trace-metal pollution of soils in northern england

Trace-metal pollution of soilsin northern EnglandB. G. Rawlins Æ T. R. Lister Æ A. C. Mackenzie

Abstract Data from a regional geochemical surveyof topsoils (n=818) in rural and peri-urban areasover a single parent material (Coal Measures) areused to identify two types of trace-metal pollution –severe local contamination at 20 sites and wide-spread, diffuse pollution in more densely populatedareas. Median concentrations of several trace metalsin topsoils were significantly higher in areas of high,compared to low, population density (percentageincreases in parenthesis): As (31), Cu (39), Fe (7),Mo (26–36), Ni (29), Pb (20), Sn (40), and Zn (11).Four potential pathways of diffuse trace-metalpollution are postulated: coal-ash dispersal, atmo-spheric aerosols derived from coal combustion, thehistorical spreading of sewage waste, and thoserelated to road vehicles. The statistical analysis ofgeochemical data classified by local, human popu-lation density can be an effective means of identi-fying the magnitude and extent of diffuse pollution,and could help to establish natural backgroundlevels.

Keywords Diffuse pollution Æ Backgroundlevels Æ Trace metals Æ Soil survey Æ Geochemistry ÆCoal measures

Introduction

Soil geochemistry is determined at a range of spatial scales,from national or regional surveys to local, site-specificinvestigations. Regional soil geochemical surveys haverecently been used to determine the magnitude and extentof areas affected by sources of anthropogenic contamina-

tion (Reimann and others 2000). At the local scale, siteinvestigations are undertaken to establish the total con-centration of a range of trace metals (heavy metals andmetalloid elements) in soil which may pose a risk to hu-man health, or cause the pollution of controlled waters. Inmany countries trace-metal concentrations are measuredin areas where sewage sludge is applied to agricultural landto ensure that guideline values are not exceeded. By es-tablishing some form of background concentration fortrace metals from a regional soil survey, the results oflocal, site-specific measurements can be put into context.Environmental geochemical data from soil surveys un-dertaken at the regional scale generally have non-normaldistributions which are often positively skewed (Reimannand Filzmoser 2000). Statistical techniques have beendeveloped to identify thresholds to separate ‘background’concentrations from sites which have naturally elevatedlevels or indicate some form of anthropogenic contami-nation, although there is no universally adopted method(Matschullat and others 2000). Natural or backgroundconcentrations of trace metals vary widely, even in un-disturbed soils; for example, lead (Pb) concentrations intopsoils throughout rural areas of England and Walesrange from 20 to approximately 120 mg kg–1 between the10 and 90th percentiles of the sample distribution re-spectively (McGrath and Loveland 1992). The primarycontrol on trace-metal contents of undisturbed soil intemperate regions such as the United Kingdom is typicallythe geochemical composition of the soil parent material(the bedrock geology or Quaternary deposit from which itformed).In its definitions of soil quality, the International Stan-dards Organisation distinguishes between the naturalbackground concentration, ‘.... derived solely from naturalsources’, and the background concentration which ‘arisesfrom both natural sources and non-natural diffuse sourcessuch as atmospheric deposition’ (ISO 1996). Although thelatter definition is instructive in limiting the inclusion ofcontaminants to those from diffuse sources, neither pro-vides any guidance on methods which could be used toestablish the range or upper limit of background values.One way of establishing background concentrations oftrace metals in soils is to define upper and lower limitsusing values from the analysis of samples from relativelypristine areas with the same parent material, which areneither highly mineralised nor subject to significantanthropogenic contamination. Elevated trace-metal

Received: 27 June 2001 / Accepted: 30 January 2002Published online: 6 April 2002ª Springer-Verlag 2002

B.G. Rawlins (&) Æ T.R. Lister Æ A.C. MackenzieBritish Geological Survey,Keyworth, Nottingham, NG12 5GG, UKE-mail: [email protected]

