Science of the Total Environment 409 (2011) 4344–4350

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r.com/ locate /sc i totenv

The origin of lead in the organic horizon of tundra soils: Atmospheric deposition,plant translocation from the mineral soil or soil mineral mixing?

Jonatan Klaminder a,⁎, John G. Farmer b, Angus B. MacKenzie c

a Department of Ecology and Environmental Science, Umeå University, 90187 Umeå, Swedenb School of GeoSciences, University of Edinburgh, Edinburgh, EH9 3JN, Scotland, UKc Scottish Universities Environmental Research Centre, East Kilbride, G75 0QF, Scotland, UK

⁎ Corresponding author. Tel.: +46 90 7869554; fax:E-mail address: jonatan.klaminder@emg.umu.se (J. K

0048-9697/$ – see front matter © 2011 Elsevier B.V. Adoi:10.1016/j.scitotenv.2011.07.005

a b s t r a c t

a r t i c l e i n f o

Article history:Received 1 March 2011Received in revised form 27 June 2011Accepted 1 July 2011Available online 4 August 2011

Keywords:Pb isotopesTundra soilContaminationCryoturbation

Knowledge of the anthropogenic contribution to lead (Pb) concentrations in surface soils in high latitudeecosystems is central to our understanding of the extent of atmospheric Pb contamination. In this study, wereconstructed fallout of Pb at a remote sub-arctic region by using two ombrotrophic peat cores and assessedthe extent to which this airborne Pb is able to explain the isotopic composition (206Pb/207Pb ratio) in the O-horizon of tundra soils. In the peat cores, long-range atmospheric fallout appeared to be the main source of Pbas indicated by temporal trends that followed the known European pollution history, i.e. accelerated fallout atthe onset of industrialization and peak fallout around the 1960s–70s. The Pb isotopic composition of the O-horizon of podzolic tundra soil (206Pb/207Pb=1.170±0.002; mean±SD) overlapped with that of the peat(206Pb/207Pb=1.16±0.01) representing a proxy for atmospheric aerosols, but was clearly different from thatof the parent soil material (206Pb/207Pb=1.22–1.30). This finding indicated that long-range fallout ofatmospheric Pb is the main driver of Pb accumulation in podzolic tundra soil. In O-horizons of tundra soilweakly affected by cryoturbation (cryosols) however, the input of Pb from the underlying mineral soilincreased as indicated by 206Pb/207Pb ratios of up to 1.20, a value closer to that of local soil minerals.Nevertheless, atmospheric Pb appeared to be the dominant source in this soil compartment. We conclude thatPb concentrations in the O-horizon of studied tundra soils – despite being much lower than in boreal soils andrepresentative for one of the least exposed sites to atmospheric Pb contaminants in Europe – are mainlycontrolled by atmospheric inputs from distant anthropogenic sources.

+46 90 7866705.laminder).

ll rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

A common trend found by regional surveys in northern Europe isdecreasing lead (Pb) concentrations in forest mosses and surface soilwith increasing latitude (Steinnes, 1987; Alriksson, 2001; Rühling andTyler, 2001). The lower Pb concentrations in the north may generate amisconception that soils at high latitude are representative ofuncontaminated, pristine ecosystems and thus representative ofnatural conditions. Significant anthropogenic contamination, howev-er, has also occurred at remote, high latitude regions of the globethrough long-range atmospheric transport (Murozumi et al., 1969;Hong et al., 1994). Strong evidence for significant anthropogenicinputs of Pb to high latitude environments comes from long-termaerosol monitoring (Gong and Barrie, 2005) and from reconstructedtemporal deposition trends using natural archives, such as ice cores(Hong et al., 1994; Rosman et al., 1994; Zheng et al., 2007), lakesediment cores (Brännvall et al., 1999; Bindler et al., 2001), peat cores

(Steinnes, 1997) and snow (Boutron et al., 1994; Shotyk et al., 2005).All these studies illustrate that the atmospheric input of Pb hasfollowed the anthropogenic emissions history. Even though long-range transported atmospheric contaminants are undisputed sourcesof Pb in high latitude environments (Steinnes and Friedland, 2006),their importance for Pb levels in surface tundra soils remains moreuncertain.

The organic layer of surface soils has the potential to differ fromnatural archives, such as ombrotrophic peat and ice cores,which receiveinputs solely from atmospheric sources, because additional inputs canarise from the underlying mineral soil through plant translocationprocesses and from in-mixing ofminerals resulting frombioturbation orfrom cryoturbation (Klaminder and Yoo, 2008). Enrichment of Pb andother metal contaminants in organic-rich surface soils in comparisonwith the underlyingmineral matrix is a common pattern in soil surveys(Alriksson, 2001; Reimann et al., 2008). For the Scandinavian borealforest, plant uptake rates of Pb from the mineral soil in mature foreststands are in the range of 0.002–0.04 mgm−2 yr−1 (Klaminder et al.,2008a; Hovmand et al., 2009). These values are more than an order ofmagnitude lower than atmospheric inputs, which have varied between0.5 and 20 mg m−2 yr−1 over the last four decades (Rühling and Tyler,

4345J. Klaminder et al. / Science of the Total Environment 409 (2011) 4344–4350

2001). Therefore, plant pumping appears unlikely to explain Pbconcentrations in the O-horizon of boreal forest soils and surfaceenrichments of this toxic metal are mainly induced by atmosphericdeposition.

