Soil-plant interactions and the uptake of Pb at abandoned mining sites in the Rookhope catchment of the N. Pennines, UK--a Pb isotope study.

This paper examines Pb concentrations and sources in soil, grass and heather from the Rookhope catchment in the North Pennines, UK, an area of historical Pb and Zn mining and smelting. Currently, the area has extensive livestock and sports shooting industries. Risk assessment, using the source-pathway-receptor paradigm, requires the quantification of source terms and an understanding of the many factors determining the concentration of Pb in plants. A paired soil and vegetation (grass and heather) geochemical survey was undertaken. Results showed no direct correlation between soil (total or EDTA extractable Pb) and vegetation Pb concentration. However, regression modelling based on the Free-Ion Activity Model (FIAM) suggested that the underlying mechanism determining grass Pb concentration across the catchment was largely through root uptake. Spatial patterns of (206/207)Pb isotopes suggested greater aerosol deposition of Pb on high moorland and prevailing wind facing slopes. This was evident in the isotopic ratios of the heather plants. Pb isotope analysis showed that new growth heather tips typically had (206/207)Pb values of ~1.14, whilst grass shoots typically had values ~1.16 and bulk soil and peat ~1.18. However, the (206/207)Pb ratio in the top few cm of peat was ~1.16 suggesting that grass was accessing Pb from a historical/recent pool of Pb in soil/peat profiles and consisting of both Pennine ore Pb and long-range Pb deposition. Isotope Dilution assays on the peat showed a lability of between 40 and 60%. A simple source apportionment model applied to samples where the isotope ratios was not within the range of the local Pennine Pb, suggested that grass samples contained up to 31% of non-Pennine Pb. This suggests that the historical/recent reservoir of non-Pennine Pb accessed by roots continues to be a persistent contaminant source despite the principal petrol Pb source being phased out over a decade ago.


Introduction
Heavy metal contamination of agricultural land is a widespread problem in mineralised and formerly mined regions. Drainage of mine waters from adits, the weathering of spoil tips and particulate dispersion are potential sources of contamination of soil (Abrahams & Thornton, 1994;Thornton & Webb, 1975). Incidences of mineral related toxicity in newborn lambs and calves have been reported in historical lead mining districts and have been attributed to ingestion of Pb-contaminated forage or soil and severely polluted water supplies (Aslibekian & Moles, 2003;Smith et al. 2010;Thornton & Abrahams, 1983). In addition, many Pb mining areas in the UK are sited in upland moorland regions, where birds such as the Red Grouse (Lagopus lagopus scotica) live and feed primarily on the shoots, seeds and flowers of heather (Calluna sp.). Lead bioaccumulation in grouse can occur in the bones, cause mortality, neurological dysfunction, immunity suppression and a decrease in reproduction in birds (Kendall et al. 1996; Thomas et al. 2009). This paper presents a survey of paired vegetation and soil samples in the Rookhope catchment; a small area in rural, upland northern Britain subject to historical Pb mining and smelting. We use the results to assess the (i) spatial distribution, (ii) magnitude and (iii) source of lead within the catchment. In addition, we discuss the possible uptake mechanisms of Pb into grass and heather growing in this catchment, the major food sources for livestock and grouse.
The principal pathways for the enrichment of plants with Pb include (i) systemic uptake via the roots from the soil solution, (ii) dermal sorption from direct aerial deposition on vegetation and (iii) a combination of the above where aerial deposition becomes part of the soil Pb pool either directly or via the recycling of plant organic matter. Recent work on the uptake of trace metal cations by plants has shown that models based on the Free Ion Activity Model (FIAM) are far more effective at predicting uptake than total or 'extractable' concentrations (Hough et al. 2004;Hough et al. 2005). The significance of aerial deposition on the Pb content of plants was reviewed by Cannon (1976), citing studies as far back as Rains (1971) on the unresolved issue of the relative importance of Pb-rich material deposited on the plant versus direct Pb uptake. Jones and Johnston (1991) used the extensive archive of herbage collected at the Rothamstead Experimental Station, UK, to investigate the influence of changing atmospheric inputs since 1860.
The natural variation in Pb isotopes allows their use as tracers for contamination and source apportionment within environmental systems (e.g. soil, plants). Such isotope studies, coupled with the large calculated estimates of emissions to the atmosphere from both domestic and industrial sources have enabled a tool capable of separating diffuse source from local point sources (Murozumi et al., 1969;BollÖfer and Rosman, 2001;MacKenzie and Pulford, 2002). The source apportionment of Pb isotopes in different environmental reservoirs has recently been reviewed by Komarek et al. (2008) and included the use of selective soil extractions. For example, Bacon et al. (2006) used sequential extractions to examine Pb mobilisation in organic rich soils from an upland catchment in Scotland. More recently, Shepherd et al. (2009) described a large scale regional study of Pb isotopes in stream sediments in the Wear catchment, NE England, encompassing the current study area, demonstrating a dominance of local ore Pb signatures in the North Pennine Orefield, but with a significant contribution of deposited industrial or petrol Pb downstream.
In this work we use a combination of Pb isotope analysis to assess the sources of Pb in soil-plant couples and Pb isotopic dilution methods to assess the labile pools of Pb in two peat cores. The results give insights into the relationships between the impact of past local Pb smelting and mining activity and diffuse aerial pollution in relation to plant Pb content and uptake mechanisms. This is key information required for conceptual models of risk assessment for animals grazing in mining contaminated sites.

