Zn lodesinor vein-type mineralisation

tion (remote location) or local contaminationlisation known, no smelter, no major city but note proximity to Verdun (WW 1?)alisation documented, limestones, probably elevated Pb background values

Table 3Statistical parameters of Pb concentration (in mg/kg) and isotope ratios in northern and southern European agricultural soil (Ap horizon, 0–20 cm).

Northern Europe Southern Europe

Minimum Median Maximum Minimum Median Maximum

Pb 1.6 9.6 52 2.1 20 1309206Pb/207Pb 1.143 1.258 1.727 1.116 1.195 1.434207Pb/208Pb 0.287 0.397 0.414 0.379 0.404 0.417208Pb/206Pb 1.477 2.017 2.702 1.781 2.075 2.251

Fig. 3. Map of the major Pb–Zn deposits of Europe (updated after De Vos et al., 2006).

536 C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542

Fig. 4. Map of the 206Pb/207Pb isotope ratio in European agricultural soils.

C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542 537

data do not plot in the Euclidian space, but rather on the Aitchisonsimplex. Thus, only order statistics are used to present the data(note that percentiles will remain unchanged under log-transfor-mation but not under a log-ratio-transformation). For producingthe colour surface maps kriging was used to convert the valuesfrom the irregularly distributed sampling sites to a regular grid.Successful kriging is based on a careful variogram analysis. Classboundaries used for the colour maps are based on percentiles. Figs.7 and 8 were produced in Wolfram Mathematica 8.

4. Results and discussion

Lead concentrations (aqua regia extraction) in the uppermost20 cm of agricultural soils (Ap-horizon) from western Europe rangefrom 1.6 to 1309 mg/kg, with a median of 16 mg/kg (Table 1,Fig. 2). The major Pb anomalies in the Pb-concentration map arenumbered and an explanation for the most likely source of anyof the Pb anomalies is provided in Table 2. The map (Fig. 2) revealsa prominent, almost straight boundary in Pb concentrations be-tween northern and southern Europe. Table 3 demonstrates thatnorthern European soil Pb concentrations (median 10 mg/kg) arelower by a factor of 2 compared to southern Europe (median20 mg/kg). When comparing the geochemical map with the geo-logical map (Fig. 1) the boundary coincides remarkably well withthe maximum extent of the last glaciation, and occurs near theTrans-European Suture Zone, one of the main tectonic borders inEurope (Grad et al., 1999; Janik et al., 2005). Because the same pat-tern is known from deep (C-horizon) soils (Salminen et al., 2005;De Vos et al., 2006), the Pb concentration in the surface soil Ap-

horizon at the European scale reflects the natural backgroundvariation. In addition to the difference in the source rocks, thelarge-scale increase in Pb concentration from north to south alsoreflects the natural difference between the young, often coarse-grained and less weathered soils in northern Europe and the older,finer grained and more weathered soils of southern Europe. At amore local scale the majority of Pb anomalies on the map coincidewith known Pb mineral belts or deposits (Fig. 3, Table 2 – e.g., Peakdistrict, Central Wales Mining Field, Ore Mountains, Rhenish SlateMountains, Iberian Pyrite Belt). Because most of the known depos-its have been mined, local industrial Pb-sources often developed inthe same areas. At the resolution of a continental-scale map (Fig. 2)it is thus difficult to quantify the natural versus the anthropogenicorigin of the Pb causing these anomalies. Such quantification willrequire geochemical mapping at a much more detailed scale(several samples per km2), typical for local studies, e.g., urban geo-chemical investigations (Johnson et al., 2011) or mineral explora-tion projects (Reimann et al., 2009c, 2010; Cohen et al., 2011).The data presented here show anomalies indicating contaminationaround several cities. For example London, Paris, Naples, Rome andLisbon are all marked by Pb concentration anomalies in Fig. 2.However, even here it is not easy to pinpoint their exact origin:traffic or industry, or, in the case of Naples and Rome, geology(presence of volcanic rocks with elevated Pb-concentrations). Itmust also be noted that the majority of European cities are notmarked by a Pb anomaly at the scale of this survey.