Less is known about the relative importance of distance-derivedairborne contaminants as a contributor to Pb in tundra surface soils,which are less exposed to atmospheric fallout than boreal forest soilsbecause of their more remote location from heavily industrializedareas of Europe. Tundra soils, developing in the harsh arctic climate,may receive significant inputs of metals from the underlying mineralsoil through cryogenic processes driven by repeated freeze–thawcycles in addition to inputs from plant uptake. Clearly, there is arationale for expecting that Pb levels in remote tundra soils might bedriven by natural processes to a larger extent than in boreal forests. Towhat extent this may be the case, however, is not known.

In this study we investigated the extent to which atmosphericinputs alone, or in combination with additional inputs from plantuptake and cryotubation, are likely to explain Pb concentrations in theO-horizon of tundra soil profiles collected in northern Sweden. Wereconstructed atmospheric inputs of Pb to this region using 210Pb-dated peat cores to assess historical variations in atmosphericdeposition rates and temporal variations in the 206Pb/207Pb ratio ofatmospheric Pb. We then compared the 206Pb/207Pb ratio of the O-horizon with the record of atmospheric Pb input preserved in the peatand with that of the underlying mineral soil to evaluate inputs fromplant pumping and cryoturbation in addition to atmosphericdeposition. We hypothesized that atmospheric-derived Pb constitutesa significant source of Pb to the O-horizon of tundra soils.

2. Materials and methods

2.1. Site description and sampling

Two peat cores were collected from the Stordalen mire (68°20′90′N, 18°58′57 ′E) in northern Sweden (Fig. 1a) using a Wardenaar peatcorer (Wardenaar, 1987). Permafrost heave drives the topography ofthe mire where higher uplifted palsas (peat mounds containing

Fig. 1. a) Location of the study area in Scandinavian perspective and b) lo

perennial frozen ice lenses) are characterized by ombrotrophic peathummocks consisting of plant communities dominated mainly eitherby dwarf shrub (Empetrum hemafroditum, Betula nana) and lichen(Cladonia ssp) or by Sphagnum fuscum and Dicranum elongatum(Malmer et al., 2005). A peat core was taken in each of these twoombrotrophic vegetation types, about 20 m apart, at the north-eastern part of the bog (Fig. 1b).

Soil pits (n=4) were excavated from well drained sites withpodzolic soils from the Abisko region (Fig. 1b). The soil pits covered agradient in mean annual precipitation ranging from 300 mm (east) toabout 1000 mm (west). In addition, two profiles from soils affected bycryoturbation (cryosols) were sampled from the area at 800 and1100 m a.s.l. (Fig. 1b). One of the profiles was sampled at the outerdomain of an active frost boil (non-sorted circle), a common type ofpatterned ground in the area. The other soil profile had a bend anddiscontinuous E-horizon indicative of past differential soil motion dueto frost heave. Despite themorphological evidence of cryoturbation allcryosol profiles had intact vegetation cover, consisting mainly ofEmpetrum sp. and lichen communities (tundra heath) and a 5–10 cmthick O-horizon. The organic soil horizon was sampled by repeatedcoring at each pit, using a plastic auger with saw-teeth (diameter8.1 cm), while deepermineral soil was sampled from the soil pit walls.Organic horizon samples were cut in the field.

2.2. Analytical methods

Peat and soil samples were analyzed for Pb concentrations andisotope ratios (206Pb/207Pb, 208Pb/206Pb and 208Pb/207Pb) usingslightly different methods.

Peat samples (~0.25 g, dried at 105 °C overnight) were groundmanually and ashed in a furnace at 450 °C for 4 h. Ashed samples weredigested in 9 ml conc. HNO3 and 0.5 ml conc. HF (Aristar, Merck,Poole, UK) in Teflon microwave digestion vessels in a MARS 5 (CEM,Buckingham, UK) microwave digestion system (Yafa and Farmer,2006). Lead concentrations and isotope ratios (206Pb/207Pb, 208Pb/206Pband 208Pb/207Pb) in the sample solutions were determined byquadrupole inductively coupled plasma-mass spectrometry (ICP-MS)

cation of the sites used for peat and soil sampling in the Abisko area.

0 2 4 6

Pb (mg/kg)

Pb

206Pb/207Pb

206Pb/207Pb

80

60

40

20

01.15 1.20 1.25

0 2 4 6

1.15 1.20 1.25

25

20

15

10

5

0

Cum

ulat

ive

peat

mas

s(k

g m

-2)

Cum

ulat

ive

peat

mas

s(k

g m

-2)

0 10 20 30 400

5

10

15

20

252000 1950 1900 1850 1800

0 10 20 30 40 500

20

40

60

802000 1950 1900 1850 1800

Date

DateDepth (cm)

Depth (cm)

b

a

Fig. 2. Lead concentrations (square), 206Pb/207Pb ratios (circle) in the panel on the lefthand side and the 210Pb dates in relation to the depth measured in cm in the panel onthe right hand side for a) the sphagnum hummock peat core (C8) and b) the lichenhummock peat core (C1).

