Comparing results from two continental geochemical surveys to world soil
composition and deriving Predicted Empirical Global Soil (PEGS2) reference values
Patrice de Caritat
a,⁎
, Clemens Reimann
b, NGSA Project Team
1and GEMAS Project Team
2a
Geoscience Australia, GPO Box 378, Canberra, ACT 2601, Australia
bGeological Survey of Norway, PO Box 6315 Sluppen, 7491 Trondheim, Norway
a b s t r a c t
a r t i c l e i n f o
Article history:
Received 14 September 2011
Received in revised form 22 November 2011 Accepted 22 December 2011
Available online 21 January 2012 Editor: G. Henderson Keywords: regolith Critical Zone geochemistry major elements trace elements global soil composition
Analytical data for 10 major oxides (Al2O3, CaO, Fe2O3, K2O, MgO, MnO, Na2O, P2O5, SiO2and TiO2), 16 total trace elements (As, Ba, Ce, Co, Cr, Ga, Nb, Ni, Pb, Rb, Sr, Th, V, Y, Zn and Zr), 14 aqua regia extracted ele-ments (Ag, As, Bi, Cd, Ce, Co, Cs, Cu, Fe, La, Li, Mn, Mo and Pb), Loss On Ignition (LOI) and pH from 3526 soil samples from two continents (Australia and Europe) are presented and compared to (1) the composi-tion of the upper continental crust, (2) published world soil average values, and (3) data from other continental-scale soil surveys. It can be demonstrated that average upper continental crust values do not provide reliable estimates for natural concentrations of elements in soils. For many elements there exist substantial differences between published world soil averages and the median concentrations observed on two continents. Direct comparison with other continental datasets is hampered by the fact that often mean, instead of the statistically more robust median, is reported. Using a database of the worldwide dis-tribution of lithological units, it can be demonstrated that lithology is a poor predictor of soil chemistry. Climate-related processes such as glaciation and weathering are strong modifiers of the geochemical signa-ture inherited from bedrock during pedogenesis. To overcome existing shortcomings of predicted global or world soil geochemical reference values, we propose Preliminary Empirical Global Soil reference values based on analytical results of a representative number of soil samples from two continents (PEGS2).
Crown Copyright © 2012 Published by Elsevier B.V.
1. Introduction
The Earth's surface is the interface between the geosphere, pedo-sphere, biopedo-sphere, hydrosphere and atmopedo-sphere, and supports human, animal and plant life. This“Critical Zone” hosts a multitude of physical, chemical and biological processes active over a range of spatial and temporal scales; these impact mass and energy exchanges governing processes as varied and crucial as soil formation, plant growth, water storage, nutrients cycling, metal and radionuclide transport, etc. (Brantley et al., 2007). This interface is under growing stress as the world's population continues to grow and with it the demand for food, water, energy and raw materials. Therefore, im-proving our understanding of the chemical composition and vari-ability of soils at the continental, and ultimately global, scale is both important and pressing.
Many researchers have attempted to estimate the average chemi-cal composition and natural variation of element concentrations in “world soils” (e.g., Bowen, 1979; Pendias, 2001; Kabata-Pendias and Kabata-Pendias, 1984; Koljonen, 1992; Rauch, 2011; Vinogradov, 1954). The values provided are based on data from existing soil sur-veys in different parts of the world often combined with estimates about the geochemical composition of the Earth's crust. In this ap-proach, the empirical data usually come from surveys covering rela-tively small areas and with rather few samples (Bowen, 1979;
⁎ Corresponding author. Tel.: +61 2 6249 9378.
E-mail address:[email protected](P. de Caritat).
1
E. Bastrakov, D. Bowbridge, P. Boyle, S. Briggs, D. Brown, M. Brown, K. Brownlie, P. Burrows, G. Burton, J. Byass, P. de Caritat, N. Chanthapanya, M. Cooper, L. Cranfield, S. Curtis, T. Denaro, C. Dhnaram, T. Dhu, G. Diprose, A. Fabris, M. Fairclough, S. Fanning, R. Fidler, M. Fitzell, P. Flitcroft, C. Fricke, D. Fulton, J. Furlonger, G. Gordon, A. Green, G. Green, J. Greenfield, J. Harley, S. Heawood, T. Hegvold, K. Henderson, E. House, Z. Husain, B. Krsteska, J. Lam, R. Langford, T. Lavigne, B. Linehan, M. Livingstone, A. Lukss, R. Maier, A. Makuei, L. McCabe, P. McDonald, D. McIlroy, D. McIntyre, P. Morris, G. O'Connell, B. Pappas, J. Parsons, C. Petrick, B. Poignand, R. Roberts, J. Ryle, A. Seymon, K. Sherry, J. Skinner, M. Smith, C. Strickland, S. Sutton, R. Swindell, H. Tait, J. Tang, A. Thomson, C. Thun, B. Uppill, K. Wall, J. Watkins, T. Watson, L. Webber, A. Whiting, J. Wilford, T. Wilson, A. Wygralak.
2S. Albanese, M. Andersson, A. Arnoldussen, R. Baritz, M.J. Batista, A. Bel‐lan, M.
Birke, D. Cicchella, A. Demetriades, E. Dinelli, B. De Vivo, W. De Vos, M. Duris, A. Dusza-Dobek, O.A. Eggen, M. Eklund, V. Ernstsen, P. Filzmoser, T.E. Finne, D. Flight, S. Forrester, M. Fuchs, U. Fugedi, A. Gilucis, M. Gosar, V. Gregorauskiene, A. Gulan, J. Halamić, E. Haslinger, P. Hayoz, G. Hobiger, R. Hoffmann, J. Hoogewerff, H. Hrvatovic, S. Husnjak, L. Janik, C.C. Johnson, G. Jordan, J. Kirby, J. Kivisilla, V. Klos, F. Krone, P. Kwecko, L. Kuti, A. Ladenberger, A. Lima, J. Locutura, P. Lucivjansky, D. Mackovych, B.I. Malyuk, R. Maquil, M. McLaughlin, R.G. Meuli, N. Miosic, G. Mol, P. Négrel, P. O'Connor, K. Oorts, R.T. Ottesen, A. Pasieczna, V. Petersell, S. Pfleiderer, M. Poňavič, C. Prazeres, U. Rauch, C. Reimann, I. Salpeteur, A. Schedl, A. Scheib, I. Schoeters, P. Sefcik, E. Sellersjö, F. Skopljak, I. Slaninka, A.Šorša, R. Srvkota, T. Stafilov, T. Tarvainen, V. Trendavilov, P. Valera, V. Verougstraete, D. Vidojević, A.M. Zissimos, Z. Zomeni. 0012-821X Crown Copyright © 2012 Published by Elsevier B.V.
doi:10.1016/j.epsl.2011.12.033
Contents lists available atSciVerse ScienceDirect
Earth and Planetary Science Letters
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e p s l
Open access under CC BY-NC-ND license.
Kabata-Pendias, 2001; Vinogradov, 1954). It is debatable how repre-sentative these values are of real soils from large and varied regions, whole continents, or indeed all continents. In addition, the samples behind the estimates often were analysed at different times, in differ-ent laboratories and using differdiffer-ent analytical techniques, and are thus scarcely comparable.