Original article

612 Environmental Geology (2002) 42:612–620 DOI 10.1007/s00254-002-0564-5

concentrations are often found in soils developed overmineralised bedrock, and methods have been devised toidentify thresholds in geochemical data for the purpose ofmineral exploration (Sinclair 1974). Several methods havebeen proposed for establishing the range of backgroundvalues by removing the highest and lowest values from thesample distribution (Matschullat and others 2000). Forexample, the ‘natural geochemical baseline’ has been de-fined by Tidball and Ebens (1976) as encompassing thecentral 95% of the observed concentrations (i.e. from 2.5to 97.5% of the distribution). By statistical analysis of adata set, it may be possible to identify a threshold betweenbackground values and a limited number of high trace-metal concentrations (resulting from severe contaminationor mineralisation). However, in areas with a legacy of at-mospheric pollution where trace metals have been de-posited over considerable distances, the cumulativedistribution may reflect a continuum from almost pristinesites, through sites affected by diffuse contamination, togrossly contaminated sites. In such circumstances, it maybe difficult to identify the subtle impact of diffuse pollu-tion on trace-metal contents. It would also require fargreater effort to remove the impact of diffuse pollution inestablishing natural background concentrations (accordingto the ISO definition).This study focussed on several trace metals (As, Cu, Cr, Ni,Pb and Zn) which are often cited as being of environ-mental concern. The aim of this paper was to investigate adata set from a regional soil survey to identify local anddiffuse trace-metal contamination, and suggest their po-tential sources and pathways. The implications of diffusepollution for establishing natural background concentra-tions of trace metals are discussed.

Study region

The geology of the region consists of Westphalian se-quences (ca. 300 million years old) of the CarboniferousCoal Measures (Fig. 1) and has been described by Downie(1960), whilst the soils have been described by the SoilSurvey of England and Wales (Soil Survey 1984). TheLower and Middle Coal Measures in this region consist ofmudstones, shales and inter-bedded sandstones. Thesandstones and shales alternate to form a succession ofsandstone, seat earth, coal, marine shales, mudstones andsiltstones. Sample sites in the study region occur in abroad range of land-use settings, from semi-natural, ruralenvironments with limited development to the fringes oflarge, urban conurbations. The main land-cover types inareas around the soil sample locations include tilled land(39%), meadow (21%), suburban (18%), and grazed turf(10%; ITE land cover map; Fuller and others 1994). Ele-vation across the region varies between around 50 and250 m above sea level.The soil geochemical atlas of England and Wales showsthat the trace-metal content of the soils developed in theregion are generally high (McGrath and Loveland 1992).Soils throughout parts of northern England were subject to

considerable atmospheric pollution during the lastcentury, resulting from the deposition of aerosols fol-lowing the domestic and industrial combustion of coalmined from the underlying Carboniferous Coal Measures.A resume of industrial activity throughout the region,which gives an indication of the magnitude of coal com-bustion, has been provided by Gilbertson and others(1997). Air pollution in Sheffield (a large conurbation inthe region) at the start of the 20th century was widespreadand intense, and by the middle of the century had con-tributed to the establishment of legislation, most notablythrough the Clean Air Act of 1956. The level of air pollu-tion subsequently declined, as did the size of the steelmanufacturing industry, in the 1970s and 1980s.

Soil survey and chemical analysis

A regional soil geochemical survey covering the entireHumber-Trent region (Fig. 1) has been undertaken by theBritish Geological Survey as part of its G-BASE (Geo-chemical Baseline Survey of the Environment) Pro-gramme. As part of this regional survey, a total of818 topsoil samples were collected over the Lower andMiddle Coal Measures Formations, where there were noextensive Quaternary deposits, ensuring that all soils werederived from the same parent material type. The samplesdescribed in this paper were collected in the summermonths of 1994, 1995 and 1996, and the sites extendedacross four counties (South Yorkshire, West Yorkshire,Derbyshire, Nottinghamshire), covering an area of ap-proximately 2,100 km2. This gives an average sample res-olution of one site per 2.5 km2. Sites were selected on asystematic basis from every second kilometre square of theBritish National Grid. Site selection in each square wasrandom, subject to the avoidance where possible of roads,tracks, railways, human habitation and other disturbedground. Samples were not collected in urban areas, butwere collected around them (in peri-urban areas).At each sample site, five holes were augered at the cornersand centre of a square with a side length of 20 m using ahand auger. There are two main soil sampling approachesfor comparing the concentrations of elements across aregion – the collection of samples from the same soil ho-rizon in each land-use type or, alternatively, sampling overa specified depth range. Sampling specific soil horizons(which occur at different depths across the landscape)would lead to variations in the amounts of any pollutantspresent due to their increased attenuation with depth fromthe surface. It was decided that soil samples would becollected at the same depths at each site, between 0 and15 cm. Soil samples from each of five holes were combinedto form an aggregated sample. At each site, information onthe location, catchment geology, contamination, land useand other features required for data interpretation wasrecorded on a field card and subsequently transferred toan electronic database.All soils were disaggregated following drying and sievedto 2 mm. All samples were coned and quartered, and a