4346 J. Klaminder et al. / Science of the Total Environment 409 (2011) 4344–4350

using an Agilent 7500ce (with octopole reaction system) instrument(Agilent Technologies, Stockport, UK). A common Pb isotopic referencematerial from theNational Institute of Standards andTechnology (NIST),SRM 981, was used for mass bias correction. Employing appropriatedilutions and bracketing standards in the 5–80 μg l−1 concentrationrange, sample concentrations were calculated using isotope analysisacquisition and the fully quantitative mode (three points per unit mass,with integration times of 0.3 s per point and five replicate runs persample). For reference materials NIMT/UOE/FM001 (peat) and NCSDC73349 (GBW07603) (bush, branches and leaves), mean Pb values(n=6) obtained were 178±7 mg kg−1 (certified value 174±8 mg kg−1) and 40±4 mg kg−1 (certified value 47±3 mg kg−1),respectively. Overall analytical precision for Pb determination induplicate peat samples (n=27) in the 0.7–6.2 mg kg−1 concentrationrange averaged ±17%. The mean isotope ratios of 206Pb/207Pb, 208Pb/206Pb and 208Pb/207Pb determined inNIMT/UOE/FM/001 (n=6,±1 SD)as 1.176±0.003, 2.094±0.005 and2.463±0.0011, respectively,were ingood agreementwith corresponding reported “information only” valuesof 1.176±0.001, 2.092±0.002 and 2.461±0.003 (Yafa et al., 2004). ForNCS DC73349 (GBW07603), which is not certified for Pb isotopes, meanvalues (n=6,±1 SD) of 1.171±0.003, 2.106±0.003 and 2.467±0.008were obtained for 206Pb/207Pb, 208Pb/206Pb and 208Pb/207Pb, respective-ly, in good agreement with the values of 1.170±0.001, 2.109±0.002and 2.466±0.002 reported previously (Farmer et al., 2010). Overallanalytical precision for Pb isotope ratio determination in duplicate peatsamples (n=27) averaged ±0.3%.

For soil, Pb concentrationsand stable Pb isotope composition (206Pb/207Pb) of freeze-dried homogenized soil samples (0.1 g) were deter-mined using ICP-MS (Perkin-Elmer model ELAN 6100) after a strongacid digestion (conc. HNO3+conc. HClO4, 10:1) for 2–5 h at b130 °C inopen Teflon vessels. Concentrations of Pb were verified against thecertified multi-element standard, SPEX ICPMS-2 (SPEX CertiPrepCertified Reference materials). A ten-point calibration within therange of 0.5 to 75–80 μg l−1 was used. Analyses of 206Pb and 207Pbwere made using dwell times of 50 ms and the NIST SRM 981 was usedfor mass bias correction. The method was validated via the analysis of acertified lake sediment (IAEA SL-1), where a mean Pb concentration±SD value of 35±2mg kg−1 (certified 37±8mg kg−1) and a 206Pb/207Pbratio of 1.213±0.002 were obtained. The latter ratio agreed well withpreviously reported values of 1.214±0.012 (Viczian et al., 1990) and1.214±0.006 (Farmer et al., 1996). Average Pb concentrations andthe 206Pb/207Pb and 208Pb/207Pb ratios of some of the podzolic profileshave previously been published as supporting information (Klaminder etal., 2010). Freeze-dried soil samples, treated with 4 M HCl to removecarbonates,were analyzed for carbon content using aPerkin-ElmerCHNS/O analyzermodel 2400. The carbon content (%)was used tonormalize thePb concentration against themineralmatter content,where the latterwascalculated as [(−1.724×carbon content)+100] %.

Lead-210 activities of the peat core sections were obtained directlyby gamma-spectrometry for core C8 and indirectly for core C1 viaalpha-spectrometry analysis of the decay series descendant 210Po.

Samples for gamma spectrometry analysis, with masses in therange 1 to 5 g, were compressed into disks of uniform geometry usinga 12 tonehydraulic press, afterwhich theywere sealed in polycarbonatecontainers and stored for at least three weeks to ensure radioactiveequilibrium between 222Rn and 226Ra. Gamma spectra were recordedusing an Ortec low background, co-axial HPGe detector (diameter57.3 mm; length 27.6 mm; resolution 675 eV at 122 keV) housed in a10 cm thick Pb shield with a Cd–Cu lining. The total 210Pb activity wasderived from its photopeak at 46.5 keV,while the supported componentwas calculated using the 214Pb/214Bi peaks at 295, 352 and 609 keV.Detection efficiencies for the sampleweights usedwere calculatedusingpeat, fromadepth containingnon-detectable levels of unsupported 210Pb,spiked with known activities of 210Pb and 226Ra.

Alpha-spectrometry was carried out using an alpha spectrometer(Canberra 4701 vacuum chamber) based on a planar silicon (PIPS)

detector of 450 mm2 surface area, ~20% counting efficiency and20 keV resolution. The peat was digested in hydrogen peroxide at90 °C to near dryness and leached with 1 M HCl at 90 °C for 1 h. Thepeat samples from core C1were counted for N1000 min, with a typicalchemical recovery ranging from 60 to 100% estimated using 209Po as ayield tracer. Excess activities of 210Pb (unsupported by in situ sources)were calculated by subtracting background activities from the total.

Lead-210 dates were derived using the Constant Rate of Supply(CRS) model described in detail elsewhere (Appleby and Oldfield,1978). Care was taken to ensure that there was no loss of surfacevegetation to ensure that the 210Pb contained in/on the vegetationwas included in the calculated inventory [i.e. specific activity(Bq kg−1) of section×section mass (kg)/cross-sectional area (m2)].The 210Pb activities of the surface vegetation samples were includedbecause interception of 210Pb by living plants may significantly affectthe 210Pb inventories used in the CRS dating model (Olid et al., 2008).