Study Area and Historical Context
Rookhope Burn, is a tributary of the River Wear, and is one of several small catchments in an area of abandoned mining associated with the North Pennine Pb-Zn ore-field, U.K. The Rookhope Burn rises to the east of Allenheads and flows through an alluvial channel in an easterly direction towards Rookhope village, before turning south to join the River Wear at Eastgate. A defining feature of the northern catchment boundary is the heath and bog clad ridge of Black Hags and Redburn Common, formed in Grindstone Sill sandstones of the Stainmore Formation (Yoredale Group). This moorland is used for seasonal grouse shooting. The southern boundary of the catchment is defined by a ridge of the same sandstone, covered by shrub heath and used for rough grazing. The steeper sides to the valley have better drainage and are a mixture of rough and improved grassland. These are mainly used for sheep grazing and occur on an inter-bedded sequence of thin sandstones and mudrocks of the Stainmore Formation. The upper valley bottom offers rough grazing.
As the stream gradient decreases below Rookhope village into the underlying Alston Formation (a mixed sequence of limestones, sandstones and shales), a combination of improved grassland, with intermittent broad leaved and mixed woodland, is found lining the river banks down to Eastgate. This area also coincides with the development of superficial Devensian till deposits. Despite the clear evidence of Roman occupation in the Northern Pennines (Dunham, 1990) there is little evidence to suggest that there was Roman Pb mining and smelting in the study area. By the Middle Ages, mining and smelting of Pb was certainly in progress (Drury, 1987;King, 1982;Fairburn, 1996). Perhaps the peak of activity in the Rookhope Burn catchment occurred in the mid to late 19 th century. Figure 1 illustrates the location of mineral veins and gives an indication of the scale of workings in the Rookhope catchment. Several examples of the mining legacy remain (spoil heaps, processing plants, mine shafts). One of the most prominent is the line of the Rookhope smelter chimney that runs up the fell side to Redburn Common, a distance of almost 2.5 km. Smith and Blackburn (2001) state that this smelt mill was built in 1737 and replaced an earlier smelt mill a little way up the valley at Scar Sike. The later smelt mill operated until 1919 (Houston, 1964), a period of Pb production and processing of at least 180 years. A paper by Louis (1917) indicates that three mines were operational in Rookhope; these were Wolfcleugh, Grove Rake, and Boltsburn. Mining continued at Boltsburn until 1940. In addition, fluorite mining continued at Grove Rake at the head of the Rookhope Burn up to 1999.
There are three major soil associations within the Rookhope catchment as designated by the Soil Survey of England and Wales 1:250000 maps. These are very acidic blanket peat (1011b; Winter Hill), soils formed from superficial deposits of Palaeozoic sandstone, mudstone and shale which have an organic rich / peaty surface horizon and are very acidic when not limed (721c, Wilcocks 1), and soils found in the lowland pastures being formed from superficial deposits of Palaeozoic sandstone and shale (713g, Brickfield 1).
The Wilcocks 1 and Brickfield 1 soils are slowly permeable and are prone to waterlogging.

Soil Survey
Soil samples (0-15 cm) were collected on a series of transects mostly within the catchment at a nominal spacing of 500 m. There were three sample points just outside the catchment (shown in Fig 8a & b). Each sample consisted of a bulked sample of five sub-samples collected on the corners and at the centre of a 5m square grid. This catchment wide sampling strategy was undertaken to be compatible with the Geochemical Baseline Survey (G-BASE) of the UK (Johnson et al. 2005). The soils were returned to the laboratory in kraft bags, air dried and sieved to < 2 mm. Approximately 30 g of the < 2 mm fraction was sub-sampled, ground in an agate ball-mill to produce a fine homogeneous powder used for total metal and Pb isotope analysis.