Though this approach may still be useful to obtain reasonable estimates of the total concentrations of major elements, it is ques-tionable whether it provides reliable values for trace elements. Furthermore, in environmental sciences it is not always realised that the average world soil or continental crust values provided in the literature are based on total concentrations, while soil guid-ance values or action levels are generally defined for element con-centrations in aqua regia extractions (ISO, 1995; USEPA, 1996). For instance aqua regia extraction is widely used and is recommended for the analysis of solid materials in Europe (e.g.,BBodSchV, 1999; Hjelmar and Holm, 1999; Langenkamp et al., 2001; REACH, 2008; Rodríguez Martín et al., 2006; Twardowska, 2004) and in Asia (Jung and Osako, 2009; Oh et al., 2010, 2011), and of sampling media for mineral exploration in Australia (e.g.,Hamlyn, 2011). Reli-able values for trace element concentrations in aqua regia extraction at the continental-scale have, until now, not been available, but must be expected, for some elements at least, to be very different from the total concentrations. This is because aqua regia only has a limited effect on minerals such as phlogopite (Mg-rich mica), diocta-hedral mica (muscovite, sericite), quartz, feldspar, plagioclase, am-phibole, barite, cassiterite, chromite, gahnite, garnet, ilmenite, monazite, rutile, sphene and zircon (Chen and Ma, 2001; Dolezal et al., 1968; Foster, 1973; Hamlyn, 2011; Klassen, 2001; Räisänen et al., 1992; Ryan et al., 2002; Tarvainen, 1995).
Here, new soil data collected and analysed to consistent protocols from two continents, one in the northern hemisphere (Europe) and one in the southern hemisphere (Australia), are presented and com-pared. The data come from continental-scale geochemical mapping programmes where a large number of samples were collected accord-ing to detailed and documented protocols, and analysed followaccord-ing a tight external quality control scheme. Internal project analytical stan-dards were exchanged between the two projects and they, as well as international Certified Reference Materials, were analysed with the same techniques to guarantee comparability of analytical results be-tween the two continents and estimation of bias, as outlined in
Reimann et al. (2012).
The geology, geomorphology and pedology of Europe and Austra-lia are complex subjects worthy of detailed discussions far beyond what can be covered in a brief article. Nevertheless, the most salient and relevant characteristics of, and differences between, these two continents (or at least the large parts thereof under study here) are summarised in the following. Major geological provinces in Europe are, in decreasing order of prevalence, (1) extended continental crust, (2) shield, and (3) orogen; in Australia, they are (1) shield, (2) platform and (3) orogen (USGS, 2011). The most common lithol-ogies in Europe are (1) plutonic and metamorphic, (2) shales, and (3) carbonate rocks; in Australia, they are (1) shales, (2) sand, and (3) plutonic and metamorphic rocks (Amiotte Suchet et al., 2003). Gen-erally speaking, fresher rock exposures are more common in Eu-rope, especially northern EuEu-rope, than in Australia, mainly because of the more recent last glaciation in Europe (Holocene, ca 20 ka) compared to Australia (Early Permian, ca 290 Ma), but also because Australia has remained tectonically relatively stable for tens or even possibly hundreds of millions of years allowing weathering under varying climates to affect surface materials both extensively and deeply in many places (BMR Palaeogeographic Group, 1990; Gale, 1992; Pillans, 2007; Veevers, 1984). Both continents span a range of present-day climate zones (Europe: from polar to arid, domi-nantly temperate; Australia from tropical to temperate, domidomi-nantly arid; Peel et al., 2007) and ecoregions (Europe: from tundra to
Mediterranean forests, woodlands and scrub, dominantly temperate broadleaf and mixed forests; Australia from tropical and subtropical grasslands, savannas and shrublands to Mediterranean forests, woodlands and scrub, dominantly deserts and xeric shrublands;
Olson et al., 2001). Soils vary enormously across such diverse set-tings (Europe: from spodosols dominating in the north, through alfisols in the centre, to inceptisols in the south; Australia from ulti-sols and inceptiulti-sols in the north, through vertiulti-sols, entiulti-sols and ari-disols in the centre, to alfisols in the southeast and southwest;
USDA, 2005). Given this range of conditions, and the differences be-tween these two continents, it should be instructive to compare em-pirical soil geochemistry data from Europe and Australia with world soil reference values and investigate if any observed differences can be related back to these conditions.
Thus, the present study aims to address the following questions: 1. How do extensive and consistent empirical datasets from two new
continental-scale surveys compare with world soil values? 2. Can lithology be used as a predictor of soil chemical composition? 3. Can this work covering two continents provide improved world soil reference values, including– for the first time – aqua regia extractable concentrations for several elements?
4. What is needed to obtain a robust estimate of global soil composition?
2. Methods
During the last four years, continental-scale geochemical surveys have been conducted in Europe and Australia covering 5.6 and 6.2 million km2, respectively. These are the Geochemical Mapping of
Agricultural Soils (GEMAS) and the National Geochemical Survey of Australia (NGSA;www.ga.gov.au/ngsa) projects, briefly described below. Average sampling densities were 1 site/2500 km2for GEMAS
and 1 site/5200 km2for NGSA. The GEMAS project sampled
agricul-tural soils (hereafter referred to as‘Ap’ samples for A ploughed hori-zon), whereas the NGSA project focused on soils developed on catchment outlet sediments generally similar to floodplain sedi-ments. The Ap samples from GEMAS (N = 2211) were taken as com-posites of 0 to 20 cm depth, air-dried and sieved tob2 mm using nylon mesh sieves. The NGSA samples considered here (N = 1315) are the Top Outlet Sediments (TOS) collected as composites from 0 to 10 cm depth, oven-dried at 40 °C and sieved tob2 mm (or ‘coarse’ as the project also used a b75 μm ‘fine’ fraction) using nylon mesh sieves (hereafter referred to as‘Tc’ for TOS coarse). Details re-lating to survey design, sample collection, and preparation are found inEGS (2008) for GEMAS and in Caritat et al. (2009) and
Lech et al. (2007)for NGSA.
In both projects, samples were analysed for an extensive suite of total and aqua regia soluble element contents, as well as for other parameters (Caritat and Cooper, 2011; Caritat et al., 2010; Reimann et al., 2009, 2011a). In both the GEMAS and NGSA projects, total major element contents were obtained by X-Ray Fluorescence (XRF) for Al2O3, CaO, Fe2O3, K2O, MgO, MnO, Na2O, P2O5, SiO2and TiO2.
Total trace element contents (As, Ba, Ce, Co, Cr, Ga, Nb, Ni, Pb, Rb, Sr, Th, V, Y, Zn and Zr) were determined by XRF for the GEMAS sam-ples and by total digestion (HF + HNO3digestion of fused XRF bead)
followed by ICP-MS analysis for the NGSA samples. The aqua regia extracted elements (Ag, As, Bi, Cd, Ce, Co, Cs, Cu, Fe, La, Li, Mn, Mo and Pb) were determined in both cases by a similar aqua regia digestion followed by ICP-MS analysis. Further details are provided inReimann et al. (2012).
Early on, Internal Project Standards (IPSs) were exchanged be-tween GEMAS and NGSA to allow demonstration of inter-comparability between both geochemical datasets despite the minor differences in analytical protocols discussed above. These IPSs were (1) a representative agricultural soil ‘GEMAS-Ap’ and (2) a
representative grazing land soil‘GEMAS-Gr’ from Europe (EGS, 2008), and (3) a representative catchment outlet sediment‘ORIS’ (Ovens River Internal Standard) from Australia (Caritat and Cooper, 2011). In addition, an IPS from another continental-scale geochemical sur-vey,‘SoNE-1’ (Soil of Nebraska-1) from the North American land-scape geochemical survey (Smith et al., 2009) was also analysed in the analytical streams of GEMAS and NGSA. Reimann et al. (2012)discuss in further detail the comparability of the continental soil datasets from GEMAS and NGSA, and conclude that the two sampling media are indeed comparable at this scale for the elements under consideration here. Based on the comparison of results for the exchanged IPSs results for all elements/parameters presented here were found to be directly comparable between the two continents (seeReimann et al., 2012, for further detail).