3. Results

3.1. Ombrotrophic peat cores

Lead concentrations of the ombrotrophic peat ranged from 0.6 to6.2 mg kg−1 (Fig. 2a,b, Tables S1, S2). Highest concentrations werefound as sub-surface peakswithin the ombrotrophic section of the coresat a cumulative peatmass (depth)of about13 kg m−2 for the sphagnumhummock (Fig. 2a) and at about 9 kg m−2 for the lichen hummock(Fig. 2b). For the ombrotrophic part of the cores, the 206Pb/207Pb ratiovaried between 1.15 and 1.17, while minerotrophic peat close to theunderlying sediment had a ratio of 1.20–1.26. However, minerotrophicpeat of low 206Pb/207Pb ratio~1.15 was also found in the lichenhummock at a cumulative peat mass depth of about 36 kg m2.

Previous reconstructed temporal variations in the 206Pb/207Pbratio of atmospheric Pb and the atmospheric Pb deposition rates for anorthern boreal site (Klaminder et al., 2006) are shown in Fig. 3a,representative of the established contamination trend in Europe. Alsoshown are the reconstructed temporal trends in Pb accumulation rate

4347J. Klaminder et al. / Science of the Total Environment 409 (2011) 4344–4350

(Fig. 3b) and in the 206Pb/207Pb ratio of the peat from the studied sub-arctic site (Fig. 3b,c). As with the Pb deposition rates in the borealregion, the Pb accumulation rates, calculated as the peat accumulationrate multiplied by the Pb concentration, began to increase at the startof the 19th century and peaked (~0.5–1.3 mg m−2 yr−1) between1960 and 1975 (Fig. 3b). In comparison, the Pb accumulation rates(0.1-0.3 mg m-2 yr-1) in more recently deposited peat are similar tothose found at the beginning of the 20th century but slightly higherthan the typical pre-1800 values of ~0.1 mg m−2 yr−1 (Fig. 3b).

For peat deposited since 1900, the 206Pb/207Pb ratio showed agenerally similar temporal trend in the two cores (Fig. 3c) and followedthe general atmospheric trend in Europe. Here values of the 206Pb/207Pbratio decreased from ~1.16 in 1900 to values between 1.14 and 1.15 forthe period 1920–1985, after which the ratio increased again towardsvalues of ~1.17 in peat deposited during the last few years. In earlieryears, however, the 206Pb/207Pb ratio of the lichen hummock peatclearly departed from the expected atmospheric trend, exhibiting adecrease towards values as low as 1.15 in the 18th century.

3.2. Tundra soil profiles

The mineral soil of the sampled tundra profiles had a 206Pb/207Pbratio that varied from 1.25 to 1.38 (Fig. 4a). In contrast,the 206Pb/207Pb ratio of the material within the O-horizon of thepodzolic soil profiles was less variable at 1.14–1.18, which wasdistinct from the underlying mineral soil, but overlaped with that ofthe ombrotrophic peat (Fig. 2a,b). The average 206Pb/207Pb ratio of the

Fig. 3. a) Temporal trends in the 206Pb/207Pb ratio of atmospheric Pb and the Pbdeposition rate reconstructed for the boreal zone (Klaminder et al., 2006) and b) thecalculated temporal variations in Pb accumulation rates in the lichen and spaghnumhummock peat-cores and c) the temporal variations in the 206Pb/207Pb ratio in thesetwo cores. The age of peat deposited before 1900 is calculated using extrapolated peataccumulation rates.

O-horizon of the podzolic soils was less variable and typically around1.170±0.002 (mean±SD). The cryosols differed from the podzolicsoils, exhibiting greater variability in 206Pb/207Pb ratios at depth in themineral soil and having a higher 206Pb/207Pb ratio of 1.18±0.02 in theO-horizon. In the plot of 206Pb/207Pb ratio against 208Pb/207Pb ratio(Fig. 5), the positions of samples from the O-horizon of the podzolicsoils overlapped with those of the ombrotrophic peat, while samplesfrom the O-horizon of the cryosols plotted along a mixing-linebetween the ombrotrophic peat and Pb from the local soil minerals.

Lead concentrations in the O-horizonmatrix (2–23 mg kg−1) werewithin the range of those of the deeper (N50 cm) underlying mineralsoil (Fig. 4b), but the mineral matter in the O-horizon was highlyenriched in Pb in comparison with that of the mineral soil (Fig. 4c).The Pb concentration in the O-horizon was higher than that of theupper part (~10 cm) of the mineral soil of the podzolic soils, the latterconsisting of a weathered E-horizon slightly depleted in Pb incomparison with the deeper mineral soil. The depletion in Pb of theE-horizon relative to the deeper and less weathered soil was moreevident at the moderate to high precipitation sites. At the latter sites,the O-horizon also had a sub-surface peak in Pb similar to that of thepeat cores (Fig. 4b,c).

4. Discussion

4.1. Long-range contamination of the atmosphere at the study area

Atmospheric fallout of Pb in ancient times as a result of the Romanmining era ca. 2000 years ago has been detected in polar ice cores(Rosman et al., 1997; Zheng et al., 2007). The studied ombrotrophicpeat cores, providing a chronology of atmospheric inputs over the last~200 years, are therefore insufficient to cover a significant part of thepollution history in the area. Lake sediment studies have revealed thatthe importance of ancient Pb gradually decreases towards the north inScandinavia and is hardly detectable in sediments from northernScandinavia due to significant dilution from local Pb-containingminerals (Brännvall et al., 1999). In line with this, the first clearlydetectable inputs of Pb in sub-arctic lakes around our study areaappear to occur during the onset of the medieval mining industry ca.900 AD, while the main input has occurred since the beginning of the20th century (Klaminder et al., 2010). In other words, the peat corescover the most intense phase of anthropogenic contamination in thestudy area.

