Applying lead (Pb) isotopes to explore mobility in humans and animals

Lead (Pb) isotopes provide a complementary method to other provenance tools for tracking the origin and movement of humans and animals. The method is founded in the geographic distribution of Pb isotope ratios. However, unlike the Sr isotope method that is closely linked to the lithology of underlying rocks, Pb more closely reflects the tectonic regimes. This makes it particularly pertinent to use in Britain as there is major tectonic boundary (the Iapetus Suture) that runs between Berwick-upon-Tweed and the Solway Firth providing a compositional boundary in Pb isotope domains that approximates to the geographic areas of Scotland versus England and Wales. Modern pollution makes it difficult to use modern floral or faunal samples to characterize biosphere variation, and so we use geological datasets to define isoscape variation and present the first Pb isotope map of Britain. We have validated the use of these data form biosphere studies using well provenanced samples. Reference fields of diagnostic compositions, are created in μ-T space and these have been used in a test case to assess the geographic origins of Neolithic animals in Great Britain.


Aims
Pb isotope analysis of tooth enamel is becoming an increasingly important method of analysis for constraining geographic origins in humans and animals [1][2][3][4][5][6] as it provides another dimension in the assessment of origins of populations in Britain by combining multiple isotope proxies. The primary aims of this paper are to 1) demonstrate the potential of lead (Pb) isotope analysis to provide a geographic discriminant between the two major tectonic zones of Great Britain, 2) to show that geological Pb signatures are transferred to the fauna exploiting these in the past and 3) to use the method to refine the geographic constraint of the origins of pigs found at several Neolithic henges in southern England. parts of Great Britain. The Pb isotope composition of rocks and minerals tends to be dominated by major geological tectonic events such as mountain building, which is accompanied by metamorphism and the intrusion of granites, the heat from which drives the re-mobilisation of Pb to create ore deposits. Thus, as described by Blicher-Toft et al. 2016 [14], Pb isotope compositions across Europe reflect large-scale tectonic events. During the Caledonian Orogeny (490-400Ma), the tectonic plates of Laurentia and Avalonia collided as the Iapetus Ocean between them closed. The junction between these two tectonic plates (Fig 1) is called the Iapetus Suture and it runs on a NE-SW line from Berwick-upon-Tweed to the Solway Firth and projects into Ireland [22]. The underlying geology to the north and south of this suture is fundamentally different; the Laurentian basement, to the north, is geologically much older (> 3000Ma-c. 1750Ma) and is depleted in uranium (U) [23] whereas the Avalonia basement in the south is geologically much younger (c. 700Ma) [22], and this means the Pb isotope compositions, related to the basements of the two areas, are different. As this geological boundary essentially defines the modern political border of Scotland with England, it provides a potential method of discriminating between Scotland and the rest of Great Britain.
Marine carbonates, such as Chalk and limestone, often display 'anomalous Pb' values [10] because they are formed as precipitates from seawater, in which the U and Pb present do not retain a proportionate parent/daughter association. This is because U and Pb have very different solubility coefficients in seawater with marine carbonates accommodating an upper limit of 3 mg/kg for uranium but only 1 μg/Kg for Pb and thorium (Th). This imbalance in uptake of U and Pb can result in modern carbonates with very high 206 Pb/ 204 Pb values, which leads to a wide range of negative model ages being calculated [24]. Such features are documented in a study of Pb isotope data from Cenozoic limestones from South America [4] and in soil leach and re-equilibrated dentine values from British Chalk sites [25].

Test case Neolithic animals
Stonehenge, and other henges in southern England were a focal point for gatherings during Neolithic times. The geographic origin of the people who created and used them is a key question to understanding their construction and use. However, Neolithic human remains are not common at many of these henges and those that are present tend to have been cremated [26], reducing the range of analytical approaches that can be employed. Consequently, studies have focused on using the feasting remains from domestic animals as a proxy for human movement. The Sr isotope values from human cremated bone are consistent with a southern British source, but bone represents a longer biological time scale and is therefore subject to biogenic averaging [27] whereas Sr isotope values suggested that Scotland could not be excluded as a source for the animals, though recent advances in mapping raise other possibilities [28]. The discovery of highly radiogenic 87 Sr/ 86 Sr values (up to 0.7172) in Neolithic animals at the Wiltshire henges has raised the issue of whether some pigs could have been brought, as livestock, from northern Britain to Wessex during the Neolithic [29,30], a possibility that has sparked debate [31,32]. This study explores the value of Pb isotopes to this debate.
A multi-isotope study (strontium (Sr), oxygen (O), carbon (C), nitrogen (N) and sulphur (S)) on tooth enamel from pigs observed that the animals had diverse isotope compositions (δ 34 S VCDT = -1.6 to 19.6, δ 13 C (collagen) = -23.1 to -19.7, δ 15 N (collagen) = 3.4 to 9.1, 87 Sr/ 86 Sr: 0.7080 to 0.7172 and δ 18 O (carbVSMOW) = 23.2 to 28.9). The S isotope range covers areas multiple lithologies from the coast to inland Britain and the Sr isotope values encompass virtually the full extent of the British biosphere range. There are also clear dietary differences recorded in C & N isotope data and the O isotope range (5.7‰), while not directly comparable human data, is not far from the full British human enamel range (6.4‰, n = 824) [33]. This data diversity

