Geochemical patterns in the soils of Cyprus

Science of the Total Environment 420 (2012) 250–262

Contents lists available at SciVerse ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Geochemical patterns in the soils of Cyprus David R. Cohen a,⁎, Neil F. Rutherford a, Eleni Morisseau b, Andreas M. Zissimos b a b

School of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney, NSW, 2052, Australia Cyprus Geological Survey, Ministry of Agriculture, Natural Resources and Environment, Lefkonos 1, Lefkosia 1415, Cyprus

a r t i c l e

i n f o

Article history: Received 26 October 2011 Received in revised form 5 January 2012 Accepted 15 January 2012 Available online 11 February 2012 Keywords: Cyprus Soil Mapping Aqua regia INAA Trace elements

a b s t r a c t The soil geochemical atlas of Cyprus is a recent addition to the series of national to continental-scale geochemical mapping programmes implemented over the last two decades for environmental and resource applications. The study has been conducted at the high sampling density of 1 site per 1 km2, with multi-element and multimethod analysis performed on samples of top soil (0–25 cm) and sub soil (50–75 cm) from a grid of over 5350 sites across a major portion of Cyprus. Major and most trace elements display sharp concentration changes across the main geological boundaries but a high degree of spatial continuity and consistency of values within those boundaries. Some elements display one to two orders of magnitude difference in median concentrations between the soils developed over ultramafic or mafic units and those developed over sedimentary rocks or alluvial units. The ratio of aqua regia-extractable to total metal contents provides an indication of the general mineralogical host for a number of trace elements. The majority of soils are near-neutral to alkaline with the small proportion of areas with soil pH b 5 largely restricted to the major Cu deposits. There is strong correlation between top soil and sub soil geochemical values. Where the concentrations of some elements (including Pb, Hg and Sn) are indicative of contamination, the values are typically higher in the top soil samples in these areas. Variations in the concentration of elements with strong redox controls on mobility are linked to changes in sedimentary environment between deep and shallow marine conditions. Some element patterns can be related to the effects of urbanisation and sulphide mining operations; however the dominant control on soil geochemistry is the parent geology and regolith forming processes. The atlas demonstrates the effectiveness of high-density sampling in mapping local to regional-scale features of the geochemical landscape. © 2012 Elsevier B.V. All rights reserved.

1. Introduction A major application for applied geochemistry is the implementation of large-scale geochemical mapping programmes for a variety of purposes. Regional geochemical mapping was originally directed towards mineral exploration (Garrett et al., 2008). In post-mining and industrialised areas, soil and sediment geochemical mapping is now primarily used for environmental purposes, including separation of natural from anthropogenic sources of metals and organics, evaluation of soils for agricultural purposes, environmental management, medical geochemistry and land use classification (Tan, 1989; Plant et al., 2000, 2003; Reimann et al., 2011). Geochemical atlases have potential as a reference to measure spatial and temporal changes in the geochemistry of the natural environment (Darnley et al., 1995) and to assist with understanding the factors that control regional geochemical variations (Reimann and Garrett, 2005). The development of

⁎ Corresponding author at: School of Biological, Earth and Environmental Sciences, University of New South Wales, NSW, 2052, Australia. Tel.: + 61 2 93858084. E-mail address: [email protected] (D.R. Cohen). 0048-9697/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2012.01.036

government policies, such as future EU directives and other initiatives relating to soil quality (Barth et al., 2009), requires detailed and extensive geochemical data on soil. The areal coverage of existing atlases ranges from regional to continental scales, with sampling densities as low as one site per 15,000 km 2 (Xie and Cheng, 1997). Even at low sampling densities, regolith geochemical surveys typically deliver coherent spatial patterns that can be linked to natural and some anthropogenic processes (Smith and Reimann, 2008). However, at low densities the underlying geology and regional climatic characteristics tend to dominate regional geochemical patterns with features relating to anthropogenic effects difficult to isolate. This is demonstrated in geochemical atlases of Europe, Australia, China and North America (Salminen et al., 2005; de Vos et al., 2006; Wang et al., 2007; Smith, 2009; Woodruff et al., 2009; de Caritat and Cooper, 2011). Regional mapping at higher densities is generally required to detect major sources of contamination, or specific mineral deposits as opposed to generally mineralised districts (Reimann et al., 1998; Cohen et al., 1999; Cornelius et al., 2008). Whereas some surveys prior to the 1990s were conducted at high densities, such as those of Fauth et al. (1985) in Germany and Thalmann et al. (1989) in Austria, the detection limits were significantly higher and the range of analytes far more restricted

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

251

Troodos Massif

Mesaoria Plain

0

10

Quaternary units Circum-Troodos Sedimentary Sequence Keryneia Terrane Mamonia Terrane Troodos Ophiolite Complex basaltic pillow lavas mafic intrusive units ultramafic units

20 km

Fig. 1. Main geological terranes of Cyprus.

than current surveys. High density soil surveys are increasingly being applied to urban environments to map the relationship between contaminants and land use (Johnson et al., 2005). The G_BASE programme in the UK is now sampling at 1 site per km 2, with expected applications of the data ranging from resource assessment to archaeology and forensics. In the case of I, the recent high density mapping of Northern Ireland (Smyth and Johnson, 2011) displays highly contiguous patterns, with elevated values generally restricted to coastal areas and organicrich soils. The soil geochemical atlas of Cyprus has been conducted at the high density of 1 site per km 2 (Cohen et al., 2011). This paper will highlight some of the features of the atlas, especially the effects of differences in parent geology and primary environmental setting, mineralisation occurrences and various anthropogenic effects on soil geochemistry, along with implications for determining environmental baseline soil characteristics of the island.