The map also indicates several anomalies where neither a Pbdeposit nor contamination are a likely Pb source. These anomaliesare directly related to geology, for example the Island of Mull(evolved acid igneous rocks), Bavarian Forest (granites), southern

Fig. 5. Map of the 207Pb/208Pb isotope ratio in European agricultural soils.

538 C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542

Slovenia (karst), central Spain (Pedochres Batholith) and severalgranitic intrusions in France (see Fig. 1, Table 2). For all local anom-alies, the Pb concentrations decrease on a scale <100 km distancefrom source towards the regional geochemical background(Reimann and Garrett, 2005).

Although theoretically independent of Pb concentration, thecontinental scale distribution of the 206Pb/207Pb isotope ratios inEuropean agricultural soils in Fig. 4 closely resembles a mirror im-age of the Pb concentration pattern (Fig. 2). The 206Pb/207Pb isotoperatios range from 1.116 to 1.727, and the median value for Euro-pean agricultural soils is 1.202 (Table 1). The map of the206Pb/207Pb isotope ratios (Fig. 4) is dominated by high ratios inthe old, granitic terrains of northern Europe (FennoscandianShield) and predominantly low ratios in southern Europe. In-creased concentrations of the isotope 206Pb in soils developed onArchaean and Palaeoproterozoic granitic rocks are related to theradioactive decay of 238U on a geological timescale. Higher valuesof the 206Pb/207Pb isotope ratio, therefore, occur north-east of theTrans-European Suture Zone but also in those regions of southernEurope where younger granitic bedrocks, rich in 238U, bias the Pbisotopes towards a higher ratio (e.g., Northern Portugal/Spain with-in the Iberian Variscides, central Spain, Massif Central, BohemianMassif, Ore Mountains). Also the Scandinavian Caledonides clearlyshow lower 206Pb/207Pb isotope ratios than the older bedrocksforming the remainder of Scandinavia. Anomalously low isotoperatios are often observed in coastal areas (especially well visiblein Norway, Denmark and Scotland – Fig. 4). They coincide withwet and/or cold climatic zones, where special conditions forweathering and evolution of soils are suspected to influence the

Pb isotopic system (Reimann et al., 2008a, 2009b, 2011). Althoughno samples were taken directly within any major city the use ofBroken Hill Pb (206Pb/207Pb = 1.04, Sangster et al., 2000) in Euro-pean leaded gasoline (e.g., Novak et al., 2003) will still be the mostlikely source of decreased isotope ratios in local areas with densetraffic (e.g., London, Paris). In both cases the small size of the neg-ative 206Pb/207Pb anomaly pattern in the map (Fig. 4) related tothese cities demonstrates how local the influence remains at theEuropean scale. The location of the majority of the large Europeancities is not even indicated by unusually low 206Pb/207Pb isotoperatios, though many case studies demonstrate massive Pb contam-ination within the city borders of almost any investigated urbanarea (see examples in Johnson et al., 2011). This demonstratesthe truly local scale of the anthropogenic impact. It is also not pos-sible to establish a relationship between traffic density and thepatterns observed in the map (Fig. 4). The influence of Pb derivedfrom fertilizers should, in contrast, increase the value of the isotoperatio (Négrel and Roy, 2002) but again remains invisible at thescale of this survey. Agricultural soils are prone to wind erosionand thus the natural variation of Pb isotopes in these soils willinfluence the isotopic composition observed in the atmosphere(Négrel and Roy, 2002).

Fig. 5 shows the spatial distribution of the 207Pb/208Pb isotoperatio in the European agricultural soils. Again a pattern emergesthat clearly reflects the ice-age boundary. In general the207Pb/208Pb isotope ratios are lower in northern Europe than inthe south. Great Britain and Ireland, The Netherlands and parts ofNorthern Germany and Belgium as well as the Alps show increased207Pb/208Pb isotope ratios. High 207Pb/208Pb ratios are a characteristic

Fig. 6. Binary plot of 206Pb/207Pb versus 208Pb/206Pb in European agricultural soils.The isotopic composition of Broken Hill and Mississippi Valley type ore is shown inaddition (data from Sangster et al., 2000).