For the last two centuries, the reconstructed trends in Pbaccumulation rate and 206Pb/207Pb ratio using the two peat coresfollow the established pollution history in Europe (Fig. 2a–c), wherethe anthropogenically driven Pb deposition rates accelerated in the20th century, peaked around the 1960s–70s and then rapidly declinedbetween the 1970s and 2000s (Rühling and Tyler, 2001; Bindler,2006; Cloy et al., 2008). As pollution sources intensified and changed,the 206Pb/207Pb of the atmosphere decreased from ~1.16–1.17 at thebeginning of the 20th century to ~1.12–1.14 during 1930–1980 andthereafter increased again to approach pre-20th century values(Farmer et al., 2002). Similar trends between the pollution historyin Europe (Fig. 3a) and the studied peat cores (Fig. 3bc) suggest thatatmospheric Pb deposition at the study sites was driven mainly bylong-range transported aerosols derived from sources at lowerlatitudes. Importantly, the fallout at the studied site has been low asindicated by a five times lower Pb accumulation rate in peat duringthe 20th and 21st centuries (Fig. 3b) in comparison with atmosphericdeposition rates estimated for the boreal site (Fig. 3a).

Some uncertainties in the peat record, however, need to behighlighted. While the peat core from the ombrotophic part of thesphagnum hummock indicates a stable 206Pb/207Pb ratio of ~1.17 inthe peat deposited in the 19th century, which is a trend seenelsewhere in Europe (Cloy et al., 2008), the lowest 206Pb/207Pb ratio inthe lichen hummock of ~1.15 occurs in fen peat deposited in the 18th

Fig. 4. a) The 206Pb/207Pb ratio of the studied cryosol and podzol tundra profiles (left panel), b) the Pb concentration in these tundra soil profiles (middle panel) and c) the Pbconcentration normalized to the mineral matter content.

4348 J. Klaminder et al. / Science of the Total Environment 409 (2011) 4344–4350

century. The latter trend cannot be explained from the Europeanpollution history and the reconstructed fallout history appears to havebeen affected by Pb mobilized and transported downwards in thepeat. Desiccation cracks commonly occur on palsa lichen hummocksuplifted by permafrost (Seppälä, 1986) and preferential flow ofcontaminant Pb along these cracks, and subsequent accumulation ofPb deeper in the peat, is a likely explanation of the observed trend.Mass-balance calculations suggest that downward translocation ofonly 4–9 mg m−2 of Pb from the surface peat, having a 206Pb/207Pbratio of 1.14–1.15, would be enough to generate the decrea-sing 206Pb/207Pb ratio trend at depth between a cumulative peatmass of 27–36 kg m−2 (Fig. 2b), assuming that the peat originally hada ratio around 1.16, similar to that of peat deposited in the sphagnumhummock during this time (Fig. 2a). Indications of such smalldownward redistribution of Pb can also be visualized in Fig. 3b,where the accumulation rate in the lichen hummock is lower duringthe 1935–1985 period than in the sphagnum hummock, butsystematically higher in the older peat layers. That cryoturbicprocesses might negatively affect the chronology of peat cores haspreviously been hypothesized (Bindler et al., 2005). Yet thesimilarities in the main reconstructed temporal trends in the Pbaccumulation rate and the 206Pb/207Pb ratio of the peat, as seenbetween the cores and reconstructed trends using other independent

Fig. 5. The average 206Pb/207Pb and 208Pb/207Pb ratio of the O-horizons of the twostudied soil types (cryosols, n=4; and podzols n=6) in comparison with ombro-trophic peat (approximating the isotopic composition of Pb in atmospheric aerosols)and bulk soil samples from the mineral soil, where samples sampled at depth N50 cmare likely to approximate the isotopic composition of Pb in local soil minerals.

records, such as lake sediments and preserved herbarium samples(Fig. 3a), suggest that the major chronology of Pb fallout during thelast two centuries is preserved within the records.

4.2. Importance of atmospheric deposition for Pb levels in the O-horizon

Lead concentrations in the O-horizon of the studied podzolictundra soil of the cryoturbated profiles are 8±5 mg kg−1 and 10±7 mg kg−1, respectively (Fig. 4b). These concentrations are more thanthree orders of magnitude lower than concentrations found in theO-horizon close to local pollution sources such as smelters (Klaminderet al., 2008b) and much lower than the average concentration of69±3 mg kg−1 found for the O-horizon of podzolic soils formedwithin the Swedish boreal forest (Alriksson, 2001). Interpretationsbased solely on the Pb concentration of the O-horizon would falselysuggest that the studied tundra soils are representative of naturalbackground conditions. However as argued below, the isotopiccomposition of the O-horizon clearly indicates that anthropogenicPb from long-range transported aerosols dominates the Pb cyclingin this soil compartment of the tundra soil rather than local soilminerals.