2. The Cyprus setting 2.1. Geology Cyprus is situated within a chain of ophiolites that represent the last vestiges of the Mesozoic Tethys oceanic crust preserved at the modern African–Eurasian plate boundary (Robinson and Malpas, 1990). In the initial stage of formation of Cyprus, back-arc spreading and magma upwelling generated a vertical sequence of ultramafic units, mafic cumulates and sheeted dykes, and pillow basalts (Morris et al., 1998). This formed the Troodos Ophiolite Complex (TOC) which was obducted in the Mid Miocene. The subsequent period of low tectonic activity was dominated by deposition of carbonaterich sediments, commencing with the deep marine Lefkara Formation and followed by the shallow marine Pakhna Formation and other units of the Circum-Troodos Sedimentary Succession (CTSS). A number of shallow depositional basins contain gypsiferous sub-units

Mixed skeletal residual and transported regolith

Residual regolith

Colluvium Terraced hillslopes Palaeo and modern gravels and alluvium Floodplain deposits

Troodos Ophiolite Complex

CTSS units Fig. 2. Simplified regolith-landscape model for Cyprus.

252

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

34 ° E

33 ° E

Keryneia Sampling site Area outside the effective control of the government of the Republic of Cyprus

Lefkosia

A’

Ammochostos

Polis

35 ° N Kokkinokhoria region

Larnaca

Pafos Lemesos

0

10

20 km

A Fig. 3. Soil sampling locations. Data for profile A–A′ is plotted in Fig. 11.

a Top soil 0

10

(0 – 25 cm)

20 km

ar-Ca

5,377 samples

(%) 30 23 20 18 15 12 8 5 3 2 1

Lefkosia

Larnaca

Pafos Lemesos

ar-Fe Lefkosia

b

(%)

Larnaka

Pafos Lemesos

Fig. 4. Geochemical maps of (a) ar-Ca and (b) ar-Fe in top soil.

7.0 6.0 5.2 4.5 4.0 3.5 3.0 2.5 2.0 1.4 0.8

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

253

a Top soil 0

10

(0 – 25 cm)

20 km

ar-Cr

5,377 samples

(mg/kg) 500 110 85 75 63 57 50 45 36 31 20

Lefkosia

Larnaca

Pafos Lemesos

tot-Cr

b

(mg/kg) 2000 600 450 340 220 180 160 140 120 90 50

Fig. 5. Geochemical maps of (a) ar-Cr and (b) tot-Cr in top soil.

reflecting the Messinian salinity crisis that occurred when the Mediterranean was closed and sea levels dropped (Rouchy et al., 2001). An abrupt uplift of the area occurred during the Pleistocene at ~ 2 Ma resulting in deposition of Pleistocene fanglomerates, mainly on the northern flanks of the Troodos Massif. Directly associated with the TOC ultramafics are podiform chromites and the large (now abandoned) Amiantos asbestos mine. Cyprus-type, Cu-rich massive sulphides are associated with the pillow basalts and there are twelve main deposits that have been mined in open cut operations. The Cyprus-type deposits have elevated Zn and Au in some sections but relatively low concentrations of Pb and large ion lithophile elements such as Ba (Constantinou and Govett, 1972; Sawkins, 1980). The geology of Cyprus can be divided into five main geological terranes — the TOC which has been subdivided into the ultramafic/mafic cumulate units and the overlying basaltic volcanic units; the Mamonia Terrane; the Keryneia Terrane; the Circum-Troodos Sedimentary Succession; and Quaternary units (Fig. 1). 2.2. Regolith and topography The regolith development in Cyprus reflects the young age of the terrane and rapid tectonic uplift, which has limited the development and preservation of deeply weathered profiles. A large volume of erosional products have been transported by gravitation and fluvial processes to form the Pliocene–Recent valley-fill sediments. In central Troodos the steep relief and historical deforestation have resulted in limited preservation of residual weathering profiles on ridge tops, pockets of colluvium or alluvium in small depressions

or stream beds, and either skeletal or no regolith cover elsewhere. In the gabbroic and ultramafic areas, weathering generally extends to depths of ~ 5 m with primary rock fabric partly preserved through feldspars and spinels. In steeper parts of the TOC terrane the nature of the regolith varies from extensive exposures of C/D-horizons to thick sequences of preserved or partly preserved colluvial mantles. The Troodos flanks, containing a variety of rock types that include TOC and Mamonia Terrane inliers and overlying CTSS units, are characterised by a series of radial ridges and streams extending towards the coastal fringe or Mesaoria Plain. Much of this region has been terraced. Colluvial pediment depths progressively increases away from central Troodos, with hills and ridges typically containing thin regolith cover surrounded by thicker colluvial accumulations in lower parts of the topography. The Mesaoria Plain contains sheetwash colluvium, deltaic deposits and river palaeogravels, containing mixtures of TOC and CTSS-derived clasts, and capped by recent alluvium. Along the coastal fringe the regolith is highly disturbed, especially where residential and industrial development is widespread. The TOC is characterised by brown to reddish calcaric leptosols and cambisols (Hadjiparaskevas, 2005). In some areas the soil profiles are skeletal and in others there is a well-preserved residual Bhorizon. In the forested areas a well-defined and organic-rich Ahorizon up to 10 cm thick is developed over a variably-thick Bhorizon characterised by Fe-oxide accumulation, and a slightly bleached underlying C-horizon. Dark red terra rossa soils cover carbonate-rich units in areas with low relief; elsewhere the CTSS carbonates break down to chalky soils with up to 25% clays. These areas have been extensively cultivated for centuries, especially the broad river valleys, due to the availability of groundwater. The Mesaoria

254

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

a Top soil 0

10

(0 – 25 cm)

20 km

tot-Cr ar-Cr

5,377 samples

8.0 6.0 5.0 4.5 4.1 3.6 3.2 2.8 2.5 2.0 1.2

Lefkosia

Larnaca

Pafos Lemesos

tot-Ca ar-Ca

b

4.0 3.2 2.8 2.5 2.2 2.0 1.8 1.6 1.4 1.2 1.0

Fig. 6. Ratios of tot- to ar-values for (a) Cr and (b) Ca in top soil.