PLOS ONE
Lead (Pb) isotopes domains for human and animal mobility means that there are few areas of Great Britain that can be clearly excluded for the overall source of animals found in the Neolithic henges.
Since the publication of Madgwick et al. 2019 [30] new research has uncovered new sources of highly radiogenic Sr in England. Biosphere mapping of southwest England, guided by geochemical soil profiles [34], has highlighted areas around the granite intrusions in Devon and Cornwall where Sr in plants reaches values of 87 Sr/ 86 Sr = 0.72 [28] and Roman and Medieval sheep from this region record 87 Sr/ 86 Sr ratios up to 0.717 [35,36]. In addition, it has been shown [37,38] that well-established forest on certain terrains, typically those with soils derived from both siliceous and carbonate bedrock, will leach out the carbonate component over time, resulting in a more radiogenic signature for the forest area. These developments mean there are now areas of England that can be considered as a possible source region as well as parts of Scotland. This paper establishes the use of Pb isotopes as a tool for discrimination between origins in northern and southern Britain.

Analytical details
The Pb analysis in this paper come from a number of sources: 1) new analyses of tooth enamel and bone from Durrington Walls, Cladh Hallan (CH) and Kings Gate, and from mineral samples from the Malvern Complex and the Lundy granite.
The tooth enamel selected was fully mineralized with good surface preservation and was prepared as follows: samples of enamel between 50-100mg were cut from the teeth using a rotary diamond wheel. The surfaces were fully abraded to a depth of 10-100 microns and all dentine was removed using a diamond burr. Bone samples were similarly abraded and cut using a rotary blade. The resulting sample was transferred to a clean (class 100, laminar flow) working area for further preparation. Samples were then cleaned by placing in >18 megaohmcm resistivity Milli-Q water at 60˚C for an hour and washed twice in Milli-Q water. They were then leached in 2% Teflon© distilled HNO 3 for approximately 2 minutes. After a final rinse, the samples were dried and transferred into a pre-cleaned Teflon© beaker where they were dissolved in Teflon© distilled 16MHNO 3 , evaporated to dryness and converted to bromide form using Romil© UpA HBr.
The samples from the Malvern Complex and the Lundy Granite were drilled out from feldspar crystals exposed on the clean-cut surface of the hand specimens held in the National Geoscience Data Centre (NGDC). The resulting mineral powder was dissolved and dried down in Teflon distilled 29M HF and 16M HNO3 then 6MCl and then converted to Bromide form using Romil© UpA HBr.
2) We were unable to access many of the samples we wanted to analyse due to Covid departmental and museum closures. However, we found we could use residual aliquots of solutions previously prepared for Sr isotope analysis for pig [30] and sheep enamel [35,36]. This had the advantage of minimizing unnecessary further sample damage of the teeth but meant we were restricted to those samples for which sufficient solution remained. These solutions were in chloride form and were dried down and converted to Bromide form using Romil© UpA HBr.
3) Previously published Pb analyse of pig enamel samples from Neolithic pig assemblage [5] is included in Table 1 and the preparation details for these samples are given in the source paper.
Separation of Pb from samples was undertaken using standard AG-1 X8 ion exchange techniques. In some cases, the residual solution from the AG-1 X8 column was converted to chloride and passed through an Eichrom© 50AG X8 column to separate out Sr. Table 1. Compilation of Pb and Sr isotope data from this study and previously published data. $-previously published Pb data [5], # samples solutions [30] x-samples solutions [35,36]. Unmarked samples are new analysis from tooth enamel and bone prepared for this study. Contribution of laboratory blank to analysis. The laboratory Pb blank is between 5-30pg with a fall in blank composition, measured on the clean lab windowsill outwith the laminar flow hoods, of 206 Pb/ 204 Pb = 17.72. Therefore a 50-milligram sample with 0.1ppm Pb, combined with 30pg blank contribution would provide a worst-case blank contribution of 0.6%. For a sample with a 206 Pb/ 204 Pb ratio of 18.80 this would result in a shift of -0.01 and is therefore negligible for the purposes of this study.
Samples analysed by Thermo Fisher Scientific Neptune Plus MC-ICP-MS were spiked with a thallium (Tl) solution and introduced into the instrument via an ESI 50 μl/min PFA microconcentric nebuliser attached to a de-solvating unit (Cetac Aridus II) and normalised to NBS981 [39]. Average 2 SD reproducibility for the following ratios was 206 Pb/ 204 Pb = 0.008%; 207 Pb/ 204 Pb = 0.008%; 208 Pb/ 204 Pb = 0.009%. Sr was loaded onto a single Re Filament [40] and the isotope composition and Sr concentrations were determined by Thermal Ionisation Mass Spectrometry (TIMS) using a Thermo Triton multi-collector mass spectrometer. The international standard for 87 Sr/ 86 Sr, NBS987, gave a value of 0.710262 ± .000020 (2SD, n = 8) during the analysis of these samples. Data are corrected to an accepted value of for NBS987 of 0.710250.