Original article

Environmental Geology (2002) 42:612–620 613

50-g subsample ground in an agate planetary ballmill.The total concentration of 24 major and trace elements(listed in Table 1) was determined in each sample bywavelength dispersive XRF (X-ray fluorescence). Thelower limits of detection for each element are alsoshown. The detection limit for Cd (2.5 mg kg–1) by XRFanalysis was too high to provide reliable data on itsconcentration in samples throughout the study region,and the data has therefore not been included in thispaper.

Throughout the broader regional survey (Fig. 1), twosamples were collected at 84 duplicate sites; the duplicateswere collected immediately adjacent to the original surveysites. Prior to chemical analysis, the duplicate sampleswere split into two subsamples, giving a total of 336 sep-arate analyses. This sampling design is suitable for theapplication of analysis of variance (ANOVA) using anested design (Snedecor and Cochran 1989). The results ofsuch an analysis provide estimates of the components ofvariance accounted for by differences between sites

Fig. 1The study region and samplinglocations

Original article

614 Environmental Geology (2002) 42:612–620

(geochemical variance), sampling at-a-site (samplingvariance), and analytical precision (analytical variance).The two latter components have been referred to as thetechnical variance, and it has been suggested that theirsum should not be greater than about 20%, with a maxi-mum analytical component of 4% (Ramsey and others1992). The results of a nested ANOVA using the duplicateand subsample data are presented in Fig. 2 for each ele-ment. In each case (with the exception of Co and Sb) thetechnical variance is less than 20%, and in most cases theanalytical variance is well below 4%. We therefore feeljustified in asserting that the analytical data meet our re-quirements for precision.

Identifying local contamination

A statistical summary is provided in Table 1, and cu-mulative frequency graphs of each of six trace metals(As, Cr, Cu, Ni, Pb, Zn) are shown in Fig. 3. Each of thedistributions of the six trace metals is positively skewed(Table 1). The median concentrations (50th percentile)of Cr, Cu, Pb and Zn from the National Soil Inventory(McGrath and Loveland 1992) are shown as filled arrowsin Fig. 3. The median concentrations of each of the tracemetals are, with the exception of Ni, higher in the studyregion than for the rest of England and Wales. Theconcentrations of Cr and Pb in the study region areparticularly high when compared to the median values atthe national scale. No survey data are currently availablefor As at the national scale.There are abrupt increases in trace-metal concentrationsin the upper decile of their distributions (Fig. 4), sug-gesting that a limited number of sites have either beencontaminated by local, point sources or reflect local min-eralisation. Five of the trace metals have sharp inflectionsin the upper decile of their distributions (Ni, Cu, Cr, Znand Pb) whereas As exhibits a more gradual rise (Fig. 4).The 97.5 percentile appears to be below the level at whichconcentrations rise sharply in each of the distributions.Lead (Pb) and Cr concentrations rise sharply between thetwo percentiles, whereas the rapid increase in Cu and Ni isabove the 99th percentile. This suggests that if it werenecessary to establish a uniform upper boundary in eachdistribution of the data set, and in so doing removeanomalously high values, it should be drawn at the97.5 percentile. The number of sites with concentrationsabove the 97.5 percentile is shown in Table 1 for each ofthe six trace metals. These sites were selected using a GISto determine whether their locations might explain theelevated concentrations. Many of them were in the peri-urban areas of medium to large conurbations throughoutthe study region, including Rotherham, Barnsley, Wake-field, Leeds, Bradford and Halifax. This suggests that manyof the sites with concentrations above the 97.5 percentileare the result of local contamination, and we wouldtherefore wish to exclude them in any attempt to define anupper threshold for background values. This findingshows that it is possible to identify and remove highly

Tab

le1

Stat

isti

cal

sum

mar

yo

fth

eto

tal

con

cen

trat

ion

of

maj

or

and

trac

eel

emen

tsin

top

soil

(n=

818)

.A

llva

lues

are

inm

gk

g–1

(un

less

oth

erw

ise

stat

ed)

Al 2

O3

As

Ba

CaO

Co

Cr

Cu

Fe 2

O3

MgO

Mn

OM

oN

iP

2O

5P

bR

bSb

SnSr

TiO

2U

VZ

nZ

r(%

)(%

)(%

)(%

)(%

)(%

)(%

)

Det

ecti

on

lim

it0.