Plain soils have high clay contents. Apart from residual soils on restricted parts of ridges on Troodos, the majority of regolith in Cyprus has undergone some degree of transportation. A preliminary regolith-landform model for Cyprus has been developed (Fig. 2) that is largely based on the residual–transported– depositional domain model outlined by Anand and Paine (2002) and the terrain model used in the design of the Geochemical Atlas of Europe (Salminen et al., 2005). Cyprus has experienced a long history of civilisation, commencing in the early Neolithic Period, due to its strategic position in the eastern Mediterranean. The Cu and timber resources have been exploited for millennia. In recent times there has been significant urban expansion along the southern coast with some associated industrial and tourist developments, and a substantial cement production industry. 3. Sampling and analysis Sites were pre-selected on a nominal 1 × 1 km grid (1.4 × 1.4 km over central Troodos) across much of Cyprus. This generated 5377 sites in the main survey, with a further 140 sites directed towards more detailed studies around some of the mines and main geological boundaries (Fig. 3). At each site a top soil (0–25 cm) and a sub soil (50–70 cm) sample was collected using an auger, and sieved to b2 mm in the field. Sampling depth followed the general protocols for the FOREGS Geochemical Atlas of Europe (Salminen et al., 1998), with a comparison between the two depths designed to provide information about enrichment or depletion processes between the

layers and potential anthropogenic contamination of the top layer. In areas with calcrete development or skeletal regolith, it was necessary to collect the sub soil sample from a shallower depth. Each field crew was led by a geologist or geochemist and a variety of field parameters were recorded, including local geology and land use. Samples were split in the laboratory using a riffle splitter. Splits of the unmilled fraction were used for 1:5 soil:water slurry pH and EC. Around a 30 g split was milled in mild steel. A 0.5 g sub-sample of the milled material was digested in aqua regia (1:2.5 HCl:HNO3) and subjected to multi-element analysis by ICP-MS and a further 2 g sub-sample analysed by INAA at Actlabs laboratories. The top soil samples were analysed for major oxides by XRF; total S, total C and soil organic carbon by automatic CS analysers; and soluble ions by chromatography at the GSD Laboratories in Lefkosia (Cohen et al., 2011). Detailed quality control procedures were implemented, including use of certified and project-specific reference materials representative of end-member lithologies, and site and sub-sample duplicates. These procedures were based on various protocols used in the exploration and environmental geochemistry sectors (ISO, 1995; Reimann et al., 1998; Reimann, 2005). Four blanks, six reference materials and six sub-sample duplicates were inserted into each sub-batch of 100 unknown samples. Analytical precision was better than ±15% (and typically better than ±10%) for all analytes in excess of 10 × the detection limit. The target zone for each analyte for each reference material was the certified or recommended values ± one standard deviation of the analyses after trimming values outside the 10th–90th percentile for the entire analytical run, with no more than 10% of

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

255

a

b

c

d

e

Fig. 7. Distribution of area with high ar-Ba, Bi, Mo and U concentrations in top soil and location of the Miocene-Messinian carbonate-dominated depositional basins.

observations permitted outside the target zone. All reference material values in excess of two standard deviation from the recommended value were followed up with the laboratory to check calibrations

and, if necessary, reanalysis of the sub-batch. Other project controls were emplaced, including “blind” resampling of a number of sites by different field crews.

pH 4

6

ar-Cu (mg/kg)

ar-Ca (%) 8

10

0.1

1

10

10

100

1k

10k

Ultramafic Mafic intrusive Basaltic volcanic Mafic clastic Silicic clastic Carbonate Alluvium-colluvium

ar-Fe (%) 1

ar-Cr (mg/kg) 10

10

100

tot-Cr (mg/kg) 1k

10

100

Ultramafic Mafic intrusive Basaltic volcanic Mafic clastic Silicic clastic Carbonate Alluvium-colluvium Fig. 8. Comparative boxplots of the distributions of selected elements in top soil divided by major lithological groups.

1k

10k

256

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

a Top soil 0

10

(0 – 25 cm)

20 km

ar-Cu

5,377 samples

(mg/kg)

Cu Mines

800 220 135 115 100 75 65 55 45 35 15

Lefkosia

Larnaca

Pafos Lemesos

pH

b

5.0 6.5 7.0 7.3 7.5 7.8 8.0 8.2 8.4 8.6 8.8

Fig. 9. Geochemical maps of (a) ar-Cu and (b) pH in the top soil. Major urban centres are also indicated.

Due to the high sampling density, colour-sliced gridded data was found to be more visually effective than individual sample symbol plots. Inverse-distance weighting (IDW) was used to generate the grids for the geochemical maps. In order to preserve the “point data” characteristics of the sample and limit propagation of anomalies from points with very high or low values (Johnson et al., 2005), while allowing some smoothing to generate a continuous coverage, the final IDW model selected had a grid-cell size of 0.33 km, a distance weighting exponent of 1.6 and a maximum search radius from the centre of each grid-cell of 2 km. An 11-division colour or shading scale, roughly corresponding to variable deciles, was used for maps in the Atlas and this paper. The samples were divided into the seven major lithological groups prior to generation of comparative boxplots. In referring to analytical methods, the aqua regia-extractable and ICPMS-analysed values for element are labelled “ar-” and the total values determined by INAA as “tot-”. Although the effects of compositional data on statistical analysis are recognised (Filzmoser et al., 2009), the effects on spatial patterns and element ratios were not considered sufficient to justify the use of adjustment procedures, such as the use of log-ratios, prior to generating the maps. 4. Results and discussion 4.1. Geological controls on element patterns The ar-Ca and ar-Fe data strongly reflect the effects of parent lithology on soil major element geochemistry. The ar-Ca values across the TOC are uniformly b4%, whereas the CTSS carbonate-rich units