C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542 539

of many European Pb ores (see data in Sangster et al., 2000). Anumber of well localised high values in the 207Pb/208Pb isotope ra-tio also occur in Norway and Sweden, marking more or less the lineof Laisvall type Pb–Zn deposits in Fig. 3. However, in general thesoils developed on the Precambrian igneous rocks in Finland arecharacterised by especially low 207Pb/208Pb isotope ratios.

A common approach to use Pb isotopes for source identificationis to use cross-plots of the isotope ratios, e.g. 206Pb/207Pb versus208Pb/207Pb, or 206Pb/207Pb versus 208Pb/206Pb (Fig. 6). Thesediagrams are intended to reveal trends from, for example, theupper crust composition (206Pb/207Pb = 1.20) towards end-mem-bers characteristic for ‘‘pollution’’, e.g. the Pb isotope compositionof the Broken Hill deposit in Australia as ‘‘typical for the European

Fig. 7. Ternary plot of the relative proportions of the three Pb isotopes 206Pb, 207Pb and 2

deposits and of Broken Hill and Mississippi Valley type ore (from Sangster et al., 2000) isouth (red).

leaded gasoline signal’’ (e.g., Novak et al., 2003). Such trends arethen interpreted as a proof of contamination. However, due tothe use of dependent variables (the same isotope occurs in the ra-tios on both axes, e.g., 206Pb/207Pb versus 208Pb/207Pb or206Pb/207Pb versus 208Pb/206Pb), these diagrams are mathemati-cally flawed because intrinsic correlations must be expected whentwo ratios that share one or more variables are compared (Kenney,1982; Ellam, 2010; Lenahan et al., 2011). The interpretation ofthese cross-plots is, therefore, much less straight-forward or easythan usually presented in the environmental literature. Fig. 6 dem-onstrates that the diagram, when all the European soil samples areincluded, could be used to argue for almost anything: the input ofBroken Hill Pb as well as the input of Mississippi Valley Pb andeven the existence of extra-terrestrial Pb in the agricultural soilsof Europe because many soil samples show isotope ratios that can-not be traced to any Pb deposit world-wide. To derive at a conclu-sion about the source of the Pb additional information, for examplethe spatial distribution of the samples on a map, is needed.

According to Ellam (2010) independent diagrams can be plottedonly when the isotope 204Pb is measured, i.e. usually only whenthermal ionisation mass spectrometry (TIMS) or multi-collectorSF-ICP-MS measurements are available and all four stable isotopesare involved in the plot. Most environmental data presented in theliterature, however, are obtained by either high resolution (HR)ICP-MS or even quadrupole (Q) ICP-MS analysis and 204Pb is notmeasured. Moreover, even a plot of 206Pb/207Pb versus 208Pb/204Pbis not completely independent because the fractional contents ofall four stable Pb isotopes add up to 100%.

In Fig. 7 the three Pb isotopes 206, 207 and 208 are plotted in aternary diagram as suggested in Reimann et al. (2008b) to solvethis dependence problem. On this diagram the colour of the sym-bol changes continuously with location from dark blue (north) tored (south). In addition to all soil samples, the 50 most importantEuropean Pb deposits, as well as the two often used ‘‘Pb end

08Pb in European agricultural soils (Y). The isotopic composition of 50 European Pbs also provided (circles). Symbol colours change with location from north (blue) to