3. Results and discussion
Table 1shows the composition of the upper continental crust, several published global or world soil concentration values, average or median soil composition from other available continents, and the median total topsoil composition from the new European and Australian continental-scale surveys (GEMAS Ap and NGSA Tc).
Starting with Clarke and Washington's (1924) famous “Clarke values,”Taylor and McLennan (1995),Wedepohl (1995)and, more
recently,Rudnick and Gao (2003) have all put considerable effort into estimating the average geochemical composition of the conti-nental crust for almost all elements of the periodic table based on models of the proportions of major rock types making up the crust and the existing knowledge about their average chemical composi-tion. The latest values of Rudnick and Gao (2003) are shown in
Table 1for those elements that are considered in this paper. Published estimates of the chemical composition of global (Vinogradov, 1954) or world soils (Bowen, 1979; Kabata-Pendias, 2001; Koljonen, 1992), shown inTable 1, are up to now based on (1) results of relatively small soil surveys over different lithologies, (2) estimates of the distribution of different rock types in the Earth crust, and (3) the existing knowledge about soil-forming processes in different climatic settings. As such, many of these values can fair-ly be regarded as representing“expert guesses.” In the next sections, we will discuss these world estimates in light of continental data-sets, and compare these datasets to other available results from sim-ilar soil surveys.
Amiotte Suchet et al. (2003)presented a database of worldwide lithology based on six major rock types: (1) sand and sandstones, (2) carbonate rocks, (3) shales, (4) plutonic and metamorphic (i.e., shield) rocks, (5) acid volcanic rocks, and (6) basalts with a 1° × 1° resolution. Table 2shows the proportions of these rock types for the whole world (Amiotte Suchet et al., 2003) and those we derived from their data for the three continental-sized study areas that will
Table 1
Average (AVE) upper continental crust (UCC), several published median (MED) global soil (GS) and world soil (WS) reference values, and average or median values from continental-scale geochemical survey data from the USA and China (as described in the footnote) compared to the new median soil total concentrations determined by continental-scale geochemical surveys in Europe (GEMAS Ap) and Australia (NGSA Tc) (this study).
UCC GS WS1 WS2 WS3 USA USA China Ap Tc
AVE MED MED MED MED AVE MED AVE MED MED
Majors— total (wt%) Al2O3 15.4 13.5 15.1 13.4 13.6 9.7 12.3 10.5 8.1 CaO 3.6 1.9 2.0 2.1 3.4 1.4 3.1 1.2 0.5 Fe2O3 5.6 5.4 5.0 5.7 3.7 2.8 4.4 3.6 3.2 K2O 2.8 1.6 1.7 1.7 1.8 1.8 2.3 1.9 1.2 MgO 2.5 1.0 1.5 0.8 1.5 1.0 1.5 1.0 0.5 MnO 0.10 0.11 0.07 0.13 0.07 0.07 0.05 0.08 0.08 0.04 Na2O 3.27 0.85 1.35 0.67 1.62 1.10 1.52 0.79 0.30 P2O5 0.15 0.18 0.17 0.18 0.10 0.15 0.18 0.06 SiO2 66.6 70.6 59.9 70.6 66.3 66.8 77.5 TiO2 0.64 0.77 0.67 0.83 0.48 0.42 0.68 0.62 0.58 Traces— total (mg/kg) As 5 5 5 6 5 7 6 7 3 Ba 624 500 500 500 362 580 502 391 315 Ce 63 (50) 65 50 49 75 59 42 Co 17 8 10 8 7 9 7 13 9 8 Cr 92 200 80 70 42 54 50 58 64 48 Ga 18 20 12 17 15 12 10 Nb 12 12 10 12 11 13 9 Ni 47 40 20 50 18 19 15 25 21 15 Pb 17 10 17 35 25 19 17 25 21 13 Rb 84 60 65 150 50 67 75 51 Sr 320 300 240 250 147 240 148 186 102 68 Th 11 6 9 9 8 9 9 8 V 97 100 90 90 60 80 67 78 70 55 Y 21 20 40 12 25 23 28 21 Zn 67 50 70 90 62 60 52 69 62 31 Zr 193 300 230 400 300 230 188 263 304
UCC: Upper continental crust, fromRudnick and Gao (2003). GS: Global soil, fromVinogradov (1954).
WS1: World soil, fromBowen (1979). WS2: World soil, fromKoljonen (1992). WS3: World soil, fromKabata-Pendias (2001).
USA: Conterminous USA soils, AVE fromShacklette and Boerngen (1984), and MED fromGarrett (2009)orGustavsson et al. (2001). China: Chinese soil, fromLi and Wu (1999).
be further discussed in this paper. This database of world lithologies was used in preference to either local continental lithological cover-ages (where they exist, e.g.,Raymond, 2009) or recently developed world lithological coverages withfiner resolution (Dürr et al., 2005). This is because we are here focussing on how soil composition over two continents compares to other continents and indeed to world average soil composition. Thus we need a consistent litho-logical database at an appropriate resolution. Future work may in-vestigate the more detailed relationships between lithologies and soil composition within Europe or Australia, but this is outside the scope of the present contribution.
Amiotte Suchet et al. (2003)argued that“The main difficulty in constructing a global data set of rock type exposures on the conti-nents is that the information given by geological maps is inadequate. Indeed, geological maps focus on the age of rocks (for sedimentary rocks), on their deformation and on their structural position (sedi-mentary basin, mountain range, etc.), but information concerning the chemical and physical nature of rocks is often insufficient.” We couldn't agree more. We will examine if differences in soil composi-tion on different continents (Table 1) reflect different rock propor-tions (Table 2), and discuss whether these six rock types provide a representative basis for a good estimate of soil chemical composition at the continental, and ultimately, global scale.
3.1. Comparison with world soil reference values
A detailed comparison of the Australian and European continental-scale data was presented inReimann et al. (2012)and here we focus on how these new data compare with global soil ref-erence values from previous workers. Most estimates provided for global or world soil averages (GS, WS1 to WS3 in Table 1) are quite similar to one another; our empirical results from two continental-scale soil surveys do, however, show some deviations relative to these estimates. For Al2O3the estimated global soil values
(13.4 to 15.1 wt.%) are much higher than the values presented here for Europe (10.5 wt.%) and Australia (8.1 wt.%). Rauch (2011)
reported a geospatially weighed mean global soil Al concentration of 3.9 wt.% (7.4 wt.% Al2O3), which is lower than the data
pre-sented here but certainly supports our observation that world soil estimates inTable 1may be overestimating soil Al content.
CaO in world soil estimates (1.9 to 2.1 wt.%) is also much higher than empirical data from continental-scale data for either Europe (1.2 wt.%) or Australia (0.45 wt.%). The higher estimate is not backed by a higher proportion of carbonates (Table 2) for the world com-pared to Europe. The same situation is found for Fe2O3, where global
soil values (5 to 5.7 wt.%) are significantly above what is measured in Europe (3.6 wt.%) and Australia (3.2 wt.%).Rauch's (2011)soil Fe concentration of 2.5 wt.% (3.6 wt.% Fe2O3) is consistent with our data.