display spatially contiguous patterns with Ca values 18–40% (Fig. 4a). The soil geochemical boundary between the CTSS and the TOC is very sharp, with Ca values dropping from >20% to b4% within the 1-km grid sampling distance. The variability in ar-Ca to the southwest of the TOC is a reflection of the complexity of the geology, with small outcrops of a variety of lithologies relating to the Troodos, Mamonia and CTSS terranes. Some elevated Ca values in areas mapped as basalts or mafic intrusive rocks relate to soils of mixed composition derived from both TOC and CTSS materials (e.g. colluvium with cobbles ranging in composition from calcarenite to weathered gabbro). The patterns for ar-Fe are a mirror image to that of ar-Ca, with a marked difference between the Fe contents of soils in the TOC and volcanic units of the Mamonia Terrane, and the surrounding CTSS units (Fig. 4b). The ultramafic core of the TOC and basaltic volcanic units contains ar-Fe contents >6% (and tot-Fe >7%). The gabbros surrounding the ultramafic core have low ar-Fe as the soils are immature leading to a significant proportion of the Fe being contained within magnetite and other resistate phases. The CTSS and other carbonate-dominated areas display very low soil Fe contents (b2%). As with Ca, there are very sharp soil geochemical boundaries (at the 1-km scale of sampling) between high and low soil Fe values along the southern boundary of the TOC. The Fe contents are slightly elevated along the coastal fringe, reflecting the accumulation of magnetite grains and mafic cobbles derived from the TOC. Soil ar-Cr and tot-Cr also display strong controls exerted by the combination of parent geology and subsequent sedimentological processes. Tot-Cr concentrations are an order of magnitude higher over the ultramafic lithologies (including the slivers of serpentinite on the western end of the island and east of Pafos) than any other

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

257

a 5,377 samples 0

10

20 km

Top soil (0 – 25 cm) Lefkosia

Larnaca

Sub soil (50 – 75 cm) Pafos Lemesos

ar-Pb (mg/kg)

b

60.0 12.7 11.5 10.6 9.6 8.7 8.1 7.3 6.2 3.9 1.0

Fig. 10. Geochemical maps of ar- Pb in (a) top soil and (b) sub soil.

lithology. Values are also elevated in low-lying areas along the southern coastline, the Kokkinokhoria region and the northern end of the Polis Valley (Fig. 5). A comparison of the ar-Cr and tot-Cr data indicates significant variations in signature between the areas of elevated soil Cr content. There is generally a converse relationship between soil tot-Cr and the tot-Cr:ar-Cr ratio, suggesting that in areas with high tot-Cr values most of the Cr is bound up in resistate chromite or Cr-substituted magnetite (Fig. 6a). The ratios are very low along the entire southern coast, and especially in the fanglomeratedominated Kokkinokhoria area, indicating that the majority of Cr in soils of those areas is present in either clasts of transported ultramafics or as chromite in the heavy mineral-rich sands observed along many of the beaches. By contrast, the tot-Ca:ar-Ca is close to 1 across the CTSS and Quaternary units, but rises to >2 over much of the central TOC where a portion of the Ca is bound up in partiallyweathered feldspars (resistate to aqua regia attack) rather than Casubstituted clays and carbonates (Fig. 6b). Soil geochemical patterns may be linked to the geochemical and depositional evolution of the marine basins and CTSS for some of the elements that display strong redox controls on chemical mobility. The distribution of strongly and marginally elevated values for ar-Ba, Bi, Mo and U (relative to the whole dataset) and comparison of the quartile values for these elements (where ar-Ca >10%), ar-Cr and ar-Cu are plotted in Fig. 7. Bismuth is elevated around the flanks of the Polemi Basin and Polis Graben, with contiguous spatial patterns cutting across a number of lithologies. Elevated ar-Ba (and tot-Ba) is largely restricted to the deep-marine Lefkara Formation carbonates that lap onto the pillow basalts of the TOC, with much lower values in the overlying shallower Pakhna Formation that contains some

reefal limestone facies. By contrast, Mo and U are elevated within the Miocene-Messinian Basins that are dominated by the Pakhna Formation. The high ar-U:tot-U ratios indicate that most U is not substituted into monazite or other resistate minerals, but into the carbonates. The increased Ba in the Lefkara Formation carbonates deposited in the reduced deeper waters but increased Mo and U in the Pakhna Formation deposited under the more oxidising conditions expected in a shallow basin, suggest element mobility in this case is more important in determining carbonate trace element contents for Bi, U and Mo (and possibly Ba) than the precipitation mechanisms outlined in Tribovillard et al. (2011). Comparative boxplots between the main lithologies for selected variables also emphasise the strong lithological controls on the distribution of elements in the soils (Fig. 8). 4.2. Effects of mining and urbanisation Apart from low values in the vicinity of the ultramafics and some mafic intrusive units, median soil ar-Cu concentrations do not vary greatly between the main lithology groups (although the differences are statistically significant). Values are highest in the pillow basalts (>200 mg/kg), especially near sulphide mineral deposits that generate most of the outliers in the boxplots (Figs. 8 and 9a). All the Cu mines display elevated soil ar-Cu values up to 2 km away from exposed mineralisation and the (former) mining operations but little indication of contamination or naturally elevated soil ar-Cu values beyond this distance. One feature of the ar-Cu is the 10-km wide corridor of elevated values that runs north–south through the sheeted dykes that display sporadic zones of hydrothermal alteration on the

258

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

ar-Cu (mg/kg)

ar-Ca (%) 40

A′

A

10,000

30

1,000

20

100

10

10

0

1

tot-Cr (mg/kg) 2,000

A′

A

ar-Pb (mg/kg)

A′

A

1,000

1,500

100

1,000

10

500

1

0

0.1

A′

A

Lemesos

Kalavasos

Kokkinopezoula

Lefkosia

Lemesos

Kalavasos

Kokkinopezoula

Lefkosia

area

Cu mines

Cu mines

area

area

Cu mines

Cu mines

area

Carbonates and other CTSS sedimentary units

Basaltic volcanic units (TOC)

Intrusive mafic units (TOC)

Fig. 11. Effect of variation in underlying lithology and the location of major urban centres and Cu mines on the ar-Ca, ar-Cu, ar-Pb and tot-Cr in top soil on traverse A–A′ (see Fig. 3 for traverse location).