Ombrotrophic peat is a recognized archive of atmospheric Pbfallout (Kylander et al., 2010). That the 206Pb/207Pb and 208Pb/207Pbratios of the O-horizon of the podzolic tundra soil are distinct fromthose of the mineral soil and overlap with those of the ombrotrophicpeat, which represents the average isotopic composition of anthro-pogenic derived aerosols (Fig. 5), indicates only a small influence of Pbfrom the underlying mineral matrix upon the O-horizon for thepodzolic profiles. The proportion of Pb in the O-horizon that is derivedfrom local soil minerals can be estimated using a simple binarymixingmodel where the 206Pb/207Pb ratio of the upper mineral soil(206Pb/207Pb ratio 1.25; Fig. 4a) and the average isotopic compositionof the ombrotrophic peat (206Pb/207Pb ratio 1.16; Fig. 2a,b) serve asreference values of the local soil minerals and the atmosphericderived Pb, respectively. Using these two end-members, we estimatethat about 11% of the Pb in the O-horizon is derived from local soilminerals in order to explain the 206Pb/207Pb ratio of about 1.17 in theO-horizon. Hence, the input of Pb from the mineral soil to the O-horizon of podzolic tundra soils must be small in comparison withatmospheric deposition rates. Similar conclusions have been drawnfrom boreal forest sites where the 206Pb/207Pb ratio of the O-horizon istypically ~1.15–1.17 while the corresponding ratio of the underlyingmineral soil varies from 1.3 to 2.3 (Bindler et al., 2008; Klaminder etal., 2008a). Small uptake fluxes of Pb from the mineral soil relative toatmospheric deposition rates have also been indicated by studiescombining the use of isotopic tracers and plant growth rates, whichhave suggested that only 0.002–0.04 mg Pb m−2 yr−1 is translocated

4349J. Klaminder et al. / Science of the Total Environment 409 (2011) 4344–4350

through plant uptake from the mineral soil to the O-horizon(Klaminder et al., 2008a; Hovmand et al., 2009). As a comparison,atmospheric Pb deposition rates are commonly one to two orders ofmagnitude higher in northern Sweden (Rühling and Tyler, 2001). Lowplant uptake rates of Pb from the mineral soil can be explained by alow bioavailability of Pb species in the soil (Turpeinen et al., 2000) anda predominant uptake of mainly atmospheric Pb through root uptakein the O-horizonwhere atmospheric derived Pb dominates or throughinterception of anthropogenic derived aerosols (Klaminder et al.,2008a).

Plant ‘pumping’ from themineral soil has been advocated to be themain process explaining the enrichment of Pb in surface soils even inhighly urban areas (Reimann et al., 2008). Biological fractionation hasbeen hypothesized to be a possible mechanism of a harmoni-zed 206Pb/207Pb ratio of the O-horizon, a hypothesis that has beenhighly questioned (Bindler, 2008; Le Roux et al., 2008; Shotyk, 2008).Biological processes such as plant uptake fractionate toward isotopesof lower atomic mass (von Blanckenburg et al., 2009), however, andwould, if operating, generate higher 206Pb/207Pb ratios for the O-horizon than for the underlying mineral soil — a trend in contrast tothat observed (Fig. 4a). On the other hand, the bulk mineral soilconsists of minerals having variable 206Pb/207Pb ratios (Harlavan andErel, 2002) and a selective plant uptake of Pb from minerals havinga 206Pb/207Pb ratio of~1.16-1.17 could explain the isotopic composi-tion of the O-horizon, even though this systematic uptake, indepen-dent of geological matrix, appears less likely. In fact, Pb with ahigher 206Pb/207Pb ratio than that of the bulk soil is generally moreeasily extractable (Harlavan and Erel, 2002), which would make thisPb more likely to be pumped to the O-horizons from the underlyingmineral soil and thus contribute to higher 206Pb/207Pb ratios in the O-horizon than in the bulk soil — again a trend in contrast to thatobserved (Fig. 4a). A preferential weathering loss of Pb fromradiogenic minerals having higher 206Pb/207Pb could explain theslightly lower 206Pb/207Pb ratio of the bulk soil from the upper ~5 cmof the mineral soil (Fig. 4a). It is more likely, however, that thelower 206Pb/207Pb ratio of the top horizon of themineral soil is generatedby a small degree of downward transport of contaminant Pb from the O-horizon. An active downward transport of Pb from the surface soil is alsoindicated by the generally higher Pb concentrations in the upper 0.6 m ofthe soil in comparison with the parent material (Fig. 4b,c).

Local soil minerals constitute a more important source of Pb in theO-horizon of the cryosols than in the podzolic soil, as indicated bysome O-horizon samples having 206Pb/207Pb ratios around 1.20,approaching values found for the underlying mineral soil (Fig. 4a).However, the average 206Pb/207Pb ratio of 1.18±0.02 for the O-horizon is only slightly higher than that of 1.16±0.02 for theombrotrophic peat, suggesting that the Pb in the O-horizon of thecryosols mainly has an atmospheric origin. A binary mixing modelbetween Pb derived from the upper mineral soil, with a 206Pb/207Pbratio of 1.25 (Fig. 4a), and Pb from atmospheric sources, with a ratio of1.16, suggests that about 22% of the Pb in the O-horizon is typicallyderived from local soil mineral soil and the remaining 78% fromatmospheric deposition to produce the observed average 206Pb/207Pbratio of 1.18. Given that the atmospheric deposition rate of about0.5 mg Pb m−2 yr−1 is known, as indicated by the peat record (Fig. 2a,b), the total input rate of Pb to the O-horizon can be calculated as theannual deposition rate divided by the relative contribution fromatmospheric deposition. This calculation suggests a total input rate ofabout 0.64 mg Pb m−2 yr−1 and that the upward transportation rate,calculated as the total input rate less the deposition rate, would haveto be around 0.14 mg Pb m−2 yr−1. In the O-horizon of the podzolicsoils, where about 11% of the Pb is estimated to be derived from localsoil minerals, a similar series of calculations suggests that only about0.06 mg Pb m−2 yr−1 is transported from the mineral soil to the O-horizon. Clearly, the higher upward transportation rates in thecryosols indicate that soil mixing is a more effective transporter of

Pb from the mineral soil into the O-horizon than plant uptaketranslocation.