Collection of peat cores
To complement the original soil and plant survey, the characteristics of Pb lability and solubility were examined in peat cores from two areas of the Rookburn catchment. Peat cores provide an undisturbed historical record of Pb deposition in the area and enable us to use analytical Pb isotope techniques to derive the source and historical mobility of lead.
In addition, we used isotopic dilution methods to assess the mechanistic processes involved with the lability and solubility of Pb. Peat cores (0-50 cm) (n=3) were extracted from two locations using a Russian auger. Core 1 was from a lowland peat site (Grid Ref: BNG 387761 544196) whilst Core 2 was taken on the high ridge (Grid Ref: BNG 390157 544786), close to the smelter chimney site. The first core from each site was cut in two lengthwise and then into 12.5 cm sections. Half of each section had pore water extracted using the centrifugation method of Kinniburgh et al. (1983). The corresponding half of each core was used to determine characteristics including moisture content, pH, LOI, Pb isotopic composition and isotopically exchangeable pools of Pb. The second core was cut into sections to determine high-resolution information (1 cm) on Pb concentrations and isotope ratios. The 1 cm sections were freeze-dried to facilitate size reduction, pulverised using a hand agate mortar and pestle to > 95 % passing a 100 µm mesh and stored in plastic tubes until required for further analysis.

Total metal concentrations in soil and peat samples
Sample digestion for total metal concentrations in soils and peats were undertaken by accurately weighing 0.25 g of soil into a Savillex™ vial to which 4 ml of concentrated HNO 3 acid was added before heating at 80C overnight until dryness, the purpose being the digestion of reactive organic phases. Digests were cooled prior to the addition of 2.5 ml HF, 2 ml HNO 3 and 1 ml HClO 4 concentrated and analytical grade acids, with a subsequent stepped heating program to 160C overnight, the purpose being the digestion of silicate and oxide phases. The dry residue was re-constituted after warming with 1.25 ml MQ water, 1.25 ml HNO 3 and 2.5 ml H 2 O 2 , to 25 ml of 5% v/v HNO 3 and stored in HDPE bottles. Reference materials (NIST SRM2710, SRM2711, GSS-6, BGS102 and BCR-2), duplicate samples and blanks were all prepared in a similar manner to check accuracy of the analytical and digestion method. The concentrations of Pb, was determined on the total digest of the soil samples using an ARL3580 ICP-AES instrument. Calibration was undertaken, using synthetic chemical multi-element standards produced in-house from a variety of commercial sources. Similarly, multi-element synthetic quality control standards for calibration verification were produced in-house, but, where possible, from different sources to the calibration standards.
For the peat cores, Pb were determined using an Agilent 7500 quadrupole ICP-MS instrument. The instrument was calibrated using a series of synthetic chemical solutions diluted from multi-element stock solutions (SPEX Certprep™), the calibration being validated using synthetic chemical standards from a separate source. Similarly, multielement synthetic quality control standards for calibration verification were produced inhouse, but, where possible, from different sources to the calibration standards. The calibration and quality control standards were inserted every 20 samples to check possible drift over the run. Data were corrected for blank contributions and possible interferences by running a number of blanks and synthetic chemical solutions of Ba, Ce, Gd, Nd, and Sm. The digest solutions were diluted to within the caibration range prior to analysis.

Soil Extractions
A sub-set of 87 of the total of 165 soil samples were extracted using 0.05 M NH 4 -EDTA, the BCR-EU standard method (Horckmans et al. (2007) to provide an assessment of bioavailable Pb. These samples matched those sites from which vegetation samples were collected and analysed. The EDTA extraction was undertaken by shaking 1 g of milled sample with 10 ml 0.05M EDTA in a plastic centrifuge tube for 17 hours. After centrifugation, the supernatant was acidified to 1% v/v HNO 3 and stored in HDPE bottles. NH 4 -EDTA was used as it is the most amenable form for direct Pb isotope analysis by ICP-MS. The concentration of Pb was determined using a Thermo Elemental PQ ExCell™ quadrupole ICP-MS instrument. The instrument was calibrated using a series of synthetic chemical solutions diluted from multi-element stock solutions (SPEX Certprep™), the calibration being validated using synthetic chemical standards from a separate source.

Vegetation
All vegetation was sampled using a clean pair of stainless steel shears from an area within 10 m of the associated soil sample and registered grid reference. Each sample was transferred to a plastic bag. Grass was cut down to within 2 cm of the soil surface; no specific species were sampled as different species were prevalent at different sites.
Heather was sampled by removing 2-4 cm of fresh tip material. On return to the laboratory all vegetation samples were air dried at 40C for at least 24 hours before milling to a fine powder using a large volume, low alloy steel, Mixermill and subsequently stored in plastic containers. A 0.5 g aliquot of the powdered vegetation was weighed into a Savillex™ vial and 10 ml of concentrated HNO 3 acid added before heating at 90C for 14 hours. After cooling a further 1 ml of H 2 O 2 was added and the solution evaporated to dryness. The dried digest being reconstituted in 25ml of 10% v/v HNO 3 before storage in HDPE bottles.
For the purposes of quality control duplicate samples, reference materials (GBW07602bush branches and leaves; GBW07605tea, from IGGE, China; and LSG-1grass, an in-house material) and blanks were prepared with each batch of samples for analysis. The elemental concentrations of the digested vegetation samples and quality control materials were determined using a Thermo Elemental PQ ExCell™ quadrupole ICP-MS instrument for 16 elements. The instrument was calibrated using a series of synthetic chemical solutions diluted from multi-element stock solutions (SPEX Certprep™), the calibration being validated using synthetic chemical standards from a separate source. The digest solutions were diluted by a further factor of 20 prior to analysis, for some samples a further dilution being necessary for all elements to fall within the defined calibration range.