Median K2O from Australia (1.2 wt.%) is lower than the global soil
estimates (1.6 to 1.7 wt.%), while the European value (1.9 wt.%) is only slightly above. For MgO, MnO, and Na2O, the European soil
me-dians (1, 0.08, and 0.79 wt.%) fall within the reported global soil range (0.8 to 1.5 wt.%, 0.07 to 0.13, and 0.67 to 1.35 wt.%, respective-ly), but the Australian data are all below that range (0.5, 0.04, and 0.3 wt.%). Median for SiO2in Europe (66.8 wt.%) is within the global
soil range (59.9 to 70.6 wt.%), but the Australian median soil value is significantly higher than this range (77.5 wt.%). In terms of TiO2,
the new continental data from Europe (0.62 wt.%) and Australia (0.58 wt.%) closely agree and suggest that the previously published global soil estimates (0.67 to 0.83 wt.%) may be too high.
For total trace elements, the European median for As (7 mg/kg) is above, and the Australian median (3 mg/kg) below, the available world soil values (5 to 6 mg/kg). The Australian median value for Ba (315 mg/kg) is well below the world soils range (362 to 500 mg/ kg). For Ce and Ga, the median soil values from Australia (42 mg/kg and 10 mg/kg) are below the global soil ranges (49 to 65 mg/kg and
12 to 20 mg/kg). With a median of only 15 mg/kg of Ni, the Australian soils are again below the published soil range (18 to 50 mg/kg). The world values for Sr (147 to 300 mg/kg) appear to be significantly above empirical data from both Europe (102 mg/kg) and Australia (68 mg/kg). The Australian data suggest lower V (55 mg/kg) and much lower Zn (31 mg/kg) content than global soil estimates (60 to 100 mg/kg and 50 to 90 mg/kg). For Co, Cr, Nb, Pb, Rb, Th, Y and Zr, the new European and Australian median values fall within the previ-ously published world soil range of values.
3.2. Comparison with data from other continental scale surveys
Table 1also compares the new European and Australian median values to the median or average values from two other continental-scale surveys, the conterminous USA (Garrett, 2009; Gustavsson et al., 2001; Shacklette and Boerngen, 1984) and China (Li and Wu, 1999). Afirst observation is that several authors of large geochemical datasets provide average values, i.e., arithmetic means, while other authors have chosen to use median values as the more representative value for the central tendency of geochemical datasets. The different methods of estimating a central tendency for a given dataset are dis-cussed in detail inReimann et al. (2008), who concluded that for geochemical data the median is the preferred statistical measure. Geochemical data are closed data (Aitchison, 1986) and do not plot in the Euclidian space, however, the calculation of the mean is based on Euclidian distances (Filzmoser et al., 2009). Furthermore, the mean will invariably yield biased (typically high) values for skewed data. The comparability of these datasets is thus limited from the outset due to fundamental differences between the statisti-cal estimators.
Still, some striking differences can be observed between these datasets. The largest difference observed between continental median values is for Na2O. The 1.4 and 3.6 times higher Na2O content of
American soils (median 1.10 wt.%) relative to European (0.79 wt.%) and Australian (0.30 wt.%) soils can not be explained by a coastal/sea-spray influence, since the conterminous USA has a lower coastal length (~10,000 km for all States except Alaska) than Australia (26,000 km) or the GEMAS area (91,000 km for EU plus Norway) (CIA, 2008). The USA does not have a greater abundance of exposed Na-silicate or salt-bearing rock types either (see sandstones, shales and plutonic/metamorphic rocks inTable 2), with the exception of acid volcanics, which are considerably higher in the USA (14%) com-pared to Europe (0.5%) and Australia (3.7%). These rocks contain Na-feldspars as shown byEberl and Smith (2009; their Fig. 9). Still, this does not satisfactorily explain why the order of median Na2O soil
content is USA > Europe≫Australia. Thus, the explanation likely lies at least in part in differences in weathering regimes being re-sponsible for the Na distributions observed, in particular the leach-ing of this soluble cation from Australia's mature soils.
The CaO content of soil is marginally higher in the USA (1.4 wt.%) than in Europe (1.2 wt.%) but 3.1 times higher than in Australia (0.5 wt.%). This is in overall agreement with the lithological distribu-tion (Table 2), which suggests that the USA study area has more car-bonate rocks (19%) than the European (14%) and especially the Australian (4%) study areas. The median MgO contents for the three continents follow a similar pattern of USA (0.99 wt.%) ~ Europe (0.95 wt.%) > Australia (0.5 wt.%) and is probably related to the car-bonate distribution as well. Of course, Ca and Mg are not exclusively sourced from carbonate lithologies, and calc- and mafic-silicates also release these cations during hydrolysis. In many ways, it was expected that Australia, with its predominantly semi-arid to arid climate and relatively common pedogenic carbonates, which are reflected in the distribution of soil pH (Caritat et al., 2011), would have a higher median CaO soil content than the more temperate con-tinents, but this is not the case. Instead the soil CaO and MgO contents
are approximately proportional to the carbonate rock proportions for these three study areas, namely USA ~ Europe = 2−3×Australia.
North American soils contain less Fe2O3(2.8 wt.%) than Australian
soils (3.2 wt.%), which are themselves below the European value (3.6 wt.%). If Fe2O3abundance is related mainly to the abundance of
ba-saltic rocks, then one would expect a ranking for the conterminous USA (1.5% basalt) somewhere between Europe (0.5%) and Australia (4%), but this is not the case. It must be noted here that the original American survey ofShacklette and Boerngen (1984)deliberately sampled sub-soils at a depth >20 cm to avoid possible effects of surface contami-nation; this could explain some of the observed differences, particularly for Mn and Fe oxides. This observation exemplifies how difficult it is to arrive at really comparable soil values for any large area. Small differences in sample medium, sample preparation or an-alytical protocols can have substantial effects on the observed ele-ment concentrations and render such datasets unfit for comparison, unless measures to ensure comparability, like exchanging IPSs and checking the results for mutual consistency, are implemented.
China has a much higher Al2O3content (12.3 wt.%) than Europe
(10.5 wt.%), USA (9.7 wt.%) or especially Australia (8.1 wt.%), but this could be partly due to the fact that it is an average value rather than a median (see discussion above).
It is perhaps for K2O that soil composition is most obviously not
reflecting rock composition: whereas the order of decreasing K-silicate-bearing shale abundance is Australia (44%) > Europe (37%) > USA (23%), the order of K2O abundance is Europe (1.9 wt.%) ~ USA
(1.8 wt.%) >Australia (1.2 wt.%). Of course K2O can also originate from
feldspars and micas in sandstones and granites, and here Australia has similar sandstone and plutonic/metamorphic abundances (24 and 20%) to the USA (23 and 19%). Europe has a low proportion of sand-stone (10%) but a much higher proportion of plutonic/metamorphic (39%) rocks. For K2O, this simple lithology schema does not
satisfac-torily explain the observed distributions either. As for Na2O above,
it is likely that weathering plays a dominant role in the abundance of K2O, also a relatively mobile element in soils, on those three
continents.
As far as total trace elements concentrations are concerned, the largest difference between the continents is observed for Sr, where the median for the USA (148 mg/kg) is over 2.1 and 1.4 times greater than for Australia (68 mg/kg) and Europe (102 mg/kg), respectively. This is consistent with the similarfindings regarding Ca above, and to some extent also Mg, and is related to the preponderance of car-bonate lithologies there.