11

10.

051

11

0.01

0.1

0.00

51

10.

051

11

11

0.00

51

11

1M

ean

14.1

21.6

492

0.8

26.2

92.5

39.2

6.6

0.8

0.2

3.9

26.0

0.3

100

76.8

1.1

8.4

69.8

0.9

2.6

103

112.

226

3M

in.

4.2

3.0

126

0.0

3.0

25.0

4.0

0.3

0.1

0.00

11.

02.

00.

121

.08.

00.

51.

021

.00.

40.

323

.010

.099

.0M

ax.

22.1

101

7,90

010

.578

.02,

534

703

18.5

6.3

0.9

43.2

233

1.3

1,86

815

922

.016

124

51.

25.

923

81,

982

4,32

0M

edia

n14

.218

.043

70.

627

.084

.030

.56.

70.

80.

23.

324

.00.

379

.075

.00.

56.

067

.00.

92.

610

310

125

6SD

2.5

11.7

397

0.9

6.8

97.9

36.0

1.7

0.4

0.1

2.5

12.2

0.1

101

21.8

1.7

9.4

19.5

0.2

0.6

23.9

85.3

154

Co

eff.

of

var.

(%)

17.9

53.7

80.7

116

26.1

105

91.8

25.7

48.1

49.1

64.9

46.9

44.7

100

28.5

148

111

28.0

17.7

23.1

23.2

76.0

58.6

Skew

nes

s–

0.1

2.3

12.3

5.2

0.3

20.3

9.0

0.4

6.4

2.1

6.6

7.2

1.9

9.0

0.4

7.1

7.5

2.7

–1.

10.

30.

513

.922

.4K

urt

osi

s0.

47.

419

835

.84.

748

314

43.

868

.916

.682

.410

76.

612

70.

568

.294

.717

.50.

82.

32.

228

659

02.

5%il

e8.

99.

027

30.

112

.050

.012

.03.

20.

30.

01.

710

.40.

135

.036

.40.

53.

043

.00.

51.

460

.444

.415

097

.5%

ile

19.5

55.0

1,01

83.

238

.015

412

610

.11.

50.

39.

346

.60.

727

212

55.

030

.111

21.

13.

915

324

738

8Si

tes

abo

vea

2019

2121

1820

1820

1526

2018

1320

2113

2118

818

2020

20

aV

alu

ein

dic

ates

the

nu

mb

ero

fsa

mp

les

inth

ed

atas

etab

ove

the

97.5

per

cen

tile

Original article


contaminated sites from a regional survey data set, andestablish an upper boundary at a specific point in thedistribution to provide an upper limit to backgroundconcentrations over a single parent material.

Investigating the impact of diffusepollution

To investigate the potential effects of diffuse pollutionthroughout the region, we used human population densityas an indicator of historical, anthropogenic contamination.However, before assessing whether any effect was appar-ent, it was first necessary to ensure there were no signifi-cant differences between trace-metal contents in soils

developed over the two geological formations, the Lower(LCM: n=415) and Middle Coal Measures (MCM: n=413).As demonstrated in the preceding section, soils at a fewsample sites had far greater trace-metal contents than therest of the distribution, which can significantly increasethe mean value. In such cases, the median is a more robustmeasure of centrality in a data set. Hence, we performed aMood’s median test on the data classified into the twoseparate formations (LCM and MCM) to identify thoseelements where the distributions were significantly dif-ferent (Table 2). The Mood’s median test is a non-para-metric test (highly resistant to the outlying values presentin the data set) and assumes that the data from eachpopulation are independent random samples.Of the elements analysed Ni, Pb, Mg, Rb, Sr, Zr, Mo and Bahad significantly different median values in topsoil sam-ples (Table 2). In subsequent statistical analyses, data forthese elements were analysed within each formation toensure that geological influences were not responsible forsignificant differences between sites.Population density information was obtained from the1991 UK Census Small Area Statistics data, part of the UKEDINA database. The data surface models were con-