540 C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542

members’’ Broken Hill and Mississippi Valley (Sangster et al., 2000)are shown as circles and similarly colour-coded according to loca-tion (the non-European deposits Broken Hill and Mississippi Valleyin black). A first important observation is that the soil samples cov-er a larger range of isotopic compositions than the ore deposits.This demonstrates that the natural variation of Pb isotopes in soilsis wider than the variation observed for Pb deposits. Broken Hill Pbfalls far outside the range covered by the European agriculturalsoils. The majority of European Pb deposits and the majority ofthe soil samples plot within a limited field in the ternary diagramof 24% < 206Pb(%) < 28%, 20% < 207Pb(%) < 22%, 52% < 208Pb(%) < 54%.Because the distribution patterns of Pb concentrations and isotopicratios in Figs. 2, 4 and 5 closely follow climatic and geological units,they provide no evidence that atmospheric-Pb contamination isthe prevailing source of Pb in European agricultural soils. In thiscase one would expect to see anthropogenic activity or recentpaths of wind transport dominating the Pb distribution patterns.The absence of such patterns in the maps for agricultural soil doesnot exclude the existence of Pb-contamination as recorded by spe-cial media like peat bogs (e.g., Kylander et al., 2010; Bindler, 2011),but questions the continental scale conclusions drawn from thesestudies. In order to identify and characterise contamination viaatmospheric transport local-scale, detailed studies in the sur-roundings of likely sources are required. Ideally, the source emis-sions should have a well defined Pb-isotope ratio, deviating fromthe local background. Such studies exist, and they have repeatedlydemonstrated that the atmospheric Pb contamination signal disap-pears in the background variation at a distance of metres to somekilometres from the source (e.g., Ault et al., 1970; Ettler et al.,2006; Hou et al., 2006; Klaminder et al., 2008).

Figs. 4 and 5 imply that at the continental scale the large isoto-pic variability of the geological background masks any additionalsignal from atmospheric-Pb contamination. However, combiningthe data of Pb concentration with the isotopic composition makesit possible to isolate an isotopic composition of the excess Pb thatgenerates local variations in concentration. The idea is that localsites with low Pb concentration should be less atmosphericallycontaminated than neighbouring sites with high Pb concentration.The isotopic composition of the excess Pb, therefore, should clearlyreflect the atmospheric contamination signal. The calculation pro-ceeds in the following way: For each site S a cell C with a radius of400 km around S is considered. Within C there are sites Smin withminimal Pbmin(S) = Pb(Smin), and Smax with maximal Pb concentra-tion Pbmax(S) = Pb(Smax). The excess Pb Pbex at S is the difference be-

Fig. 8. Isotopic composition of calculated ‘‘excess Pb’’ in the Ap samples (+) plottedagainst ‘‘location’’ (Longitude + Latitude). The isotopic composition of 50 EuropeanPb deposits (Sangster et al., 2000) in relation to their location in Europe is alsoprovided (filled squares). The geographical location of a few outliers that plotoutside the isotopic range shown in this diagram is provided in the form of blackcircles.

tween Pbmax(S) and Pbmin(S). Using the isotopic data at Smin andSmax one can separately calculate the concentrations206Pbex = 206Pb(Smax) � 206Pb(Smin), as well as 207Pbex and 208Pbex.The ratio 206Pbex/207Pbex varies for south-western Europe (longi-tude + latitude < 70�) between 1.16 and 1.22 (Fig. 8). These valuescan be related to the typical low 206Pb/207Pb ratios of SW-EuropeanPb ores which provide a potential natural source of local Pb-anom-alies in soils. They, therefore, do not prove predominant atmo-spheric contamination. Further towards the NE the206Pbex/207Pbex variation increases mainly due to the occurrenceof very high 206Pbex/207Pbex ratios, again consistent with the occur-rence of very high 206Pb/207Pb ratios in some NE-European Pb ores,but also with the general geology (old granitic bedrocks with high206Pb/207Pb ratios). The outstanding break between low and veryhigh variation in this diagram marks exactly the location of theTrans-European Suture Zone. For some sites in NE-Europe206Pbex/207Pbex deviates towards very low isotope ratios (Fig. 8).Such deviations would usually be interpreted as a predominantatmospheric contamination by Broken Hill Pb with its characteris-tically low 206Pb/207Pb ratio (e.g., Rosman et al., 1994; Renberget al., 2002; Novak et al., 2003; Kylander et al., 2010), especiallywhen plotted in a traditional binary diagram of isotope ratiosagainst a proposed hypothetical source as shown in Fig. 6. How-ever, here also this interpretation is not conclusive, because206Pbex/207Pbex in these soils again most likely simply reflects a rel-atively low 206Pb/207Pb isotope ratio of the local ore deposits or lo-cal deviations in the bedrock. Fig. 8 shows that deposits with suchlow isotope ratios occur in NE-Europe. A second argument, againstpredominant atmospheric contamination at these sites, is that traf-fic density is much lower here than in most other locationsthroughout Europe. Another potential influence on the Pb systemat these northern locations that needs consideration is simply ahigher content of organic material in some of the soils (Reimannet al., 2008a, 2009b).