For As, the USA soils (5.6 mg/kg) contain 1.8 times the median content in Australia (3.1 mg/kg) but are somewhat lower than in Europe (7 mg/kg). Zn follows the same pattern with the USA soils (52 mg/kg) falling between the Australia (31 mg/kg) and European (62 mg/kg) medians. Both As and Zn are more abundant in shale and schist than in any other common rock type (Reimann and Caritat, 1998), so lithology fails to explain the As distribution in soil as Australia has considerably more abundant shale than either Europe or the USA (Table 2). For Zn, another rich source can be
found in basaltic rocks, which are more common in Australia than in the USA, and very rare in the GEMAS area, thus not provid-ing a satisfactory explanation for the high soil Zn values beprovid-ing found in Europe.
Median Ba in the USA (502 mg/kg) is 1.6 and 1.3 times greater than in Australia (315 mg/kg) and Europe (391 mg/kg), respectively. Median Pb in the USA (17 mg/kg) is intermediate between Europe (21 mg/kg) and Australia (13 mg/kg). Both Ba and Pb are abundant in feldspar, plagioclase and mica, which are main constituents of granite and shale, lithologies that are more represented in both Australia and Europe than in the USA. Thus, like for K2O, the simple
rock type schema does not explain the levels of these trace elements observed in soil from these three continents.
For the geochemically‘inert’ Zr, which is most abundant in grey-wacke and sandstone, the median soil composition (Australia 304 mg/kg > Europe 263 mg/kg > USA 188 mg/kg) is also inconsistent with the relative proportions ofAmiotte Suchet et al.'s (2003) sand-stone category (Australia 24% ~ USA 23%≫Europe 9.5%).
In terms of total Cr concentration, the soils of Australia (48 mg/kg) and Europe (64 mg/kg) are respectively similar to, and more ele-vated than, those from the USA (50 mg/kg). For Ni, the soils of Aus-tralia and the USA have the same median value of 15 mg/kg, much lower than for Europe (21 mg/kg). Since both Cr and Ni are consid-erably more common in ultramafic and mafic rocks than in other rock types (e.g.,Reimann and Caritat, 1998), one would expect a strong control being exerted on their soil concentration by the abun-dance of these lithologies. However, here too, the distribution of basalt lithology, which is Australia≫USA>Europe (Table 2), falls short of explaining soil composition as revealed by continental-scale sampling. One problem with the database ofAmiotte Suchet et al. (2003)may be the“plutonic and metamorphic rocks” category con-taining rocks of very different chemical composition. For example, greenstone belts in Scandinavia, with their high Ni and Cr values, fall into this category although, based on geochemical considerations, they would probably be better classified as “basalts.”
The data presented herein for soil composition from two conti-nents and compared to a third continent, thus appear to indicate that parent material, i.e., the type of rock substrate, is a poor predictor of soil composition at this scale, and/or that the existing geological databases on the spatial distribution of rock types (as opposed to rock age) are deficient.
Parameters other than rock composition of course play a signi fi-cant role in soil formation by modifying the chemical composition inherited from the parent material during pedogenesis (i.e., climate, organisms, relief, time; seeJenny, 1941; Shaw, 1930). Only now that continental geochemical surveys and world lithological databases are becoming available, can it be shown how strong continental-scale climate is as a modifier of the chemical signal inherited from parent material. In the case of Europe, the southerly extent of the ice sheet during the last (Quaternary) glaciation has a profound im-pact on geochemical patterns expressed in soils (Reimann et al., 2012). This is a result of scraping clean the Baltic shield and exposing fresh plutonic and metamorphic rocks at the surface, as well as leav-ing behind coarse glacial sediment consistleav-ing of crudely ground bed-rock material, with a low capacity for adsorbing trace elements. Areas receiving unusual amounts of rainfall also often show specific geo-chemical signals. In Australia, the influence of past climates is ex-pressed by the extensive and, in places, intensive, weathering that has occurred at least episodically probably since the Paleozoic, and through the Mesozoic and Cenozoic (BMR Palaeogeographic Group, 1990; Gale, 1992; Li and Vasconcelos, 2002; Nott, 1995; O'Sullivan et al., 2000; Pillans, 1997, 2007). This has resulted in an intense leach-ing of mobile cations released by the breakdown of rock-formleach-ing minerals, with a subsequent impoverishment of these elements in the soil, and concomitantly an increase in levels of elements held in weathering resistant mineral phases. Weathering end-member Table 2
Relative areal proportions of six major lithological types as presented byAmiotte Suchet et al. (2003)for the world, and those recalculated from their data for the project areas of NGSA and GEMAS (this study) as well as the conterminous USA (Shacklette and Boerngen, 1984) denoted as‘USA48’. ‘SAND’ stands for sand and sandstones, ‘CARB’ carbonate rocks, ‘SHAL’ shales, ‘PLME’ plutonic and metamorphic (i.e., shield) rocks,‘ACVO’ acid volcanic rocks, and ‘BASA’ basalts.
SAND CARB SHAL PLME ACVO BASA TOT WORLD 20% 13% 28% 31% 2.4% 6.0% 100% NGSA 24% 4.3% 44% 20% 3.7% 4.0% 100% GEMAS 9.5% 14% 37% 39% 0.5% 0.5% 100% USA48 23% 19% 23% 19% 14% 1.5% 100%
minerals and regolith types are commonly reported to be enriched in Si (quartz, clay minerals; silcrete), Al (illite, smectite, kaolinite and its polymorphs, gibbsite; bauxite), Fe (goethite, hematite; laterite), Zr (zircons and other resistates; heavy mineral sands) and Ti (ilmenite, rutile, anatase; silcrete) (Dixon and Weed, 1977; Dixon and Young, 1981; Gilkes and Suddhiprakarn, 1979; Kronberg and Nesbitt, 1981; Nesbitt and Markovics, 1997; Taylor et al., 1992; Wilson, 1999; Zeissink, 1969). Judging from the new national-scale soil dataset and especially comparing it to Europe and the USA, it appears that the only elements in which the Australian samples are enriched are Si and Zr, and not Al or Fe or Ti as could have been expected. This raises questions about the real effects of weathering at this scale and the perhaps greater than hitherto speculated mobility of Al, Fe and Ti. A confounding factor, however, is that bauxites, laterites and heavy mineral sands were not targeted sampling media in Australia (but nor were they in Europe and the USA).
3.3. The Predicted Empirical Global Soil reference values based on two continents (PEGS2)
It follows from the above discussion that published world or global soil reference values have some similarities but also some im-portant differences when compared to empirical soil geochemical data obtained from surveys of whole continents. This is attributed to the fact that these reference values are calculated often starting from an upper crust composition, which is a calculated estimate based on geophysical and some geochemical models. It is clear from
Table 1that the upper continental crust chemistry is at best a very crude estimator or predictor of soil composition. For instance, all major elements except SiO2and TiO2are overestimated by between
25% (P2O5) and 600% (Na2O) compared to the median of soil
concen-trations measured over large areas such as Europe and Australia. The situation is similar for the total trace elements reported: only As, Ce, Nb, Pb and Y are close to (within 25% of) crustal values; Th, Rb, Zn, V, Ga, Cr, Ba, Co, Ni, Sr are overestimated up to nearly four-fold, and Zr is underestimated by 68%.
InTable 3an improved, but interim, reference value for world soils based on the median values for the data from the two new continental-scale soil geochemical datasets from both hemispheres is presented. This value is called the Preliminary Empirical Global Soil value based on two continents, or PEGS2; it is calculated as the median value between the GEMAS and NGSA medians. In the future as more (inter-comparable) continental datasets become available, a PEGS3, PEGS4, etc., can be put forward in the literature, by similarly taking medians of the three, four, etc., values available. Apart from presenting data for total major and trace element contents, the PEGS2 values include for the veryfirst time a global soil estimate for aqua regia extractable elements. Note that a strict quality control pro-gram was implemented in the GEMAS and NGSA projects to allow the resulting data to be demonstrably comparable (Reimann et al., 2012).