Fig. 3Log-normal cumulative percentage distributions for six trace metals(n=818). The vertical arrows show the median (50th percentile)concentration of each trace metal in topsoils throughout England andWales (McGrath and Loveland 1992)

Fig. 4The upper decile of Fig. 3. Dashed lines highlight the 97.5 and the 99thpercentiles

Fig. 2Analytical, within- and between-site variance from the analysis ofduplicate samples (and their subsamples) from 84 sites throughoutnorthern England using a nested analysis of variance

Original article


structed from the population-weighted centroids definedfor each 1991 enumeration district (Bracken and Martin1989). The data set forms the basis for population surfaceconstruction, and the counts associated with each centroidare redistributed into the cells of a regular grid (with a cellsize of 200 by 200 m). Using a GIS, the total population ineach of the cells falling within a one-kilometre radius ofeach sampling site was calculated. The cumulative distri-bution of the total population within this radius aroundeach sample site is shown in Fig. 5. Each of the soil sam-ples was classified into one of two groups (high and lowpopulation) according to whether they had total popula-tions above or below the median value (a total populationof about 1,700 people within a 1-km radius; Fig. 5).For those elements which had no significant differencesbetween their median values over the two formations, aMood’s median test was applied to determine whether thedistributions of values in the high- and low-populationclasses were significantly different (Table 3). For thoseelements in which a geological influence was identified inthe preceding section, the same analysis was performed onthe high- and low-population classes within each forma-tion (Table 4). Trace elements with significantly highermedian concentrations in areas of high compared to lowpopulation density (percentage increases in parenthesis inTable 3) across both formations include As (31.3), Cu(38.5), Zn (11.1), Sn (40) and Fe (6.5). From the within-formation analysis, soil samples with significantly highermedian values of trace elements over the LCM in areas ofhigh population (Table 4) were Ni (28.6), Pb (19.9) andMo (35.5). For samples over the MCM, significantly highermedian values of Pb (19.7) and Mo (25.9) were found inareas of high compared to low population (Table 4). Thesedifferences are, with the exception of Fe, much greaterthan the levels of analytical error (Fig. 2). There were nosignificant differences between the median values for theother major and trace elements (Cr, V, Co, U, Sb, Al, P, Ca,Ti, Mg, Rb, Sr, Zr and Ba; see Tables 3 and 4). Althoughcorrelation coefficients were calculated between popula-tion density and trace-metal contents, no consistent rela-tionship could be demonstrated, suggesting that the

Tab

le2

Co

mp

aris

on

of

maj

or-

and

trac

e-el

emen

tco

nce

ntr

atio

ns

into

pso

ils

ove

rth

eL

ow

er(L

CM

)an

dM

idd

leC

oal

Mea

sure

s(M

CM

).A

llva

lues

are

inm

gk

g–1

(un

less

oth

erw

ise

stat

ed)

Al 2

O3

As

Ba

CaO

Co

Cr

Cu

Fe 2

O3

MgO

Mn

OM

oN

iP

2O

5P

bR

bSb

SnSr

TiO

2U

VZ

nZ

r(%

)(%

)(%

)(%

)(%

)(%

)(%

)

LC

M(n

=41

5)M

ean

14.4

22.3

442

0.7

25.6

98.0

37.2

6.7

0.78

0.2

4.1

24.9

0.3

110

72.4

1.0

8.5

65.4

0.9

2.6

103

106

282

Med

ian

14.4

19.0

397

0.6

26.0

85.0

30.0

6.8

0.75

0.2

3.6

23.0

0.3

88.0

70.0

1.0

8.5

63.5

0.9

2.6

101

101

273

SD2.

512

.122

90.

77.

213

326

.21.

70.

360.

13.

014

.40.

111

722

.60.

56.

018

.40.

10.

622

.155

.620

7M

CM

(n=

413)

Mea

n13

.921

.054

40.

926

.886

.941

.26.

60.

870.

23.

627

.10.

390

.481

.31.

28.

474

.30.

82.

610

311

724

3M

edia

n14

.017

.046

40.

627

.083

.531

.06.

60.

800.

23.

026

.00.

367

.079

.00.

56.

072

.50.

92.