Lead isotope ratio determinations in soil, peat and vegetation
The Pb isotope ratios of vegetation digests, total soil and peat digests and EDTA soil partial extractions were determined using an Agilent 7500cx quadrupole ICP-MS. The instrument was prepared for isotope ratio determinations before each analytical session, by plateauing of the detector voltage, cross-calibration of the pulse counting-analogue modes and updating of the dead-time correction factors. Isotope ratio determinations were recorded on the basis of 10 replicate integrations of 30s; each integration being 1000 peak jumps (1 point per peak) across the isotopes 203,205 Tl and 206,207,208 Pb. With reference to the total Pb concentrations in the solutions, prior to analysis, all test portions were diluted using 1% HNO 3 /0.5% HCl to give a count rate as close to, but below, 1Mcps on the major lead isotope 208 Pb to produce best counting statistics within the linear range of the pulse counting detector. In addition, the test solutions were doped with Tl sufficient to measure close to 1Mcps on 205 Tl. The issue of mass bias mass in the isotope ratios was addressed by determining the measured isotope ratio for NIST SRM918 regularly throughout the analytical session and producing interpolated correction factors of sample isotope ratios with reference to the accepted isotope ratios for SRM981 (Thirwell, 2002). Stability of the instrument was such that the use of the 205/203 Tl ratio to determine and correct for mass bias was found to detract from the quality of analysis in terms of both accuracy and precision and therefore not used. Quality control for Pb isotope ratios was performed using an in-house solution produced from a naturally occurring UK galena -"Glendenning". This has been run repeatedly over several years. Long term (n=37 over 18 months) precision being 0.07% for both 206/207 Pb and 208/207 Pb ratios. Individual sample errors in Pb isotope ratios were calculated by propagating the precision of individual sample measurements and the uncertainty in the measurements of the SRM981 sample used to correct for mass bias. Typical sample uncertainty is shown on the relevant graphs.

LOI and soil pH
Both Loss on ignition (LoI) and pH analysis were undertaken on < 2 mm fraction of the samples. LoI analysis was performed as a proxy for total organic content. The samples were initially dried to 105C for four hours to remove absorbed water, before weighing into glass beakers. The beakers containing the samples were "ignited" at 450C for a minimum of 4 hours, before cooling in a desiccator and reweighing. Soil pH was measured by suspending soil in 0.01M CaCl 2 solution in the ratio 1:2.5.

Lead Isotopic Dilution assays for Peat samples
Isotopic dilution assays were carried out on peat core samples using 204 Pb as the stable metal isotope. The method was similar to that used by Marzouk et al. (2011)  Suspensions were re-equilibrated for 3 days before filtering the supernatant (0.2µm cellulose acetate filters). To discriminate between true labilie Pb species and non-labile colloidal-Pb in solution, a resin stage was undertaken whereby 10ml of solution was shaken with 0.1g Chelex 100 grade resin (Bio-Rad Laboratories, UK) for an hour. The chelex-100 resin was converted from Na to Ca form to reduce the pH of the resin from 10.05 to 7.90. The resin was then eluted using 5% HNO 3 . Solutions were analysed by ICP-MS (Thermo-Fisher XSeries II ) to determine isotopic abundances of 204 Pb, 206 Pb, 207 Pb and 208 Pb. where M PbSoil is the average atomic mass of Pb, C is gravimetric Pb concentration (mg L -1 ), V is the volume of added spike (L), W is the weight of soil (kg), IA denotes isotopic abundance of a particular isotope in the spike or soil and RSS is the ratio of isotopic abundances for 204 Pb to 208 Pb calculated for the spiked soil supernatant.

Grass Pb uptake modelling
The model used to predict plant uptake was based on the Free Ion Activity Model, after Hough et al. (2004). It is based on a pH-dependent Freundlich relation that can be used to describe metal solubility in soils (e.g. Tye et al. 2004). The equation predicts free metal ion activity for Pb in the soil pore water (M 2+ ) from total soil metal content which is assumed to Page 14 of 55 be adsorbed on humus, [M C ] (mg of a specific metal per kg of soil organic carbon), and soil pH: where k 1 and k 2 are empiric, metal-specific constants and n F is the power term from the

Soil geochemistry
Key soil geochemical parameters for each sample are summarised in Table 1 for the grass   dataset and Table 2 for the heather dataset. The majority of soils have high organic matter content and low soil pH indicative of either peat soils (Winter Hill) or those with organic rich surface horizons (Wilcocks 1 & Brickfield 1 associations). Land-use is predominantly rough grazing or moorland peat and heather. An important feature of relatively undisturbed soils with high concentrations of organic matter in the surface horizons is that aerially deposited contamination from mining, smelting and long range sources will be concentrated in reservoirs close to the surface as has often been demonstrated for peat deposits (Novak et al. 2008;Weiss et al. 2002).