The data for trace elements in world soil provided by Kabata-Pendias (2001)come closest (Table 1) to the PEGS2 (Table 3) esti-mates provided here and based on empirical results from two continental-scale soil surveys. Most interesting deviations between PEGS2 and the values provided byKabata-Pendias (2001)are clearly higher PEGS2 values for Y (by a factor of ~2), Cr and Rb, and lower PEGS2 values for Sr, Pb and Zn.
4. Conclusions
New geochemical surveys of soils in Europe (N= 2211) and Austra-lia (N= 1315), covering 5.6 and 6.2 million km2 respectively, have
been carried out in such a way as to be demonstrably comparable. Using Internal Project Standards and analysing Certified Reference Materials it was possible to demonstrate comparability between both
datasets as well as acceptable precision and bias for all 10 major oxides (Al2O3, CaO, Fe2O3, K2O, MgO, MnO, Na2O, P2O5, SiO2and TiO2) as
well as for Loss On Ignition (LOI) and pH, for 16 total trace elements (As, Ba, Ce, Co, Cr, Ga, Nb, Ni, Pb, Rb, Sr, Th, V, Y, Zn and Zr), and also for 14 aqua regia soluble elements (Ag, As, Bi, Cd, Ce, Co, Cs, Cu, Fe, La, Li, Mn, Mo and Pb).
For many elements there exist substantial differences between published world soil averages and the median concentrations ob-served on two continents. Further, average crustal values are not related to average soil values. They should not be used for providing background or reference values for soils. Only real soil analyses from very large areas provide a reasonable estimate of average world soil composition.
Table 3
Median (MED) soil values determined by continental-scale geochemical surveys in Eu-rope (GEMAS Ap) and Australia (NGSA Tc), and the derived Preliminary Empirical Global Soil estimate based on two continents (PEGS2).
Ap Tc PEGS2 MED MED Majors— total (wt%) Al2O3 10.5 8.1 9.3 CaO 1.2 0.5 0.8 Fe2O3 3.6 3.2 3.4 K2O 1.9 1.2 1.6 MgO 1.0 0.5 0.7 MnO 0.08 0.04 0.06 Na2O 0.79 0.30 0.55 P2O5 0.18 0.06 0.12 SiO2 66.8 77.5 72.2 TiO2 0.62 0.58 0.60 Traces— total (mg/kg) As 7 3 5 Ba 391 315 353 Ce 59 42 51 Co 9 8 9 Cr 64 48 56 Ga 12 10 11 Nb 13 9 11 Ni 21 15 18 Pb 21 13 17 Rb 75 51 63 Sr 102 68 85 Th 9 8 8 V 70 55 63 Y 28 21 25 Zn 62 31 47 Zr 263 304 284
Traces— aqua regia (mg/kg) Ag_AR 0.038 0.012 0.025 As_AR 5.7 1.6 3.6 Bi_AR 0.17 0.12 0.15 Cd_AR 0.18 0.04 0.11 Ce_AR 29 29 29 Co_AR 7.8 6.3 7.0 Cs_AR 1.1 0.8 1.0 Cu_AR 15 11 13 Fe_AR 17,632 16,500 17,066 La_AR 15 14 14 Li_AR 12 5.7 8.7 Mn_AR 462 279 370 Mo_AR 0.42 0.20 0.31 Pb_AR 15 7 11 Other parameters LOI (%) 8.7 5.5 7.1 pH 5.8 5.7 5.7
Using a database on the worldwide distribution of lithological units it can be demonstrated that there is little relation between the spatial distribution of these units and the observed chemistry of the soils. Perhaps only CaO and MgO in Australian soils are to a degree commensurate with the abundance of carbonate and mafic litholo-gies, and SiO2with the abundance of sandstone. Even median Al2O3,
Fe2O3, TiO2and MnO are lower in Australia than in Europe and/or
USA, suggesting a greater solubility than perhaps hitherto suspected. The poor coincidence between lithology and soil geochemistry at the continental scale may be partly due to the fact that geological maps are based on age relations while good lithological maps of the world are direly missing. Furthermore past and present climates (and climate-related processes) play an often underestimated role in de-termining the chemistry of soils.
To overcome the above shortcomings of predicted global or world soil geochemical reference values, Preliminary Empirical Global Soil reference values based on mutually consistent data from two conti-nents (PEGS2) are proposed to be used for comparison and normali-sation purposes. It is hoped that more continental datasets will be added to this estimate in the near future.
Acknowledgements
The GEMAS project is a cooperation project of the EuroGeoSurveys Geochemistry Expert Group with a number of outside organisations (e.g., Alterra in The Netherlands, the Norwegian Forest and Landscape Institute, several Ministries of the Environment and University De-partments of Geosciences in a number of European countries, CSIRO Land and Water in Adelaide, Australia) and Eurometaux. Very special thanks are due to Ilse Schoeters, who arranged the Expert Group's contact with Eurometaux and made this all possible, and to Robert G. Garrett of the Geological Survey of Canada, who established the first contact between Ilse and the Geochemistry Expert Group's chair-man. The analytical work was co-financed by the following organisa-tions: Eurometaux, Cobalt Development Institute (CDI), European Copper Institute (ECI), Nickel Institute, Europe, European Precious Metals Federation (EPMF), International Antimony Association (i2a), International Manganese Institute (IMnI), International Molybdenum Association (IMoA), ITRI Ltd. (on behalf of the REACH Tin Metal Con-sortium), International Zinc Association (IZA), International Lead Association-Europe (ILA-Europe), European Borates Association (EBA), the (REACH) Vanadium Consortium (VC) and the (REACH) Selenium and Tellurium Consortium. Finally, the Directors of the Eu-ropean Geological Surveys and the additional participating organisa-tions are thanked for making sampling of almost all of Europe on a tight time schedule possible.
The NGSA project was part of the Australian Government's On-shore Energy Security Program 2006–2011, from which funding sup-port is gratefully acknowledged. NGSA was led and managed by Geoscience Australia and carried out in collaboration with the geolog-ical surveys of every State and the Northern Territory under National Geoscience Agreements. The Directors from the Geological Survey of Queensland (GSQ), Geological Survey of New South Wales (GSNSW), GeoScience Victoria (GSV), Mineral Resources Tasmania (MRT), Primary Industries and Resources South Australia (PIRSA), Geological Survey of Western Australia (GSWA) and Northern Terri-tory Geological Survey (NTGS) are warmly thanked for supporting the NGSA. This project would not have been feasible without land-owners granting permission to access the sampling sites. This manu-script was prepared while PdC was a visiting scientist at the Geological Survey of Norway in July–August 2011; he gratefully ac-knowledges the receipt of a Development Award from Geoscience Australia and the hospitality of the Geological Survey of Norway, which made this possible. PdC published with permission from the Chief Executive Officer, Geoscience Australia.
References
Aitchison, J., 1986. The Statistical Analysis of Compositional Data. Chapman & Hall, London. 416 pp.
Amiotte Suchet, P., Probst, J.-L., Ludwig, W., 2003. Worldwide distribution of continen-tal rock lithology: implications for the atmospheric/soil CO2 uptake by continencontinen-tal weathering and alkalinity river transport to the oceans. Global Biogeochem. Cycles 17, 1038–1051.