5. Conclusions

Soil formation through weathering of bedrock and sediments isan extremely slow process, and soils must be considered an essen-tial and non-renewable resource. Knowledge and extrapolation ofthreats to soils, e.g., changes in element concentrations and thereasons thereof, are of primary importance for society. In this con-text, the addition of isotope tracing at the continental scale pro-vides important supplementary information about the origin andbehaviour of elements in soils.

The pattern dominating the Pb maps presented here is thesouthern limit of the last glaciation, almost coincident with thetectonic border between the old Precambrian craton of north-east-ern Europe and the younger Palaeozoic platform of western Eur-ope. The Pb isotope map also depicts the border between theCaledonides and the Baltic Shield. Evidently, a close relationshipof the Pb concentrations and isotopic ratios with geology is pre-served in the agricultural soils at the European scale. This provesthat the majority of Pb in European agricultural soils is at presentstill of natural origin.

The atmospheric Pb contamination of the northern hemisphereas demonstrated using ice cores (Rosman et al., 1994), lake sedi-ments (Renberg et al., 2002) or peat bogs (Kylander et al., 2010)contributes little to the total Pb inventory of European agriculturalsoils. Therefore, the quantity of diffuse atmospheric Pb contamina-tion at the continental scale does not at this time really represent athreat to agricultural soils and humanity at large, while local pol-lution certainly does. Attention clearly needs to be directed to-wards local contamination sources and the toxic potential of

C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542 541

atmospheric Pb, rather than to the quantity of regional or pan-European scale atmospheric Pb transport.

The maps presented here establish the natural geochemicalbackground for both Pb concentration and Pb isotope ratios inEuropean agricultural soils in aqua regia/HNO3-extracts and dem-onstrate the existence of two major continental-scale natural back-ground regimes for northern and southern Europe, each in turnsubstantially influenced by local-scale internal variability. Animportant conclusion, for any interpretation of local Pb isotopestudies, is that they have to demonstrate that a clearly definedand well known source of Pb contamination contrasts to the vari-ations in the regional geochemical background. Using global aver-age values, as representative for a certain source (e.g., the averageupper crust for ‘‘geogenic Pb’’), is simply not sufficient.

Availability of the data

All results from the GEMAS project will be published in the formof a book in 2013. All data will accompany that book in the form ofexcel files on an attached CD-ROM. Until release of the book, thedata will be made available to third parties via a request to the firstauthor if the Geological Surveys of Europe and the InternationalLead Association – Europe (ILA) agree on a suggested use of thedata.