Total and EDTA extractable Pb concentrations in soil
The total Pb content of the soils varied from background concentrations to highly contaminated (13-27600 mg/kg) and are comparable to those reported in Kabata-Pendias,  Figure 2 shows the proportion of extractable Pb (%) compared to the total Pb concentration. At the lowest and highest total Pb concentrations the EDTA extractable proportion is typically lower. Solid phase Pb speciation is likely to be the underlying geochemical control: at the lowest concentrations (10's mg kg -1 total Pb) a higher proportion of Pb is likely to occur within the silicate lattice of the soil and is unavailable for extraction. In the soils with high concentration (1000's mg kg -1 total Pb), Pb is likely to be present mainly locked up within the sulphide ores, thus being insoluble to the EDTA extraction.

Concentrations of Pb in Vegetation
The range of Pb concentration in grass from the Rookhope catchment spans three orders of magnitude (0.8 -48 mg kg -1 ) ( Table 1) with an inter-quartile range of 2.7 -16.7 mg kg -1 .
The highest concentrations of Pb in grass were often found close to spoil heaps or within the vicinity of smelter operations. For heather, Pb concentrations ranged from 1.17 to 7.64 mg kg -1 with an inter-quartile range of 1.6-2.9 mg kg -1 . When compared to the literature, Notten et al. (2008) found a Pb concentration range of 0.41-1.27 mg kg -1 in mixed leaf material from contaminated floodplain soil in Biesbosch, Germany, whilst Smith et al. (2009) gave median values of between 3 and 60 mg kg -1 for washed herbage from floodplain soils in Wales.

Lead uptake into grass
The relationship between grass and heather with total lead and EDTA extractable lead concentrations were investigated. Weak positive correlations between vegetation and soil Pb for grass were found but these only had probability values of 0.007 and 0.110 for EDTA extraction and the total Pb respectively. Similarly, low probability values for heather 0.100 (EDTA) and 0.069 (total) signify no or very weak correlations. The poor correlations found between total and EDTA soil Pb and the concentration in grass and heather is unsurprising. It is increasingly recognized that improved predictions of trace metal uptake into plants can be found using models such as the Free Ion Activity Model (FIAM) (Hough et al. 2005). These models assume the uptake of metals into plants is associated with the intensity factor of free metal ion activity (M 2+ ). It has been observed that even washing with EDTA can still leave traces of aerosol Pb (Rains, 1971;Notten et al. 2008)  However, in low pH soils such as peat, it is also likely that H + will act as the major competing ion, as demonstrated by Hough et al. (2005) for Zn and Cd uptake into grass (Lolium perenne).
The results suggest that systemic uptake of (Pb 2+ ) could account for much of the grass Pb concentration across the catchment. Sources of soil Pb that the grass roots may access include (i) natural geogenic Pb, (ii) recently deposited Pb, particularly from rain that enters the soil or a (iii) pool of recent and historically deposited Pb consisting of local and long range aerially transported Pb (e.g. Pb smelting and petrol combustion). These different sources of historical/recent Pb may be continually recycled from the soil to plants through senescence and management activities such as heather burning. The significance of these different Pb sources is investigated using Pb isotopes as tracers. in petrol was primarily derived from the "Broken Hill Mine" in Australia (Sugden et al. 1993). However, other sources of aerosol Pb have been recorded in the UK with a range of 1.06 -1.13 (Charlesworth et al, 2006;Noble et al, 2008 ). As 'Broken Hill Type' Pb was the major source of non-Pennine Pb in the UK, we therefore refer to all sources of non-Pennine Pb as 'Broken Hill Type' or 'BHT' Pb in the following discussion for simplicity. In addition, because deposition of non-Pennine Pb has occurred over a long period of time, the 206/207 Pb isotope ratio found in soils and peats is likely to be an integrated ratio of recent and historical deposition and as a result of ploughing, grazing and recycling through managed heather burning.