western side of the TOC. Soil ar-Cu values are typically elevated in the main urban areas relative to the surrounding areas. The relative Cu distribution between the Lefkara and Pakhna formations follows Bi whereas Cr follows U and Mo, hence there may be a component of mechanical transportation from Troodos in the patterns. The majority of soils in Cyprus display near-neutral to alkaline pH values (Fig. 9b). This reflects the influence of carbonates in the CTSS

and colluvium–alluvium areas, and alkaline earth oxides and hydroxides derived from the mafic and Ca-rich lithologies in the TOC and Mamonia Terranes. The carbonate-dominated formations, such as the Pakhna and Lefkara Formations, display soil pH >8.3. Values below pH 6.4 are largely restricted to the vicinity of the sulphide mines, though sporadic low pH values throughout the sheeted dykes in the western end of Troodos reflect isolated zones of sulphide

Mine Urban All other sites areas samples

Ratio of top soil to sub soil metal concentrations

2.0

Q3

2.0

med Q1

1.5

1.5

1.0

1.0

0.5

0.5

tot-Br ar-Ca tot-Ca ar-Fe tot-Fe ar-La tot-La ar-Cr tot-Cr ar-Cu ar-As ar-Zn ar-Hg ar-Pb Fig. 12. Comparison of the distributions for the ratio of top soil to sub soil concentrations of selected elements. Samples are separated into the Cu mine sites, urban areas and remaining samples.

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

259

1000

Cr

INAA or XRF

Sr

Top soil Ni

Sub soil

100

Ba Zn

Co Ca

Sc

10

As Nd

Fe Sm Sb 1

Yb Mo

Na

Rb La Ce

U

Th Cs

Hf

Eu Tb Ta Lu

0.1 0.1

1

10

100

1000

Atlas of Europe (~850 sites) Fig. 13. Comparison between the average tot- metal contents for top soil and sub soil in Cyprus and in the Geochemical Atlas of Europe. Values in mg/kg except Ca, Fe and Na in %.

mineralisation — a feature indicated in various locations by secondary Cu minerals such as malachite. Soil ar-Pb is elevated near some sulphide deposits and in most of the Kokkinokhoria region (Fig. 10). Other elevated values appear more closely related to the main urban areas of Lefkosia (Nicosia), Larnaca, Lemesos (Limassol) and Pafos, where top soil ar-Pb values are generally higher and form more contiguous anomalous areas than the sub soil. Sporadic anomalous Pb values with no obvious geological controls probably relate to hunting activities (heavy mineral concentrates from a number of these samples containing shotgun pellets). To further highlight the strong spatial zonation in most of the geochemical variables, the top soil ar-values for Ca, Cr, Cu and Pb are presented from a 2–3 km wide traverse that extends from the Akrotiri Peninsula, through the Lemesos urban/industrial area, the Kalavasos and Kokkinopesula Cu mining areas, and finally to Lefkosia (Nicosia) (Figs. 3 and 11). There are six major lithological boundaries crossed in this representative traverse. For major and minor elements (Ca and Cr) there are sharp geochemical steps between the different lithologies. Whereas the Cr contents are expectedly high in the area with the mafic units, there is a gradual increase in Cr in the soils overlying the carbonates to the south across the Lemesos region and north towards Lefkosia that probably reflects heavy mineral accumulations in the flat terrains at the ends of the traverse. Both Cu and Pb display erratic but generally high values over the two mining areas, but Cu values are also elevated in the mafic units generally. Many of the Pb values are elevated in the urban areas but the lack of continuity might indicate a “nugget effect” related to the presence of particulate Pb derived from vehicle batteries, wheel balancing lugs, Pb-based paints and other anthropogenic sources.

4.3. Top soil versus subsoil There is generally a high correlation between the geochemistry of the top soil and sub soils in Cyprus, with median top:sub soil ratio generally in the range 0.8–1.2 (Fig. 12). Soil tot-Br contents display a distinct tendency towards sub soil enrichment, which may be a function of the mobility of Br and flushing from the top soil to the lower soil horizons. There is weak depletion of Ca in top soil related to both development of low-Ca terra rossa soils and thin transported soils containing a component of TOC materials overlying residual

profiles developed in CTSS carbonate units. REE and Cr are slightly enriched in the top soil relative to sub soil, but this may represent heavy mineral lag effects, whereas ar-Pb is strongly enriched (a factor of ~1.5) in top soil in all areas, especially urban areas. ar-Cu displays top soil enrichment in the mines and urban areas relative to the remainder of the dataset and arAs in the mines. Conversely, ar-Zn and ar-Hg are both significantly enriched in top soil in urban areas Compared with the FOREGS Atlas of Europe, the total soil values are higher in Cyprus for most of the mafic and ultramafic-associated elements (Cr, Co, Cu, V, Fe) and sedimentary carbonate-hosted elements (Ca, Sr and Ba), but lower for most other elements including the REE and incompatible elements such as Hf, Th and Ta (Fig. 13). This simply reflects the balance in surface area exposure between the major lithologies in Cyprus (ophiolites and marine sediments) and Europe as a whole (granites, volcanic units, a wide variety of terrestrial and marine clastic sedimentary rocks and various metamorphic units). The full set of element and other variable maps are presented in Cohen et al. (2011) and key parameters are summarized in Table 1.