BBodSchV (Bundes-Bodenschutz- und Altlastenverordnung), 1999. Federal Soil Protection and Contaminated Sites Ordinance. Federal Ministry for the Environment, Nature Conservation and Nuclear Safety, Germany. Available at: http://www. umweltbundesamt.de/boden-und-altlasten/altlast/web1/berichte/pdf/bbodschv-engl. pdf. 64 pp.
BMR (Bureau of Mineral Resources) Palaeogeographic Group, 1990. Australia: Evolu-tion of a Continent. Bureau of Mineral Resources, Canberra.
Bowen, H.J.M., 1979. Environmental Chemistry of the Elements. Academic Press, London. 333 pp.
Brantley, S.L., White, T.S., Ragnasdottir, K.V. (Eds.), 2007. The Critical Zone: where rock meets life: Elements (An International Magazine of Mineralogy, Geochemistry, and Petrology), 3, pp. 307–338.
Caritat, P. de, Cooper, M., 2011. National Geochemical Survey of Australia: Data Quality Assessment. Geoscience Australia Record 2011/21. (2 Volumes), 478 pp. Caritat, P. de, Cooper, M., Lech, M., McPherson, A., Thun, C., 2009. National Geochemical
Survey of Australia: Sample Preparation Manual. Geoscience Australia Record 2009/08. 28 pp.
Caritat, P. de, Cooper, M., Pappas, W., Thun, C., Webber, E., 2010. National Geochemical Survey of Australia: Analytical Methods Manual. Geoscience Australia Record 2010/15. 22 pp.
Caritat, P. de, Cooper, M., Wilford, J., 2011. The pH of Australian soils:field results from a national survey. Soil Res. 49, 173–182.
Chen, M., Ma, L.Q., 2001. Comparison of three aqua regia digestion methods for twenty Florida sandy soils. Soil Sci. Soc. Am. J. 65, 491–499.
CIA (Central Intelligence Agency), . The World Factbook— Coastline . Available at:https:// www.cia.gov/library/publications/the-world-factbook/fields/2060.html.
Clarke, F.W., Washington, H.S., 1924. The composition of the Earth's crust. USGS Profes-sional Paper 127. 117 pp.
Dixon, J.B., Weed, S.B., 1977. Minerals in Soil Environments. Soil Sci. Soc. Am., Madison, Wisconsin. 948 pp.
Dixon, J.C., Young, R.W., 1981. Character and origin of deep arenaceous weathering mantles on the bega batholith, southeastern Australia. Catena 8, 97–109. Dolezal, J., Povondra, P., Sulcek, Z., 1968. Decomposition Techniques in Inorganic
Anal-ysis. English translation edited by In: Hughes, D.O., Floyd, P.A., Barratt, M.S. (Eds.), Elsevier, New York. 224 pp.
Dürr, H.H., Meybeck, M., Dürr, S.H., 2005. Lithologic composition of the Earth's conti-nental surfaces derived from a new digital map emphasizing riverine material transfer. Global Biogeochem. Cycles 19, GB4S10. doi:10.1029/2005GB002515. Eberl, D.D., Smith, D.B., 2009. Mineralogy of soils from two continental-scale transects
across the United States and Canada and its relation to soil geochemistry and cli-mate. Appl. Geochem. 24, 1394–1404.
EGS (EuroGeoSurveys), 2008. EuroGeoSurveys geochemical mapping of agricultural and grazing land in Europe (GEMAS)— Field Manual. Norges Geologiske Under-søkelse Report 2008.038. 46 pp.
Filzmoser, P., Hron, K., Reimann, C., 2009. Univariate statistical analysis of environmen-tal (compositional) data— problems and possibilities. Sci. Total Environ. 407, 6100–6108.
Foster, J.R., 1973. The efficiency of various digestion procedures on the extraction of metals from rocks and rock forming minerals. Can. Inst. Min. Bull. 66, 85–92. Gale, S.J., 1992. Long-term landscape evolution in Australia. Earth Surf. Process.
Land-forms 17, 323–343.
Garrett, R.G., 2009. Relative spatial soil geochemical variability along two transects across the United States and Canada. Appl. Geochem. 24, 1405–1415.
Gilkes, R.J., Suddhiprakarn, A., 1979. Magnetite alteration in deeply weathered adamel-lite. J. Soil Sci. 30, 357–361.
Gustavsson, N., Bølviken, B., Smith, D.B., Severson, R.C., 2001. Geochemical landscapes of the conterminous United States— new map presentations for 22 elements. U.S. Geological Survey Professional Paper 1648. 38 pp.
Hamlyn, P., 2011. Sampling, analysis and quality control. Chapter 11 Field Geologist's Manual, Fifth Edition: The Australasian Institute of Mining and Metallurgy Mono-graph 9. Carlton, Victoria, 479 pp.
Hjelmar, O., Holm, P.E., 1999. Determination of total or partial trace element content in soil and inorganic waste material. Nordtest Technical Report 446, Espoo, Finland. 44 pp.
ISO (International Organization for Standardization), . ISO 11466:1995: Soil quality— extraction of trace elements soluble in aqua regia. Available at:http://www.iso. org/iso/catalogue_detail.htm?csnumber=19418.
Jenny, H., 1941. Factors of Soil Formation— A System of Quantitative Pedology. McGraw Hill, New York. 281 pp.
Jung, C.-H., Osako, M., 2009. Metal resource potential of residues from municipal solid waste (MSW) melting plants. Resour. Conserv. Recycl. 53, 301–308.
Kabata-Pendias, A., 2001. Trace Elements in Soils and Plants. CRC Press, Boca Raton. 413 pp.
Kabata-Pendias, A., Pendias, H., 1984. Trace Elements in Soils and Plants. CRC Press, Boca Raton. 315 pp.
Klassen, R.A., 2001. The interpretation of background variation in regional geochemical surveys— an example from Nunavut, Canada. Geochem. Explor. Environ. Anal. 1, 163–173.
Koljonen, T. (Ed.), 1992. Geochemical Atlas of Finland, Part 2: Till. Geological Survey of Finland, Espoo, Finland. 218 pp.
Kronberg, B.I., Nesbitt, H.W., 1981. Quantification of weathering, soil geochemistry and soil fertility. J. Soil Sci. 32, 453–459.
Langenkamp, H., Düwel, O., Utermann, J., 2001. Trace element and organic matter con-tents of European soils. Progress Report, Joint Research Centre Ispra. Available at:
http://ec.europa.eu/environment/waste/sludge/pdf/heavy_metals_progress_report. pdf.
Lech, M.E., Caritat, P. de, McPherson, A.A., 2007. National Geochemical Survey of Australia: Field Manual. Geoscience Australia Record 2007/08. 53 pp.
Li, J.-W., Vasconcelos, P., 2002. Cenozoic continental weathering and its implications for the palaeoclimate: evidence from 40Ar/39Ar geochronology of supergene K–Mn oxides in Mt Tabor, central Queensland, Australia. Earth Planet. Sci. Lett. 200, 223–239.
Li, J., Wu, G. (Chief Compilers), 1999. Atlas of the Ecological Environmental Geochem-istry of China. Geological Publishing House, Beijing, China, 209 pp.
Nesbitt, H.W., Markovics, G., 1997. Weathering of granodioritic crust, long-term stor-age of elements in weathering profiles, and petrogenesis of siliciclastic sediments. Geochim. Cosmochim. Acta 61, 1653–1670.