Lead Isotope Ratio Geochemistry in soils and plants
Lead isotope ratios were determined in some of the total and EDTA extractions from soils and the vegetation (Tables 1 and 2). These were plotted together with potential endmember compositions from previously published data. A bi-variate plot ( Figure 5)  The vegetation array from the current study fall along a mixing line between the Weardale galena samples of Rohl (1996) and values obtained for end member aerosols from Liverpool and the Isle of Man sampled in 1997 (Charlesworth et al. 2006)  Weardale galena value (Rohl, 1996;Shepherd et al. 2009) and the mean petrol Pb value (Sugden et al. 1993;Vinogradoff et al. (2005) The heather shows a 'BHT' Pb contribution of between 10 and 35%, whilst the grass has a wide range of contributions between 7 and 31%.
Lead isotopic homogeneity of the different sample matrices at each site was investigated by comparing the 206/207 Pb ratio of each grass, heather and soil EDTA partial extraction with the 206/207 Pb ratio of the soil total Pb from the same site ( Figure 6). The soil EDTA extractable Pb was essentially isotopically similar to the total soil Pb. This is perhaps not surprising as EDTA is a powerful extractant and will dissolve a high proportion of the total Pb ( Figure 2). Some grass samples (n=8) give a similar isotopic ratio as their paired soil sample whilst a second group of grass samples (n=6)  To investigate the effect of different sources and element concentrations on isotope ratios, especially for samples influenced by more than one isotopic source, it is conventional to plot isotope ratio against the reciprocal of concentration. This ensures that when two isotopic end-members of fixed composition are mixed in different proportions, the samples plot along a linear array (Faure, 1977). The samples for all matrices are shown in this way in Figure 7. For clarity the vegetation and soils concentrations are plotted on different scales. Both total and EDTA extractable Pb, at concentrations lower than 200 mg kg -1 , form a single linear array with a uniform 206/207 Pb ratio of 1.178; whilst at higher concentrations the spread of isotope ratios increases to encompass 1.174-1.182, diagnostic of the addition of two or more isotopically distinct Pb sources of both lower and higher ratios. All but one of the soil Pb isotopic ratios are similar to local ore Pb in the range 1.177-1.198 as determined by Rohl (1996). The heathers form a weak array but with a greater slope than the grass. The lower "background" 206/207 Pb ratio is approximately

The spatial distribution of Pb isotopes in vegetation
The spatial distribution of Pb isotope ratios in the vegetation within the Rookhope catchment was examined. Isotope ratios were split into 4 categories (Figure 8a) than the grass plants (~1.16). It is likely that heather with it's taller architecture, and longer growth period is a better scavenger of current aerosol Pb, depending on spatial position.

Pb solubility and source apportionment using historical archives of Pb in peat cores
Many of the grass and heather plant samples were growing in peat or uncultivated (rough pasture) soils where high organic matter surface horizons have developed. These horizons typically provide historical/recent archives of pollution where atmospherically deposited metals accumulate. We used peat cores samples with Pb isotope geochemistry and isotopic dilution techniques to undertake a more detailed analysis of Pb source and bioavailability.

Total Pb concentration and source apportionment in peat cores
In the two peat cores, Pb concentration varies over two orders of magnitude ( Figure 9).
Lead is present at higher concentrations at the surface, substantially increases between 3-10 cm depth, peaking at 7 cm depth, and decreases with depth thereafter. Whilst the distribution profile is similar for both cores, the absolute concentrations of lead at 3-10 cm depth differ and reflect the proximity of Core 2 to the smelter chimney, where 2500 mg kg -1 Pb was measured at 7 cm depth. Although the enrichment of Pb at 3-10cm is lower in Core 1 (lowland core), the peak is comparatively broader, suggesting that Pb is more evenly distributed at those levels. Careful inspection of log Pb concentrations (Fig. 9, Inset) shows that Pb concentrations in Core 1 steadily decline and reach background concentrations at about 40cm, whilst in core 2, the Pb concentrations declines more rapidly and reaches background concentration at 30cm. However, the results, suggest that the vast majority of 'BHT' Pb that has accumulated has been relatively immobile and has accumulated in the top 10 cm, with possible slight enrichment of Pb down the core.
The isotope analyses of the two peat cores (Figure 9) suggests that the Pb is dominated by the Weardale galena ( 206 Pb/ 207 Pb `= 1.178 -1.206) isotope signature (Shepherd et al., 2009;Rohl, 1996). The distribution of isotope ratios with depth is also similar for both cores. The lowest 206 Pb/ 207 Pb ratios (1.161-1.165) were measured at the surface and sharply increased within the first 3 cm. The lower Pb ratios can be attributed to 'BHT' Pb inputs essentially concentrated in the upper 3 cm ( Figure 9). As described previously, the location of Core 2 increases its exposure and favours the capture of atmospheric 'BHT' lead, leading to lower 206 Pb/ 207 Pb ratios at (sub)surface levels in comparison with Core 1.
These observations suggest that Pb from sources other than the Pennine Ore remain mostly concentrated in the upper peat profile (0-3 cm). At greater depths (4-50 cm), 206 Pb/ 207 Pb ratios showed minor temporal variations within a narrow range (1.175-1.182) in good agreement with the Pennine ore signature (Rohl, 1996). Core 2 showed slightly higher 206 Pb/ 207 Pb ratios compared to Core 1 throughout the profile. Interestingly, the 206/207 Pb signiture does not decrease at the surface as has been found in recent studies of Pb in ombrotrophic peat bogs, where a increase in 206/207 Pb ratios has been found at the surface, reflecting the withdrawal of Pb used in petrol . Possible reasons for this are (i) that the surface of the peat has been bioturbed by grazing and (ii) that managed heather burning is a way in which to recycle plant Pb back to the surface of the peat (Odigie & Flegal, 2011).
If isotope and total Pb data are combined, it is apparent that the aforementioned enrichment in Pb at 3-10 cm depth in both cores probably traces the peak of Pb mining activities in the area. As mining activities ceased a few years ago, Pb releases into the environment decreased and other sources of Pb were less well masked, with the result of clear isotopic shifts. The fact that the peat array falls along a mixing line between the North Pennine ore signatures (Rohl, 1996) and the 'BHT' Pb supports them as possible endmembers of sources ( Figure 10). The North Pennine ore component is dominant as most samples tend to cluster towards this end, but the 'BHT' Pb contribution appears to be relevant in samples from the top of the core where the 206/207 Pb isotopic ratios are lower than the range for the Weardale galena (1.178 -1.206). Based on equation (6), the contribution of 'BHT' Pb was estimated at 16% of the total Pb at the surface of Core 1 whereas in Core 2 'BHT' Pb accounted for 19% due to its landscape position and higher exposure to atmospheric deposition.