Acknowledgements

The GEMAS project is a cooperation project of the EuroGeoSur-veys Geochemistry Expert Group with a number of outside organ-isations (e.g., Alterra in The Netherlands, the Norwegian Forest andLandscape Institute, several Ministries of the Environment andUniversity Departments of Geosciences in a number of Europeancountries, CSIRO Land and Water in Adelaide, Australia) and Euro-metaux. Sampling was covered by the participating organisationsin their home countries, the analytical work was co-financed bythe following organisations: Eurometaux, Cobalt DevelopmentInstitute (CDI), European Copper Institute (ECI), Nickel Institute,Europe, European Precious Metals Federation (EPMF), InternationalAntimony Association (i2a), International Manganese Institute(IMnI), International Molybdenum Association (IMoA), ITRI Ltd.(on behalf of the REACH Tin Metal Consortium), International ZincAssociation (IZA), International Lead Association-Europe (ILA-Europe), European Borates Association (EBA), the (REACH) VanadiumConsortium (VC) and the (REACH) Selenium and TelluriumConsortium. Important discussions with Arnold Arnoldussen, PeterEnglmaier and Friedrich Koller are acknowledged. Lead isotopeanalyses were carried out as an internal research project at theGeological Survey of Norway. The GEMAS project is managed bythe Geological Survey of Norway with financial support from Euro-GeoSurveys. The Directors of the European Geological Surveys andthe additional participating organisations are thanked for makingsampling of almost all of Europe, on a tight time schedule, possible.The authors thank the two reviewers for a thorough and insightfulreview of the manuscript, resulting in many improvements.

References

Aitchinson, J., 1986. The Statistical Analysis of Compositional Data. Chapman andHall, London, UK.

Asch, K. (Ed.), 2003. The 1:5 Million International Geological Map of Europe andAdjacent Areas: Development and Implementation of a GIS-enabled Concept;Geologisches Jahrbuch; SA 3. BGR, Hannover, Schweitzerbart (Stuttgart).

Ault, W.U., Senechal, R.G., Erlebach, W.E., 1970. Isotopic composition as a naturaltracer of lead in the environment. Environ. Sci. Technol. 4, 305–317.

Bindler, R., 2011. Contaminated lead environments of man: reviewing the leadisotopic evidence in sediments, peat, and soils for the temporal and spatialpatterns of atmospheric lead pollution in Sweden. Environ. Geochem. Health 33,311–329.

Blundell, D., Freeman, R., Mueller, St., 1992. A Continent Revealed: The EuropeanGeotraverse. Cambridge University Press, Cambridge, UK.

Cohen, D.R., Rutherford, N.F., Morisseau, E., Zissimos, A.M., 2011. Geochemical Atlasof Cyprus. UNSW Press, Sydney, Australia.

De Vos, W., Tarvainen, T. (Chief editors), Salminen, R., Reeder, S., De Vivo. B.,Demetriades, A., Pirc, S., Batista, M.J., Marsina, K., Ottesen, R.-T., O’Connor, P.J.,Bidovec, M., Lima, A., Siewers, U., Smith, B., Taylor, H., Shaw, R., Salpeteur, I.,Gregorauskiene, V., Halamic, J., Slaninka, I., Lax, K., Gravesen, P., Birke, M.,Breward, N., Ander, E.L., Jordan, G., Duris, M., Klein, P., Locutura, J., Bel-lan, A.,Pasieczna, A., Lis, J., Mazreku, A., Gilucis, A., Heitzmann, P., Klaver, G., Petersell,V., 2006. Geochemical Atlas of Europe. Part 2 – Interpretation of GeochemicalMaps, Additional Tables, Figures, Maps and Related Publications. GeologicalSurvey of Finland, Espoo, Finland, 692 pp. <http://www.gtk.fi/publ/foregsatlas/>(accessed 26.10.11).

EGS, 2008. EuroGeoSurveys Geochemistry Working Group. EuroGeoSurveysGeochemical Mapping of Agricultural and Grazing Land in Europe (GEMAS) –Field Manual. Norges Geologiske Undersøkelse Report, 2008.038, 46 pp. <http://www.ngu.no/upload/Publikasjoner/Rapporter/2008/2008_038.pdf>.

Ellam, R.M., 2010. The graphical presentation of lead isotope data for environmentalsource apportionment. Sci. Total Environ. 408, 3490–3492.

Ettler, V., Mihlajevic, M., Sebek, O., Molek, M., Grygar, T., Zeman, J., 2006.Geochemical and Pb isotopic evidence for sources and dispersal of metalcontamination in stream sediments from the mining and smelting district ofPribram, Czech Republic. Environ. Pollut. 142, 409–417.