610

410

124

2SD

2.6

11.3

511

1.1

6.4

32.2

43.8

1.7

0.41

0.1

1.9

9.1

0.1

80.2

20.1

2.0

10.9

19.7

0.2

0.6

25.7

107

56.5

Dif

fere

nce

bet

wee

nm

edia

nva

lues

(%)

2.5

11.8

17.0

a–

5.0

–3.

71.

8–

3.2

2.7

6.7a

–6.

3–

16.7

a13

.0a

7.7

–23

.9a

12.9

a0

014

.2a

2.8

0.0

–2.

90.

0–

11.3

a

Dif

fere

nce

bet

wee

nm

ean

valu

es(%

)

–3.

3–

6.2

23.1

23.5

4.4

–12

.89.

6–

1.5

12.6

3.4

–13

.58.

5–

6.3

–18

.412

.311

.8–

1.6

13.5

–7.

51.

20.

49.

2–

13.7

aSi

gnifi

can

td

iffe

ren

ceat

the

1%co

nfi

den

cele

vel

(Mo

od

’sm

edia

nte

st)

Fig. 5Cumulative percentage of total population within a 1-km radius ofeach soil sample location (Copyright Crown Copyright OrdnanceSurvey, an EDINA Digimap/JISC supplied service)

Original article


pattern of diffuse contamination is complex, and onlybecomes apparent when data for a large number of sam-ples are combined.There are two lines of evidence to suggest that topsoilshave been contaminated by diffuse sources in areas of highcompared to low population density throughout the studyregion. First, after accounting for potential differencescaused by variations in geological formation (in the case ofMo and Pb), there are consistent, large, statistically sig-nificant differences between trace-element concentrations(As, Cu, Mo, Pb, Sn and Zn) in areas of high and lowpopulation density (Tables 3 and 4). These elements have

close geochemical associations with coal, which was usedwidely throughout the region. Second, the concentration ofmajor and trace elements which are unlikely to have beenaffected by human activities have similar concentrations inareas of high and low population density, which suggeststhat the results do not simply reflect variations in bedrockgeochemistry.We believe there are four pathways through which traceelements may have been dispersed around human settle-ments. The first two relate to the use of coal, which was theprimary energy source in domestic dwellings and industrythroughout the area during the preceding 150 years. The

Table 3Comparison of major- and trace-element concentrations in areas of high and low population density. Elements shown are those with nosignificant difference in median values over Lower (LCM) and Middle Coal Measures (MCM; see Table 2). All values are in mg kg–1 (unlessotherwise stated)

Al2O3 As CaO Co Cr Cu Fe2O3 MnO P2O5 Sb Sn TiO2 U V Zn(%) (%) (%) (%) (%) (%)

Low population density (n=400)Mean 14.2 18.9 0.8 25.5 85.1 31.0 6.3 0.2 0.3 0.9 6.4 0.9 2.6 101 100Median 14.3 16.0 0.6 26.0 82.0 26.0 6.5 0.2 0.3 0.5 5.0 0.9 2.5 101 95.0SD 2.6 10.0 1.0 7.7 38.7 19.9 1.8 0.1 0.1 1.2 5.4 0.2 0.7 23.5 41.1

High population density (n=418)Mean 14.1 24.2 0.8 26.8 99.6 47.0 6.9 0.2 0.3 1.4 10.4 0.9 2.6 105 123Median 14.1 21.0 0.6 27.0 86.0 36.0 6.9 0.2 0.3 0.5 7.0 0.9 2.6 105 105SD 2.4 12.6 0.9 5.9 131 45.1 1.6 0.1 0.1 2.0 11.7 0.2 0.5 24.2 111

Increase in median(low to high pop.) %

–1.4 31.3a 7.1 3.8 4.9 38.5a 6.5a –6.3 7.7 0.0 40.0a –3.2 4.0 4.0 11.1a

Increase in mean(low to high pop.) %

–0.3 28.0 8.3 5.0 17.1 51.7 8.4 –2.9 10.5 56.7 62.8 –2.9 0.8 3.8 23.7

aSignificant difference at the 1% confidence level (Mood’s median test)

Table 4Comparison of major- and trace-metal concentrations in areas of highand low population density within each of the formations: theLower Coal Measures (LCM), and the Middle Coal Measures (MCM).