5. Conclusions

Although Pb concentrations in the O-horizon of high latitudetundra soil are an order ofmagnitude lower than those found in borealforests at lower latitudes, they are not representative of levelsexpected in a pristine environment. Indeed, most (≥80%) of the Pbappears to be derived from anthropogenic sources, as indicated bythe 206Pb/207Pb and 208Pb/207Pb ratios of the O-horizon, whichapproach those of atmospheric derived contaminants within thepeat core record. Although natural processes such as plant uptake andsoil mixing can both contribute Pb to the O-horizon, the latter appearsto be a more efficient transporter of geogenic Pb from the mineral soilto the O-horizon than plant uptake, especially for cryosolic soils. Itseems, therefore, of more importance to quantify soil mixing ratesrather than plant cycling rates when investigating the impact of localsoil minerals upon the O-horizon geochemistry of tundra soils.

Acknowledgements

We thank L. Jörnhagen and A. Zackrisson (Umeå) for fieldworkassistance and J.M. Cloy, F.S. Douglas, L.J. Eades (UoE) and C. Donnelly(SUERC) for assistance in the laboratory. We also thank fouranonymous reviewers for their helpful comments on this paper.Financial support was provided by the Swedish Research Council(project no 2009-3282).

Appendix A. Supplementary data

Supplementary data to this article can be found online at doi:10.1016/j.scitotenv.2011.07.005.

References

Alriksson A. Regional variability of Cd, Hg, Pb and C concentrations in different horizonsof Swedish forest soils. Water Air Soil Pollut Focus 2001;1:325–41.

Appleby PG, Oldfield F. The calculation of 210Pb dates assuming a constant rate of supplyof unsupported 210Pb to the sediment. Catena 1978;5:1–8.

Bindler R. Mired in the past — looking to the future: geochemistry of peat and theanalysis of past environmental changes. Glob Planet Change 2006;53:209–21.

Bindler R. Comment on “The biosphere: A homogeniser of Pb-isotope signals” by C.Reimann, B. Flem, A. Arnoldussen, P. Englmaier, T.E. Finne, F. Koller and O.Nordgulen. Appl Geochem 2008;23:2519–26.

Bindler R, Anderson NJ, Renberg I, Malmquist C. Paleolimnological investigation ofatmospheric pollution in the Søndre Strømfjord region, southern West Greenland:accumulation rates and spatial patterns. Geol Greenland Survey Bull 2001;189:48–53.

Bindler R, Cortizas AM, Blaauw M. Comment on “Atmospheric mercury accumulationrates between 5900 and 800 calibrated years BP in the High Arctic of Canadarecorded by peat hummocks”. Environ Sci Technol 2005;39:908–9.

Bindler R, Renberg I, Klaminder J. Bridging the gap between ancient metal pollution andcontemporary biogeochemistry. J Paleolimn 2008;40:755–70.

Boutron CF, Candelone JP, Hong S. The changing occurrence of natural andman-derivedheavy-metals in Antarctic and Greenland ancient ice and recent snow. Int J EnvironAnal Chem 1994;55:203–9.

Brännvall M-L, Bindler R, Renberg I, Emteryd O, Bartnicki J, Billström K. The Medievalmetal industry was the cradle of modern large-scale atmospheric lead pollution innorthern Europe. Environ Sci Technol 1999;33:4391–5.

Cloy JM, Farmer JG, Graham MC, MacKenzie AB, Cook GT. Historical records ofatmospheric Pb deposition in four Scottish ombrotrophic peat bogs: an isotopiccomparison with other records from western Europe and Greenland. GlobBiogeochem Cycle 2008;22:GB2016. doi:2010.1029/2007GB003059.

Farmer JG, Eades LJ, MacKenzie AB, Kirika A, Bailey-Watts TE. Stable lead isotope recordof lead pollution in Loch Lomond sediments since 1630 A.D. Environ Sci Technol1996;30:3080–3.

Farmer JG, Eades LJ, Atkins H, Chamberlain DF. Historical trends in the lead isotopiccomposition of archival Sphagnummosses from Scotland (1838–2000). Environ SciTechnol 2002;36:152–7.

Farmer JG, Eades LJ, Graham MC, Cloy JM, Bacon JR. A comparison of the isotopiccomposition of lead in rainwater, surface vegetation and tree bark at the long-termmonitoring site, Glensaugh, Scotland, in 2007. Sci Total Environ 2010;408:3704–10.

Gong SL, Barrie LA. Trends of heavy metal components in the Arctic aerosols and theirrelationship to the emissions in the Northern Hemisphere. Sci Total Environ2005;342:175–83.

4350 J. Klaminder et al. / Science of the Total Environment 409 (2011) 4344–4350

Harlavan Y, Erel Y. The release of Pb and REE from granitoids by the dissolution ofaccessory phases. Geochim Cosmochim Acta 2002;66:837–48.

Hong SM, Candelone J-P, Patterson CC, Boutron CF. Greenland ice evidence ofhemispheric lead pollution 2 millennia ago by Greek and Roman civilizations.Science 1994;265:1841–3.

Hovmand MF, Nielsen SP, Johnsen I. Root uptake of lead by Norway spruce grown onPb-210 spiked soils. Environ Pollut 2009;157:404–9.

Klaminder J, Yoo K. Contaminants as tracers for Studying Dynamics of Soil Formation:mining an ocean of opportunities. Adv Agro 2008;100:15–57.

Klaminder J, Renberg I, Bindler R, Appleby P, Emteryd O, Grip H. Estimating the meanresidence time of lead in the organic horizon of boreal forest soils using 210-lead,stable lead and a soil chronosequence. Biogeochem 2006;78:31–49.