Lability of lead and source apportionment in labile pools in peat cores
Isotopic dilution (ID) assays allow an improved assessment of the reactive pool of Pb in soils compared to the 0.05M EDTA extractant. The reactive pool of Pb is in equilibrium with the soil pore water and is therefore available for uptake by the grass via the FIAM.
Lability of Pb in the peat cores using ID was between 40-60 % (Table 3) (Table 1). Using Eqn. 2, the isotopic ratio of 206/207 Pb in the labile and non-labile pools do not show distinctive isotopic signatures attributable to a single Pb source, (e.g. Pennine Pb ore or 'BHT' Pb) and are similar ( Figure 12). This suggests that the Pb from diverse sources is almost completely mixed throughout the labile and non-labile pools in the cores despite probable differing times of deposition. Figure 12 shows that for Core 1 (lowland) there are not large differences in the 206/207 Pb isotope ratio between the total and labile pools. However in Core 2, close to the smelter chimney, there appeared to be a shift in isotope values in the top 25cm, and only small differences between 25-50cm, suggesting that the labile pool contains a greater proportion of 'BHT' Pb close to the surface.

Pb concentrations, isotope ratios and source apportionment in peat pore waters
The pore water concentrations of Pb were ~ 400 µg L -1 in the top 12.5 cm and decreased with depth (6-95 µg L -1 ) ( Table 3) However, this was found not to be the case generally, suggesting that there may be recently deposited Pb in the pore waters that has not yet reached isotopic equilibrium with the labile pool. One such source may be Pb deposited in rainwater. Recent studies by Bacon et al. (2006) and Dawson et al. (2010) have suggested that soil pore waters and spatially associated stream waters can show different Pb isotope signatures compared to the soil; indicative of a greater mobility for recently deposited anthropogenic lead of aerosol origin. However, whilst our results may suggest that the solubility of 'BHT' Pb is slightly greater, a more detailed investigation is required to demonstrate this more conclusively.