European Commission, 2006. Thematic Strategy for Soil Protection. ReportCOM(2006) 231.

Filzmoser, P., Hron, K., Reimann, C., 2009. Univariate statistical analysis ofenvironmental (compositional) data – Problems and possibilities. Sci. TotalEnviron. 407, 6100–6108.

GLC2000 database, 2003. The CORINE Land Cover Map for Europe. EuropeanCommission Joint Research Centre, Ispra. <http://edit.csica.es/Soil-Vegetation-LandCover.html>.

Grad, M., Janik, T., Yliniemi, J., Guterch, A., Luosto, U., Tiira, T., Komminaho, K., Sroda,P., Höing, K., Makris, J., Lund, C.E., 1999. Crustal structure of the Mid-PolishTrough beneath the Teisseyre-Tornquist Zone seismic profile. Tectonophysics314, 145–160.

Hou, X., Parent, M., Savard, M.M., Tasse, N., Begin, C., Marion, J., 2006. Leadconcentrations and isotope ratios in the exchangeable fraction: tracing soilcontamination near a copper smelter. Geochem.: Explor. Anal. Environ. (GEEA)6, 229–236.

Janik, T., Grad, M., Guterch, A., Dadlez, R., Yliniemi, J., Tiira, T., Keller, G.R., Gaczynski,E., 2005. Lithospheric structure of the Trans-European Suture Zone along theTTZ-CEL03 seismic transect (from NW to SE Poland). Tectonophysics 411, 129–156.

Johnson, C.C., Demetriades, A., Locutura, J., Ottesen, R.T. (Eds.), 2011. Mapping theChemical Environment of Urban Areas. John Wiley & Sons, Chichester, UK.

Jones, A., Montanarella, L., Jones, R., 2005. Soil Atlas of Europe. EuropeanCommission, Luxemburg.

Kenney, B.C., 1982. Beware of spurious self-correlations! Water Resour. Res. 18,1041–1048.

Klaminder, J., Bindler, R., Rydberg, J., Renberg, I., 2008. Is there a chronologicalrecord of atmospheric mercury and lead deposition preserved in the mor layer(O-horizon) of boreal forest soils. Geochim. Cosmochim. Acta 72, 703–712.

Komárek, M., Ettler, V., Chrastny, V., Mihaljevic, M., 2008. Lead isotopes inenvironmental studies: a review. Environ. Int. 34, 562–577.

Kownacka, L., Jaworowski, Z., Suplinska, M., 1990. Vertical distribution and flows ofPb and natural radionuclides in the atmosphere. Sci. Total Envon. 91, 199–221.

Kylander, M.E., Klaminder, J., Bindler, R., Weiss, D.J., 2010. Natural lead isotopevariations in the atmosphere. Earth Planet. Sci. Lett. 290, 44–53.

Lenahan, M.J., Bristow, K.L., Caritat, P.de., 2011. Detecting induced correlations inhydrochemistry. Chem. Geol. 284, 182–192.

McCann, T. (Ed.), 2008. The Geology of Central Europe. Geological Society, London,UK.

Négrel, P., Roy, S., 2002. Investigating the sources of the labile fraction in sedimentsfrom silicate-drained rocks using trace elements, and strontium and leadisotopes. Sci. Total Environ. 298, 163–181.

Novak, M., Emmanuel, S., Vile, M.A., Yigal, E., Veron, A., Paces, T., Kelman, W.,Vanecek, M., Stepanova, M., Brizova, E., Hovorka, J., 2003. Origin of lead in eightcentral European peat bogs determined from isotope ratios, strength andoperation times of regional pollution sources. Environ. Sci. Technol. 37, 437–445.

Nriagu, J.O., 1979. Global inventory of natural and anthropogenic emissions of tracemetals to the atmosphere. Nature 279, 409–411.