5. Conclusion The data demonstrates the effects of lowering of multielement detection limits and detailed quality control procedures on the capacity of modern surveys to map subtle spatial variations in regional soil geochemistry. The results from this highdensity, multi-depth and multi-element geochemical mapping project clearly indicate that major and trace element concentrations in the soils of Cyprus are dominated by parent lithology. Subsequent regolith processes, such as physical concentrating of heavy minerals and ocean influences along the coastal plains, also influence spatial geochemical patterns. Different combinations of elements serve to map the geology with even subtle differences between rock types being mappable. An example is the Lefkara Formation versus the overlying Pakhna Formation carbonates using elements such as Ba and U. In urban and industrial areas there is detectable contamination for a number of metals (e.g. Pb), but mainly in the top soil. Mining activities have contributed to elevated concentrations of commodity elements contained within the deposit and elements associated with the mining activities, although elevated values in soils generally

260

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

Table 1 Basic statistical summary for Cyprus soil geochemical data. Variable

Ag Al As As Au B Ba Ba Be Bi Br Ca Ca Cd Ce Ce Co Co Cr Cr Cs Cs Cu Dy Er Eu Eu Fe Fe Ga Gd Ge Hf Hf Hg Ho In K La La Li Lu Lu Mg Mn Mo Mo Na Na Nb Nd Nd Ni Ni P Pb pH Pr Rb Rb Re Sb Sb Sc Sc Se Sm Sm Sn Sr Ta Tb

Method

ICP ICP ICP INAA INAA ICP ICP INAA ICP ICP INAA ICP INAA ICP ICP INAA ICP INAA ICP INAA ICP INAA ICP ICP ICP ICP INAA ICP INAA ICP ICP ICP ICP INAA ICP ICP ICP ICP ICP INAA ICP ICP INAA ICP ICP ICP INAA ICP INAA ICP ICP INAA ICP INAA ICP ICP electrode ICP ICP INAA ICP ICP INAA ICP INAA ICP ICP INAA ICP ICP INAA ICP

Units

DL

mg/kg % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg % mg/kg mg/kg mg/kg mg/kg mg/kg % mg/kg mg/kg mg/kg % % mg/kg mg/kg mg/kg mg/kg mg/kg % mg/kg

0.002 0.01 0.1 0.5 0.002 1 0.5 50 0.1 0.02 0.5 0.01 1 0.1 0.1 3 0.1 1 0.5 5 0.02 1.0 0.1 0.1 0.1 0.1 0.2 0.01 0.01 0.02 0.1 0.1 0.1 1.0 0.01 0.1 0.02 0.01 0.5 0.5 0.1 0.05 0.1 0.01 10 0.01 1 0.001 0.01 0.1 0.02 5 0.1 20 0.01 0.1

mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

0.1 0.1 15 0.001 0.02 0.1 0.1 0.1 0.1 0.1 0.1 0.2 1 0.5 0.1

Top soil (0–25 cm)

Sub soil (25–50 cm)

(5377 samples)

(5197 samples; 209 for pH)

Min

Max

Mean

Std dev

Min

Max

Mean

Std dev

b0.002 0.12 b0.1 b0.5 b0.002 b1 b0.5 b 50 b0.1 b0.02 b0.5 0.1 b1 b0.1 0.3 b3 1.4 b1 b0.5 b5 b0.02 b1.0 1.9 b0.1 b0.1 b0.1 b0.2 0.15 0.4 0.24 b0.1 b0.1 b0.1 b1.0 b0.01 b0.1 b0.02 b0.01 b0.5 b0.5 b0.1 b0.05 b0.1 0.06 28 b0.01 b1 b0.001 b0.01 b0.1 0.18 b5 b0.1 b 20 b0.01 b0.1 1.85 b0.1 0.2 b 15 b0.001 b0.02 b0.1 b0.1 1.2 b0.1 b0.1 b0.1 b0.2 2 b0.5 b0.1

5.02 8.61 230.0 224.0 1.080 38 2490 5150 10.2 5.96 282.0 41.2 41 15.5 353.0 393 311.0 331 2270.0 > 100,000 7.22 11.0 > 12,000 11.9 7.3 3.3 3.6 31.20 32.4 16.30 19.4 1.6 1.1 12.0 4.52 2.5 23.10 1.33 191.0 205.9 117.0 1.60 1.5 13.40 12,000 58.40 66.00 7.03 6.59 3.5 99.20 97 3990 4900 0.31 4490 10.78 32.9 51.3 131 0.116 97.10 91.6 853.0 74.1 70.0 16.6 18.3 96.3 1200 7.5 2.1

0.04 1.88 4.9 6.9 0.006 2 163 227 0.4 0.09 10.9 11.9 13 0.3 14.2 25 21.2 27 73.7 518 0.42 1.3 87.9 2.0 1.1 0.5 0.7 3.57 4.3 5.31 2.5 0.1 0.1 1.8 0.03 0.4 0.03 0.24 8.3 12.4 9.3 0.14 0.3 1.32 981 0.95 1.53 0.25 0.70 0.3 7.16 10 111 102 0.04 11 8.32 1.8 9.5 21 0.002 0.41 0.6 11.2 18.0 0.7 1.8 2.5 0.8 335 0.4 0.3

0.16 1.14 7.0 7.7 0.035 2 265 340 0.3 0.14 13.7 10.4 10 0.5 11.8 19 22.5 26 129.9 2809 0.37 1.2 332.7 0.9 0.6 0.3 0.4 2.60 2.6 2.90 1.3 0.1 0.1 1.5 0.08 0.2 0.31 0.16 6.8 9.6 7.7 0.09 0.2 1.63 857 1.91 3.10 0.37 0.68 0.3 4.89 9 374 429 0.03 77 0.80 1.3 6.5 22 0.004 2.11 2.1 14.3 11.6 1.6 1.1 1.3 2.5 370 0.5 0.2

b 0.002 0.06 b 0.1 b 0.5 b 0.002 b1 b 0.5 b50 b 0.1 b 0.02 b 0.5 0.2 b1 b 0.1 0.1 b3 0.3 b1 b 0.5 b5 b 0.02 b 1.0 0.6 b 0.1 b 0.1 b 0.1 b 0.2 0.05 0.1 0.01 b 0.1 b 0.1 b 0.1 b 1.0 b 0.01 b 0.1 b 0.02 b 0.01 b 0.5 b 0.5 b 0.1 b 0.05 b 0.1 0.02 15 b 0.01 b1 b 0.001 b 0.01 b 0.1 0.03 b5 b 0.1 b20 b 0.01 b 0.1 3.11 b 0.1 0.1 b15 b 0.001 b 0.02 b 0.1 b 0.1 0.3 b 0.1 b 0.1 b 0.1 b 0.2 1 b 0.5 b 0.1