Nott, J., 1995. The antiquity of landscapes on the North Australian Craton and the im-plications for theories of long-term landscape evolution. J. Geol. 103, 19–32. O'Sullivan, P.B., Gibson, D.L., Kohn, B.P., Pillans, B., Pain, C.F., 2000. Long-term landscape
evolution of the Northparkes region of the Lachlan Fold Belt, Australia: constraints fromfission track and paleomagnetic data. J. Geol. 108, 1–16.
Oh, S.-Y., Cha, S.-W., Kim, I.-H., Lee, H.-W., Kang, S.-G., Choi, S.-J., 2010. Disposal of heavy metal-contaminated sediment from Ulsan Bay, South Korea: treat-ment processes and legal framework. Water Environ. J. doi:10.1111/j.1747-6593.2010.00231.x
Oh, S.-Y., Yoon, M.-K., Kim, I.-H., Kim, J.Y., Bae, W., 2011. Chemical extraction of arsenic from contaminated soil under subcritical conditions. Sci. Total Environ. 409, 3066–3072.
Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V.N., Underwood, E.C., D'amico, J.A., Itoua, I., Strand, H.E., Morrison, J.C., Loucks, C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W., Hedao, P., Kassem, K.R., 2001. Terrestrial ecoregions of the world: a new map of life on Earth. BioScience 51, 933–938.
Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the Köppen– Geiger climate classification. Hydrol. Earth Syst. Sci. 11, 1633–1644.
Pillans, B., 1997. Soil development at snail's pace: evidence from a 6 Ma soil chronose-quence on basalt in north Queensland, Australia. Geoderma 80, 117–128. Pillans, B., 2007. Pre-Quaternary landscape inheritance in Australia. J. Quat. Sci. 22, 439–447. Räisänen, M.L., Hämäläinen, L., Westerberg, L.M., 1992. Selective extraction and
deter-mination of metals in organic stream sediments. Analyst 117, 623–627. Rauch, J.N., 2011. Global distributions of Fe, Al, Cu, and Zn contained in Earth's derma
layers. J. Geochem. Explor. 110, 193–201.
Raymond, O.L., . Surface geology of Australia 1:1 million scale. Geoscience Australia, Canberra, Australia. Available at: https://www.ga.gov.au/products/servlet/ controller?event=GEOCAT_DETAILS&catno=69455.
REACH (Registration, Evaluation, Authorisation and Restriction of Chemicals), 2008. Guidance on information requirements and chemical safety assessment— Appendix R.7.13-2: environmental risk assessment for metals and metal com-pounds. Guidance for the Implementation of REACH, European Chemicals Agency (ECHA). Available at: http://guidance.echa.europa.eu/docs/guidance_ document/information_requirements_r7_13_2_en.pdf. 74 pp.
Reimann, C., de Caritat, P., GEMAS Project Team, NGSA Project Team, 2012. New soil composition data for Europe and Australia: demonstrating comparability, identify-ing continental-scale processes and learnidentify-ing lessons for global geochemical map-ping. Sci. Total Eniron. 416, 239–252.
Reimann, C., Caritat, P. de, 1998. Chemical Elements in the Environment— Factsheets for the Geochemist and Environmental Scientist. Springer-Verlag. 398 pp.
Reimann, C., Filzmoser, P., Garrett, R.G., Dutter, R., 2008. Statistical Data Analysis Explained. Applied Environmental Statistics with R. Wiley, Chichester. 343 pp.
Reimann, C., Demetriades, A., Eggen, O.A., Filzmoser, P., 2009. The EuroGeoSurveys geo-chemical mapping of agricultural and grazing land soils project (GEMAS)— evalu-ation of quality control results of aqua regia extraction analysis. Norges Geologiske Undersøkelse Report 2009.049.
Reimann, C., Demetriades, A., Eggen, O.A., Filzmoser, P., 2011a. The EuroGeoSurveys geochemical mapping of agricultural and grazing land soils project (GEMAS)— evaluation of quality control results of total C and S, total organic carbon (TOC), cation exchange capacity (CEC), XRF, pH, and particle size distribution (PSD) anal-ysis. Norges Geologiske Undersøkelse Report 2011.043.
Rodríguez Martín, J.A., López Arias, M., Grau Corbí, J.M., 2006. Heavy metals contents in agricultural topsoils in the Ebro basin (Spain)— application of the multi-variate geostatistical methods to study spatial variations. Environ. Pollut. 144, 1001–1012.
Rudnick, R.L., Gao, S., 2003. Composition of the Continental Crust. In: Holland, H.D., Turekian, K.K. (eds-in-chief), Treatise on Geochemistry Volume 3: Rudnick, R.L. (ed.), The Crust, 1–64. Elsevier-Pergamon, Oxford.
Ryan, P.C., Wall, A.J., Hillier, S., Clark, L., 2002. Insights into sequential chemical extrac-tion procedures from quantitative XRD: a study of trace metal partiextrac-tioning in sed-iments related to frog malformities. Chem. Geol. 184, 337–357.
Shacklette, H.T., Boerngen, J.G., 1984. Element concentrations in soils and other surficial materials of the conterminous United States: an account of the con-centrations of 50 chemical elements in samples of soils and other regoliths. U.S. Geological Survey Professional Paper 1270. 105 pp.
Shaw, C.F., 1930. Potent factors in soil formation. Ecology 11, 239–245.
Smith, D.B., Woodruff, L.G., O'Leary, R.M., Cannon, W.F., Garrett, R.G., Kilburn, J.E., Goldhaber, M.B., 2009. Pilot studies for the North American Soil Geochemical Landscapes Project-Site selection, sampling protocols, analytical methods, and quality control protocols. Appl. Geochem. 24, 1357–1368.
Tarvainen, T., 1995. The geochemical correlation between coarse andfine fractions of till in southern Finland. J. Geochem. Explor. 54, 187–198.
Taylor, S.R., McLennan, S.M., 1995. The geochemical evolution of the continental crust. Rev. Geophys. 33, 241–265.
Taylor, G., Eggleton, R.A., Holzhauer, C.C., Maconachie, L.A., Gordon, M., Brown, M.C., McQueen, K.G., 1992. Cool climate lateritic and bauxitic weathering. J. Geol. 100, 669–677.
Twardowska, I., 2004. Assessment of pollution potential from solid waste. In: Twardowska, I., Allen, H.E., Kettrup, A.A.F., Lacy, W.J. (Eds.), Waste Management Series, 4. Elsevier, pp. 173–205.
USDA (United States Department of Agriculture), . Global Soil Regions. Available at:
http://soils.usda.gov/use/worldsoils/mapindex/order.html.
USEPA (United States Environmental Protection Agency), . Method 3050B: acid diges-tion of sediments, sludges and soils. Available at:http://www.epa.gov/wastes/ hazard/testmethods/sw846/pdfs/3050b.pdf.
USGS (United States Geological Survey), . Geologic Province and Thermo-Tectonic Age Maps. Available at:http://earthquake.usgs.gov/research/structure/crust/maps.php. Veevers, J.J. (Ed.), 1984. Phanerozoic Earth History of Australia. Clarendon Press,
Ox-ford. 418 pp.
Vinogradov, A.P., 1954. Geochemie Seltener und nur in Spuren Vorhandener Che-mischer Elemente im Boden. Akademie Verlag, Berlin. 249 pp.
Wedepohl, H., 1995. The composition of the continental crust. Geochim. Cosmochim. Acta 59, 1217–1239.
Wilson, M.J., 1999. The origin and formation of clay minerals in soils: past, present and future perspectives. Clay Miner. 34, 7–25.
Zeissink, H.E., 1969. The mineralogy and geochemistry of a nickeliferous laterite profile (Greenvale, Queensland, Australia). Miner. Deposita 4, 132–152.