Nriagu, J.O., 1998. Tales told in lead. Science 281, 1622–1623.Reimann, C., Birke, M. (Eds.), 2010. Geochemistry of European Bottled Water.

Borntraeger Science Publishers, Stuttgart.Reimann, C., Garrett, R.G., 2005. Geochemical background: concept and reality. Sci.

Total Environ. 350, 12–27.Reimann, C., Flem, B., Englmaier, P., Finne, T.E., Koller, F., Nordgulen, Ø., 2008a. The

biosphere: a homogeniser of Pb-isotope ratios. Appl. Geochem. 23, 705–722.Reimann, C., Flem, B., Englmaier, P., Finne, T.E., Koller, F., Nordgulen, Ø., 2008b.

Reply to the comments on ‘‘The biosphere: a homogenizer of Pb-isotopesignals’’ by Richard Bindler and William Shotyk. Appl. Geochem. 23, 2527–2535.

Reimann, C., Demetriades, A., Eggen, O.A., Filzmoser, P., 2009a.The EuroGeoSurveysGeochemical Mapping of Agricultural and Grazing Land Soils Project (GEMAS) –

542 C. Reimann et al. / Applied Geochemistry 27 (2012) 532–542

Evaluation of Quality Control Results of Aqua Regia Extraction Analysis. NorgesGeologiske Undersøkelse Report, 2009.049. <http://www.ngu.no/upload/Publikasjoner/Rapporter/2009/2009_049.pdf> (accessed on 26.10.11).

Reimann, C., Englmaier, P., Flem, B., Gough, L., Lamothe, P., Nordgulen, Ø., Smith, D.,2009b. Geochemical gradients in O-horizon soils of southern Norway: naturalor anthropogenic? Appl. Geochem. 24, 62–76.

Reimann, C., Matschullat, J., Birke, M., Salminen, R., 2009c. Arsenic distribution inthe environment: the effects of scale. Appl. Geochem. 24, 1147–1167.

Reimann, C., Matschullat, J., Birke, M., Salminen, R., 2010. Antimony in theenvironment – lessons from geochemical mapping. Appl. Geochem. 25, 175–198.

Reimann, C., Smith, D.B., Woodruff, L.G., Flem, B., 2011. Pb-concentrations and Pb-isotope ratios in soils collected along an east-west transect across the UnitedStates. Appl. Geochem. 26, 1623–1631.

Renberg, I., Bränvall, M.L., Bindler, R., Empteryd, O., 2002. Stable lead isotopes andlake sediments – a useful combination for the study of atmospheric leadpollution history. Sci. Total Environ. 292, 45–54.

Rosman, K.J.R., Chisholm, W., Boutron, C.F., Candelone, J.P., Hong, S., 1994. Isotopicevidence to account for changes in the concentration of lead in Greenlandsnow between 1960 and 1988. Geochim. Cosmochim. Acta 58,3265–3269.

Salminen, R. (Chief-Editor), Batista, M.J., Bidovec, M. Demetriades, A., De Vivo. B., DeVos, W., Duris, M., Gilucis, A., Gregorauskiene, V., Halamic, J., Heitzmann, P.,Lima, A., Jordan, G., Klaver, G., Klein, P., Lis, J., Locutura, J., Marsina, K., Mazreku,A., O’Connor, P.J., Olsson, S.Å., Ottesen, R.-T., Petersell, V., Plant, J.A., Reeder, S.,Salpeteur, I., Sandström, H., Siewers, U., Steenfelt, A., Tarvainen, T., 2005.Geochemical Atlas of Europe. Part 1 – Background Information, Methodologyand Maps. Geological Survey of Finland, Espoo, Finland. <http://www.gtk.fi/publ/foregsatlas/> (accessed on 26.10.11).

Sangster, D.F., Outridge, P.M., Davis, W.J., 2000. Stable lead isotope characteristics oflead ore deposits of environmental significance. Environ. Rev. 8, 115–147.

Strauss, W., 1978. Air Pollution Control, Part III. John Wiley & Sons, New York.