8.23 8.71 259.0 267.0 1.670 19 3010 4770 16.6 9.40 175.0 48.4 42 21.3 101.0 162 1630.0 1570 1810.0 > 100,000 6.74 12.0 > 12,000 14.8 7.8 4.8 5.3 23.80 22.0 23.40 21.6 6.0 1.4 11.0 3.62 2.7 0.79 1.37 115.0 112.0 119.0 4.10 1.7 12.90 12,000 78.50 99.00 4.22 3.71 2.4 87.40 105 6900 5660 0.38 2040 10.50 20.2 57.6 152 0.299 51.80 27.9 450.0 69.8 98.6 17.7 20.8 52.8 1200 6.6 2.7

0.04 1.94 4.8 6.7 0.006 2 179 250 0.4 0.08 12.8 13.0 14 0.3 13.0 23 21.9 28 70.5 456 0.44 1.3 91.4 2.0 1.1 0.5 0.7 3.54 4.2 5.38 2.5 0.1 0.1 1.7 0.03 0.4 0.02 0.22 7.7 11.7 9.8 0.14 0.3 1.31 951 0.93 1.47 0.24 0.66 0.2 6.84 10 110 100 0.04 7 8.33 1.7 9.1 20 0.002 0.32 0.5 11.6 18.0 0.6 1.8 2.4 0.6 344 0.4 0.3

0.20 1.25 9.5 10.1 0.043 2 330 431 0.4 0.17 11.3 11.0 11 0.7 11.0 19 35.6 37 128.3 2768 0.39 1.2 400.8 1.0 0.6 0.3 0.4 2.62 2.7 3.13 1.4 0.1 0.1 1.5 0.08 0.2 0.03 0.16 6.7 9.7 8.6 0.11 0.2 1.60 944 2.57 3.63 0.28 0.66 0.2 5.04 10 398 447 0.03 30 1.12 1.3 6.8 22 0.005 1.08 0.9 11.0 12.2 1.9 1.1 1.4 1.5 373 0.5 0.2

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

261

Table 1 (continued) Variable

Method

Units

DL

Top soil (0–25 cm)

Sub soil (25–50 cm)

(5377 samples)

(5197 samples; 209 for pH)

Min Te Th Th Ti Tl Tm U U V Y Yb Yb Zn Zr

ICP ICP INAA ICP ICP ICP ICP INAA ICP ICP ICP INAA ICP ICP

mg/kg mg/kg mg/kg % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

0.02 0.1 0.2 0.01 0.02 0.1 0.1 0.5 1 0.1 0.1 0.2 0.1 0.1

b0.02 b0.1 b0.2 b0.01 b0.02 b0.1 b0.1 b0.5 b1 0.2 b0.1 b0.2 b0.1 b0.1

Max

Mean

5.34 47.7 19.0 1.18 4.53 1.7 129.0 11.2 485 79.2 10.0 9.7 4990 44.9

0.06 1.3 2.7 0.09 0.09 0.2 0.7 1.1 96 11.4 1.0 1.9 67 4.6

Std dev 0.18 1.4 2.5 0.08 0.15 0.1 1.9 1.2 70 5.3 0.5 0.9 118 4.7

Min b 0.02 b 0.1 b 0.2 b 0.01 b 0.02 b 0.1 b 0.1 b 0.5 b1 0.2 b 0.1 b 0.2 b 0.1 b 0.1

Max 13.50 10.7 17.3 1.08 19.90 1.6 19.6 18.5 618 80.6 9.6 10.4 7650 57.8

Mean 0.06 1.3 2.6 0.09 0.09 0.2 0.6 1.1 96 11.5 1.0 1.9 62 5.1

Std dev 0.26 1.4 2.6 0.08 0.33 0.1 0.9 1.3 74 5.9 0.6 1.0 175 5.3

ICP refers to aqua regia digestion ICP-MS; INAA is instrumental neutron activation analysis.

extend for less than 2 km beyond the mine operations. There is no single set of “background” element levels that can be applied to the soils of Cyprus, and which can be used to differentiate “natural” from “potentially polluted” sites. Median concentrations of many elements (e.g. Cr) vary by orders of magnitude between soils derived from the different lithologies. The atlas data will, however, guide the setting of lithology-specific and terrane-specific indicatory levels for natural element abundances.

Acknowledgements The authors acknowledge the funding, technical and logistical support provided by the Geological Survey Department of the Ministry of Agriculture, Natural Resources and Environment of the Republic of Cyprus, and assistance with the extensive analytical programme by Actlabs.

References Anand RR, Paine M. Regolith geology of the Yilgarn Craton, Western Australia: implications for exploration. Aust J Earth Sci 2002;49:3-162. Barth JAC, Grathwohl P, Fowler HJ, Bellin A, Gerzabek MH, Lair GJ, Barcelo D, et al. Mobility, turnover and storage of pollutants in soils, sediments and waters: achievements and results of the EU project AquaTerra. A review. Agron Sustain Dev 2009;29:161–73. de Caritat P, Cooper M. National Geochemical Survey of Australia: The Geochemical Atlas of Australia. Geoscience Australia Record 2011/20; 2011. Cohen DR, Rutherford NF, Morisseau E, Zissimos AM. Geochemical Atlas of Cyprus. Sydney: UNSW Press; 2011. Cohen DR, Silva-Santisteban CM, Rutherford NF, Garnett DL, Waldron HM. Comparison of biogeochemical and stream sediment geochemical patterns in the north eastern region of NSW. J Geochem Explor 1999;66:469–89. Constantinou G, Govett GJS. Genesis of sulfide deposits, ochre and umber of Cyprus. Inst Mining Metall 1972;81:B34–46. Cornelius M, Robertson IDM, Cornelius AJ, Morris PA. Geochemical mapping of the deeply weathered western Yilgarn Craton of Western Australia, using laterite geochemistry. Geochem Explor Enviro Anal 2008;8:241–54. Darnley AG, Björklund A, Bølviken B, Gustavsson N, Koval PV, Plant JA, et al. A global geochemical database for environmental and resource management: recommendations for international geochemical mapping. IGCP Proj 259. UNESCO Publ; 1995. de Vos W, Tarvainen T, et al. Geochemical Atlas of Europe. Part 2 — Interpretation of geochemical maps, additional tables, figures, maps, and related publications. Geol Surv Finland, Espoo; 2006. Fauth H, Hindel R, Siewers U, Zinner J. Geochemischer Atlas Bundesrepublik Deutschland. Verteilung von Schwermetallen in Wässern und Bachsedimenten. Bundesanstalt für Geowisseneschaften und Rohstoffe, Hannover. Stuttgart: Schweizerbart'sche Verlagsbuchandlung; 1985. Filzmoser P, Hron K, Reimann C. Univariate statistical analysis of environmental (compositional) data — problems and possibilities. Sci Total Environ 2009;407:6100–8.