Elements shown are those with a significant difference in medianvalues over LCM and MCM (see Table 2). All values are in mg kg–1

(unless otherwise stated)

Ba MgO (%) Mo Ni Pb Rb Sr Zr

Lower Coal Measures (LCM)Low population density (n=224)

Mean 430 0.8 3.6 22.1 103 72.3 63.4 273Median 382 0.7 3.1 21.0 80.5 70.0 61.5 274SD 240 0.4 2.1 10.1 93.0 23.8 17.9 63.3

High population density (n=190)Mean 455 0.8 4.8 28.3 119 72.5 67.8 292Median 416 0.8 4.2 27.0 96.5 70.5 65.5 273SD 215 0.2 3.6 17.7 140 21.0 18.7 299

Increase median: low to high pop. (%) 8.8 14.3 35.5a 28.6a 19.9a 0.7 6.5 –0.5Increase mean: low to high pop. (%) 5.9 3.8 35.8 28.4 14.8 0.3 7.1 6.8Middle Coal Measures (MCM)Low population density (n=175)

Mean 554 0.9 3.1 25.8 74.7 83.2 74.2 246Median 462 0.8 2.7 25.0 61.0 82.0 73.0 245SD 620 0.5 1.4 8.3 56.8 21.8 18.5 53.0

High population density (n=228)Mean 536 0.8 4.0 28.0 102 79.8 74.3 241Median 470 0.8 3.4 26.0 73.0 79.0 72.0 241SD 410 0.3 2.1 9.7 92.6 18.7 20.6 59.2

Increase median: low to high pop. (%) 1.7 0.0 25.9a 4.0 19.7b –3.7 –1.4 –1.6Increase mean: low to high pop. (%) –3.3 –8.6 30.7 8.4 37.0 –4.1 0.1 –1.9

aSignificant difference at the 1% confidence level (Mood’s median test)bSignificant difference at the 2% confidence level (Mood’s median test)

Original article


first pathway is the spreading of coal ash, either as a soilamendment or waste disposal, although there is littlepublished information on this practice. The second path-way is the atmospheric deposition of aerosols followingdomestic (and industrial) coal combustion. Arsenic, Moand Pb have been shown to be closely associated withpyrite in UK coals (Spears and Zheng 1999), whereas Cuand Zn also form common pyritic minerals. Numerousstudies have reported regional heavy-metal contaminationof topsoil from aerosol deposition following coal com-bustion (Kapicka and others 1999; Moon and others 2000).Published data on the fate of trace elements following coalcombustion are generally based on the analysis of feed coaland combustion byproducts, through which atmosphericemissions from large power plants can be calculated usinga mass-balance approach. Application of this approach fora power plant in Spain demonstrated that the trace ele-ments Cr, Ni, Cu, Zn, As, Mo, Sn and Pb were highlyenriched in the finest (<10 lm) fly-ash fractions (Queroland others 1995). There is little published data on theemission of trace elements from domestic coal combus-tion, which would have been widespread throughout thestudy region during the last century. However, the corre-spondence between elements enriched in fine ash particlesand those with significantly increased concentrations indensely populated areas suggests that the atmosphericdispersal of aerosols derived from domestic coal com-bustion is a pathway which could account for much of theobserved difference in trace-metal concentrations.The third pathway is the historical spreading of domesticsewage waste, often referred to as ‘night soil’, on land closeto urban settlements, whereby human and animal wasteswere used as fertilisers in the 18th and 19th centuries.Again, little published information is available, althoughthis practice was common. The fourth pathway is theemission of trace elements from road vehicles, includingPb in petrol and, to a lesser extent, Zn from rubber tyres.In each of these four pathways, solid wastes and aerosolsare likely to be deposited at short distances from humansettlements, leading to their accumulation in topsoils.