Klaminder J, Bindler R, Renberg I. The biogeochemistry of atmospherically derived Pb inthe boreal forest of Sweden. Appl Geochem 2008a;23:2922–31.

Klaminder J, Bindler R, Rydberg J, Renberg I. Is there a chronological record ofatmospheric mercury and lead deposition preserved in the mor layer (O-horizon)of boreal forest soils? Geochim Cosmochim Acta 2008b;72:703–12.

Klaminder J, Hammarlund D, Kokfelt U, Vonk JE, Bigler C. Lead contamination ofsubarctic lakes and its response to reduced atmospheric fallout: can the recoveryprocess be counteracted by the ongoing climate change? Environ Sci Technol2010;44:2335–40.

Kylander ME, Klaminder J, Bindler R, Weiss DJ. Natural lead isotope variations in theatmosphere. Earth Planet Sci Lett 2010;290:44–53.

Le Roux G, Sonke JE, Cloquet C, Aubert D, de Vleeschouwer F. Comment on “Thebiosphere: A homogeniser of Pb-isotope signals” by C. Reimann, B. Flem, A.Arnoldussen, P. Englmaier, TE Finne, F. Koller and O. Nordgulen. Appl Geochem2008;23:2789–92.

Malmer N, Johansson T, Olsrud M, Christensen TR. Vegetation, climatic changes and netcarbon sequestration in a North-Scandinavian subarctic mire over 30 years. GlobalChange Biol 2005;11:1895–909.

Murozumi M, Chow TJ, Patterson CC. Chemical concentrations of pollutant leadaerosols, terrestrial dusts and sea salts in Greenland and Antarctic snow strata.Geochim Cosmochim Acta 1969;33:1247–94.

Olid C, Garcia-Orellana J, Martinez-Cortizas A, Masque P, Peiteado E, Sanchez-Cabeza JA.Role of surface vegetation in Pb-210-dating of peat cores. Environ Sci Technol2008;42:8858–64.

Reimann C, Flem B, Arnoldussen A, Englmaier P, Finne TE, Koller F, et al. The biosphere:a homogeniser of Pb-isotope signals. Appl Geochem 2008;23:705–22.

Rosman KJR, Chisholm W, Boutron CF, Candelone JP, Hong S. Isotopic evidence toaccount for changes in the concentration of lead in Greenland snow between 1960and 1988. Geochim Cosmochim Acta 1994;58:3265–9.

Rosman KJR, Chisholm W, Hong SM, Candelone JP, Boutron CF. Lead from Carthaginianand Roman Spanish mines isotopically identified in Greenland ice dated from 600BC to 300 AD. Environ Sci Technol 1997;31:3413–6.

Rühling Å, Tyler G. Changes in atmospheric deposition rates of heavy metals in Sweden.A summary of nationwide Swedish surveys in 1968/70 – 1995. Water Air Soil PollutFocus 2001;1:311–23.

Seppälä M. The origin of palsas. Geogr Ann A 1986;68:141–7.Shotyk W. Comment on “The biosphere: A homogeniser of Pb-isotope signals” by C.

Reimann, B. Flem, A. Arnoldussen, P. Englmaier, T.E. Finne, F. Koller and O.Nordgulen. Appl Geochem 2008;23:2514–8.

Shotyk W, Zheng JC, Krachler M, Zdanowicz C, Koerner R, Fisher D. Predominance ofindustrial Pb in recent snow (1994–2004) and ice (1842–1996) from Devon Island,Arctic Canada. Geophys Res Lett 2005;32.

Steinnes E, editor. Impact of long-range atmospheric transport of heavy metals to theterrestrial environment in Norway. (Vol. SCOPE 31). New York: John Wiley & SonsLtd.; 1987.

Steinnes E. Trace element profiles in ombrogenous peat cores fromNorway: evidence oflong range atmospheric transport. Water Air Soil Pollut 1997;100:405–13.

Steinnes E, Friedland AJ. Metal contamination of natural surface soils from long-rangeatmospheric transport: existing and missing knowledge. Environ Rev 2006;14:169–86.

Turpeinen R, Salminen J, Kairesalo T. Mobility and bioavailability of lead incontaminated boreal forest soil. Environ Sci Technol 2000;34:5152–6.

Viczian M, Laszity A, Barnes RM. Identification of potential environmental sources ofchildhood lead-poisoning by inductively coupled plasma mass-spectrometry —

verification and case-studies. J Anal At Spectrom 1990;5:293–300.von Blanckenburg F, vonWiren N, Guelke M,Weiss DJ, Bullen TD. Fractionation of metal

stable isotopes by higher plants. Elements 2009;5:375–80.Wardenaar ECP. A new hand tool for cutting peat profiles. Can J Bot 1987;65:1772–3.Yafa C, Farmer JG. A comparative study of acid-extractable and total digestion methods

for the determination of inorganic elements in peat material by inductively coupledplasma-optical emission spectrometry. Anal Chim Acta 2006;557:296–303.

Yafa C, Farmer JG, Graham MC, Bacon JR, Barbante C, Cairns WRL, et al. Development ofan ombrotrophic peat bog (low ash) reference material for the determination ofelemental concentrations. J Environ Monit 2004;6:493–501.

Zheng JC, Shotyk W, Krachler M, Fisher DA. A 15,800-year record of atmospheric leaddeposition on the Devon Island Ice Cap, Nunavut, Canada: natural andanthropogenic enrichments, isotopic composition, and predominant sources.Glob Biogeochem Cycle 2007;21:GB2027. ISI:000247533700001.