Garrett RG, Reimann C, Smith DB, Xie X. From geochemical prospecting to international geochemical mapping: an historical overview. Geochem Explor Enviro Anal 2008;8:205–17. Hadjiparaskevas C. European Soil Bureau — Research Report No. 9, Soil Survey and Monitoring in Cyprus; 2005. ISO. Soil quality: extraction of trace elements soluble in aqua regia. ISO-11466 Standard; 1995. Johnson CC, Breward N, Ander EL, Ault L. G-BASE: baseline geochemical mapping of Great Britain and Northern Ireland. Geochem Explor Enviro Anal 2005;5: 347–57. Morris A, Anderson MA, Robertson AHF. Multiple tectonic rotations and transform tectonism in an intraoceanic suture zone, SW Cyprus. Tectonophysics 1998;299: 229–53. Plant JA, Reeder S, Salminen R, Smith DB, Tarvainen T, De Vivo B, Petterson MG. The distribution of uranium over Europe: geological and environmental significance. Appl Earth Sci 2003;112:221–38. Plant JA, Smith D, Smith B, Williams L. Environmental geochemistry at the global scale. J Geol Soc Lond 2000;157:837–49. Reimann C, Äyräs M, Chekushin V, Bogatyrev I, Boyd R, de Caritat P. Environmental geochemical atlas of the central Barents region. NGU–GTK–CKE Special Publication. Geological Survey of Norway, Trondheim, Norway; 1998. Reimann C, Demetriades A, Eggen OA, Filzmoser P. The EuroGeoSurveys geochemical mapping of agricultural and grazing land soils (GEMAS). NGU Rpt 2011;043: 2011. Reimann C, Garrett RG. Geochemical background: concept and reality. Sci Tot Environ 2005;350:12–27. Reimann C. Sub-continental scale geochemical mapping: sampling, quality control and data analysis issues. Geochem Explor Enviro Anal 2005;5:311–23. Robinson PT, Malpas J. The Troodos ophiolite of Cyprus: new perspectives on its origin and emplacement. In: Malpas J, editor. Ophiolites–oceanic crust analoguesProceedings of the Troodos 1987 Symposium, Nicosia, Cyprus Geological Survey; 1990. p. 13–26. Rouchy JM, Orszag-Sperber F, Blanc-Valleron MM, Pierre C, Rivière M, CombourieuNebout N, Panayides I. Paleoenvironmental changes at the Messinian–Pliocene boundary in the eastern Mediterranean (southern Cyprus basins): significance of the Messinian Lago-Mare. Sed Geol 2001;145:93-117. Salminen R, Batista MJ, Bidovec M, Demetriades A, De Vivo B, De Vos W, et al. FOREGS Geochemical Atlas of Europe. Part 1: background information methodology and maps. Geological Survey of Finland, Espoo; 2005. Salminen R, Tarvainen T, Demetriades A, Duris M, Fordyce FM, Gregorauskiene V, et al. FOREGS Geochemical Mapping Field Manual. GTK Guide 47. Geological Survey of Finland, Espoo; 1998. Sawkins FJ. Integrated tectonic-genetic model for volcanic-hosted massive sulfide deposits. Geology 1980;18:1061–4. Smith DB, Reimann C. Low-density geochemical mapping and the robustness of geochemical patterns. Geochem Explor Enviro Anal 2008;8:219–27. Smith DB. Geochemical studies of North American soils: results from the pilot study phase of the North American Soil Geochemical Landscapes Project. Appl Geochem 2009;24:1355–6. Smyth D, Johnson CC. Distribution of iodine in soils of Northern Ireland. Geochemistry: Exploration, Environment, Analysis 2011;11:25–39. Tan J. The atlas of endemic diseases and their environments in the People's Republic of China. Beijing: Science Press; 1989. Thalmann F, Schermann O, Schroll E, Hausberger G. Geochemischer Atlas der Republik Österreich 1:1,000,000. Geologische Bundesanstalt Wien. 1989. Tribovillard N, Algeo TJ, Baudin F, Riboulleau A. Analysis of marine environmental conditions based on molybdenum–uranium covariation — Applications to Mesozoic paleoceanography. Chem Geol 2011;268:211–25.

262

D.R. Cohen et al. / Science of the Total Environment 420 (2012) 250–262

Wang X, Chi Q, Liu H, Nie L, Zhang B. Wide-spaced sampling for delineation of geochemical provinces in desert terrains, northwestern China. Geochem Explor Enviro Anal 2007;7:153–61. Woodruff LG, Cannon WF, Eberl DD, Smith DB, Kilburn JE, Horton JD, Garrett RG, Klassen RA. Continental-scale patterns in soil geochemistry and mineralogy: results from

two transects across the United States and Canada. Appl Geochem 2009;24: 1369–81. Xie X, Cheng H. The suitability of floodplain sediment as a global sampling medium: evidence from China. J Geochem Explor 1997;58:51–62.