Discussion
The aims of this paper were to assess the spatial distribution, source and extent of Pb Spatial position within the catchment was found to influence the size of the total soil Pb pool and isotopic composition (e.g. proximity to mining operations or the position in landscape with respect to altitude and prevailing wind). However, our results suggest that across the different contaminated soils of the catchment, the underlying mechanism determining grass Pb concentration was through plant uptake. This hypothesis is strongly supported by the reasonable parameterisation of the FIAM model. This result differs slightly from the interpretation of results presented by other authors where it is suggested that plant concentrations of Pb are dominated by deposition. For example Bacon et al. (2005) suggested that only ~20% of Pb came from the soil in a study where they spiked the soil with 207 Pb. Watmough & Hutchinson (2004) also found a high proportion of pollution Pb in woodland vegetation. However, in both these studies the results did not differentiate between direct deposition on leaves or the deposition of Pb to soil and the subsequent uptake through the root system. It has been considered that Pb is rapidly fixed in soils but the recent use of isotope dilution techniques have shown that the reactive pools of Pb in soils are surprisingly large. In this study the labile pools of Pb in peat were found to be between 40-60 % whilst in another recent survey of soils from the Rookhope catchment the lability of Pb was found to be between 20 -80 % of total metal concentration (Marzouk, 2012). In another study, Degryse et al. (2007) suggested that the lability of Pb in soil was greater than what was intuitively expected, as it has always been assumed that Pb sorbed strongly to poorly crystalline FeO and organic matter. In addition, they suggested that ageing, sorption and precipitation reactions may also be limited. A further indication of the extent of Pb uptake through plant roots in contaminated soils, such as the ones in our study, can be found in experiments assessing remediation strategies.
Both lime or phosphate applications have been found to decrease dramatically the uptake of Pb by grass in contmainated soils (Gray et al. 2006;Friesl-Hanl et al. 2009) confirming that uptake of Pb by roots takes place in contaminated soils.  (Tye et al. 2003). Results in Table 1 where paired soil and grass isotopes are presented show that the dataset breaks down into two groups. One group (Sample nos: 25, 28, 509, 518, 534, 539, 1002(Sample nos: 25, 28, 509, 518, 534, 539, , 1041  are predominatly peat soils. We believe that the discrepancy between soil and plant isotopes found in the second group of soils is largely a result of the low resolution (0-15 cm) sampling used in this study. However, we stated previously (Section 3.7.1) that the Pb enrichment in the peat cores represents the integrated non-Pennine 'BHT' Pb deposition. is likely to be assimulated by the plants as it becomes part of the labile pool and before progressive sorption and fixation processes occur over time to some degree. The suggestion that plant uptake is the underlying mechanism for grass Pb concentration across the catchment is further demonstrated by the difference in paired heather and grass 206/207 Pb ratios as described in Section 3.5. It appears likely, that that the grass is obtaining Pb from a combination of Pennine and 'BHT' sources, but predominantly from the soil.
The high resolution analysis (1cm increments) of peat cores has demonstrated that much of the deposited 'BHT' Pb is relatively immobile and remains in the top few cm in peat or undisturbed organic rich soil profiles. The use of isotope dilution and pore water analysis on the peat cores has developed our understanding of the behaviour of Pb in organic rich soils/peat. We found relatively high Pb lability as well as mixing of the different Pb sources throughout the labile and non-labile pools. Thus, it can be postulated that grass roots are also likely to extract Pb from environments of high 'BHT' Pb bioavailability in the top few cm. For example, in Core 2, these environments (1cm sections) could have up to 2500 mg kg -1 Pb with 40-60% lability. This is particularly relevant when in waterlogged conditions, such as peat, much of the root growth is likely to be shallow so as to stay in more oxygenated environments. These factors may contribute to the high levels of 'BHT' Pb levels found in the grass. Whilst, this discussion has focused on the uptake of Pb to grass, it is likely that some of the uptake of Pb to heather may be through the root system, although isotope ratios ( 206/207 Pb = 1.14 -1.15) suggest a greater direct deposition on the plant surfaces. However, the soil pH data range was too narrow to model effectively in this study to assess whether it would follow the FIAM.
We estimated the proportion of non-Pennine ore lead in some grass and heather samples.
It was estimated from the simple mixing model (eqn. 6) that up to 31% in grass and 37% in heather was present. Sources of 'BHT' Pb will include both historical and current deposition. A further recycling pathway will include the redistribution of Pb from the peridical burning as the heather is managed. It is likely that the contribution from aerosol deposited Pb is not as great as it was in recent decades. This can be seen in data from the UK National Air Quality Archive, (2009). Selected data are shown in  (Hough et al. 2005).

Conclusions
A combined vegetation and soil survey was undertaken within the Rookhope Catchment to assess the extent of Pb contamination in soils, grass and heather as an aid to understanding potential bio-accumulation in grouse, sheep and cattle. We found spatial differences in Pb concentration and isotope composition related to the historical proximity to mining and smelting activity as well as landscape position, controlled largely by altitude and prevailing wind direction. The application of a FIAM for lead uptake in grass across the catchment using the data from the simple soil geochemical survey of total metal, pH and LOI concentrations produced reasonable predictions of lead concentrations, whilst there always being potential for some aerosol Pb to be retained on the vegetation. Despite recent steep declines in 'BHT' aerosol Pb, the grass and heather Pb isotope data from elevated and prevailing wind locations showed that it can still contain considerable quantities of 'BHT' Pb. If grass conforms largely to the FIAM for plant Pb uptake as suggested, it is apparent that the grass is accessing a historical/recent soil pool of lead that has built up over many decades, where 'BHT' Pb still resides in the top few cm. In the acidic peats, the use of isotopic analysis has demonstrated that there is almost complete mixing of Pb between the labile and non-labile pools of Pb and generally high lability, allowing the grass access to the reservoirs of 'Broken Hill type' Pb and increasing its content in the vegetation. These results demonstrate that although petrol Pb was phased out in 2000, it resides in soil reservoirs and is highly persistent in the environment. The results of the current study also has important ramifications for sampling strategies where plant uptake processes need to be considered in addition to soil Pb data to understand the potential bio-accumulation of Pb in cattle, sheep and grouse. Results demonstrate that sampling strategies need to be developed for particular soil types and scenarios. For undisturbed, organic rich soils there is a need for high resolution sampling strategies in the top few cm of peats and organic rich horizons to develop improved understanding of Pb lability, source and the subsequent interactions with pore waters and plant uptake.