Discussion

Analysis of the data set indicates that trace-metal contentsof topsoils have been affected by severe local pollution ataround 20 sites (giving rise to very high concentrations;Table 1), and diffuse pollution which has given rise tomedian concentrations of between 11 and 40% higher inareas of high population density. Diffuse, trace-metalcontamination of topsoils has rarely been demonstratedusing data from a regional soil survey in the UK (Anderand others 2000).Most methods of establishing background values in envi-ronmental datasets based on the identification of outlierswould be effective in identifying the sites affected by severepollution. However, they cannot account for more diffuseforms of contamination where the concentration of anelement has been raised in a significant proportion of the

sample population, as shown for As in Fig. 6. This wouldbe necessary to establish natural background concentra-tions according to the ISO definition. Likewise, methodsbased on normalising the data against a conservativevariable (one which is not significantly influenced bydiffuse contamination) are unlikely to be effective becausethe magnitude of natural variability is often greater thanthe subtle changes caused by diffuse pollution. From theanalysis presented in this study it is clear that diffusesources of pollution can increase natural backgroundconcentrations by a significant amount, and that envi-ronmental geochemical data sets need to be investigatedthoroughly to identify such trends.

Acknowledgements This paper is published with the permissionof the Director of the British Geological Survey (NERC). Theauthors would like to thank all the BGS staff and volunteerworkers involved in the collection and analysis of samples in theG-BASE programme, and two anonymous reviewers for theircomments on a preliminary draft of the manuscript.

References

Ander EL, Smith B, Fordyce FM, Rawlins BG (2000) Traceelements. In: Darcy B, Ferrier R, Ellis JB, Dils RM (eds) Diffusepollution: economic and environmental impacts in the UK.Lavenham, Suffolk

Bracken I, Martin D (1989) The generation of spatial populationdistributions from census centroid data. Environ PlanningA:537–543

Downie C (1960) The area around Sheffield. The GeologistsAssociation, Colchester

Fuller RM, Groom GB, Jones AR (1994) The land cover map ofGreat Britain: an automated classification of Landsat ThematicMapper data. Photogrammetric Eng Remote Sensing 60:553–562

Gilbertson DD, Grattan JP, Cressey M, Pyatt FB (1997) An air-pollution history of metallurgical innovation in iron- and steel-making: a geochemical archive of Sheffield. Water Air SoilPollut 100:327–341

ISO (1996) Soil quality – vocabulary, part 1: terms and definitionsrelating to the protection and pollution of the soil. InternationalStandards Organisation ISO 11074(1):22

Fig. 6Cumulative percentage of total arsenic concentrations in topsoils inareas of high and low population density

Original article


Kapicka A, Petrovsky E, Ustjak S, Machackova K (1999) Proxymapping of fly-ash pollution of soils around a coal-burningpower plant: a case study in the Czech Republic. J Geochem Expl66:291–297

Matschullat J, Ottenstein R, Reimann C (2000) Geochemicalbackground – can we calculate it? Environ Geol 39:990–1000

McGrath SP, Loveland PJ (1992) The soil geochemical atlas ofEngland and Wales. Blackie, Glasgow

Moon JW, Moon HS, Woo NC, Hahn JS, Won JS, Song Y, Lin X,Zhao Y (2000) Evaluation of heavy metal contamination andimplication of multiple sources from Hunchun basin, north-eastern China. Environ Geol 39:1039–1052

Querol X, Fernandez-Turiel JL, Lopez-Soler A (1995) Trace ele-ments in coal and their behaviour during combustion in a largepower station. Fuel 74:331–343

Ramsey MH, Thompson M, Hale M (1992) Objective evaluation ofprecision requirements for geochemical analysis using robustanalysis of variance. J Geochem Expl 44:23–36

Reimann C, Filzmoser P (2000) Normal and lognormal datadistribution in geochemistry: death of a myth. Consequences forthe statistical treatment of geochemical and environmental data.Environ Geol 39:1001–1014

Reimann C, Banks D, de Caritat P (2000) Impacts of airbornecontamination on regional soil and water quality: the KolaPeninsula, Russia. Environ Sci Technol 34:2727–2732

Sinclair AJ (1974) Selection of thresholds in geochemical datausing probability graphs. J Geochem Expl 3:129–149

Snedecor GW, Cochran WG (1989) Statistical methods, 8th edn.Iowa State University Press, Iowa

Soil Survey (1984) Soils and their use in northern England. SoilSurvey of England and Wales, Harpenden

Spears DA, Zheng Y (1999) Geochemistry and origin of elementsin some UK coals. Int J Coal Geol 38:161–179

Tidball RR, Ebens RJ (1976) Geochemical baselines for sagebrushand soil, Powder River Basin, Montana, Wyoming. US Geol SurvRep, pp 6–13

Original article


trace-metal pollution of soils in northern england

Documents