A new Pb iso-scape for Great Britain
A Pb isotope zonation map of Britain has been created which highlights the tectonically controlled geospatial variation in Pb isotope values across Britain using published data [14][15][16][17][18]20] and data from this study. It is contoured using inverse distance weighting (IDW) with ten natural break categories for 206 Pb/ 204 Pb; and 238 U/ 204 Pb (μ) which are shown in Fig 2 alongside the sample sites (n = 633). The maps show the strong zonation of the isotope compositions across the Iapetus Suture and throughout Scotland which reflects the influence of the very old Laurentian basement that underlies much of Scotland. England is dominated by Hercynian mineralisation focussed on the Pennines, Mendips, and SW England ore fields whereas the ores in Wales have an older Avalonian signature.
Superimposed over this map is an outline of Chalk outcrop. This lithology is an important substrate in southern England and its domain is designated an isotope range of 206 Pb/ 204 Pb 18.815 ±0.0195(n = 26) and 238 U/ 204 Pb (μ) of 9.71±0.24 (1SD, n = 26), using data from this study and Montgomery (2002) [25].

The construction of reference data fields
We have created four reference data fields in μ-T space, representing the main geographic Pb isotope sub-divisions defined in Fig 2. The co-ordinates for the field can be obtained from J. Evans.
Ores from Scotland: The data from Scotland define a broad sloping field from c 70Ma-2400 Ma [14]. We focus on the younger end of this array which is constructed using the

PLOS ONE
Lead (Pb) isotopes domains for human and animal mobility following data sources [14-17, 19, 41]. The boundary lines for this field encompasses 90% of these data. Ores from Wales: The 1SD field for T vs μ, for data from Wales is taken from published studies [14,41] and includes samples of Avonian basement from the Malverns Complex (Table 1).
Ores from England: The Pennines, Mendips and south-west England mineralization are all similar geological age. There is a large amount of data published from these areas [14,20,42]. The data field is based on the 1SD range of data (n = 197).
The Chalk: This is the rock that underlies much of southern Britain and is an important lithology for archaeological studies because it is both a focus of important Neolithic and Bronze Age sites and because of its preservation properties. The Chalk does not host economic Pb mineralisation so there is an absence of published Pb isotope data in the geological literature. We have derived the 1SD range for T and μ from Pb isotope analysis of Chalk soil leaches and equilibrated dentine [25], and bone samples (this study).

Testing the validity of using mineral composition as a proxy for bioavailable Pb
The transmission of these geological signals into the biosphere is critical for the application of this method to provenance studies. Fig 3 shows the reference fields in T vs μ space with archaeological faunal/human data to validate them. This exercise is limited by the paucity of biosphere data currently available. However, there is sufficient data to demonstrate the assumption that, in a situation of geogenic exposure, the faunal/human data reflect the labile Pb isotope composition of the local geological environment.
There are two sites where Pb isotope analyses are available from pre-mining (Neolithic) human enamel [45,46]. The Sr isotope composition of the tooth enamel has been used to select those individuals that had a childhood origin on Chalk. These include three juveniles [45] and one adult [46]. These four samples have Pb isotope compositions that plot within the 1SD field designated to represent the Pb isotope composition of the Chalk and thus supports the transfer of the geogenic Pb composition into the tooth enamel.
The validation of the ore field in England is based on new data presented in this study from south-west England. The samples are from sheep/goat samples of Roman [36] and Medieval [35] age excavated in Exeter [47]. The Pb data from these animals plot as a cluster within 1SD of the Pb isotope field, thus showing the relationship between the Pb isotope compositions of the mineralization in England and the faunal biosphere.
Measurements of modern animals provide evidence for transmission of geogenic Pb isotope composition from ore field for Wales into the biosphere field. These samples were specifically collected from Wales in areas where Pb mining is extensive. Consequently, natural Pb

PLOS ONE
Lead (Pb) isotopes domains for human and animal mobility levels are high and the influence of modern pollution is swamped by the geogenic component in the environment [43]. The data plot well within the 1SD data field based on mineral data from Wales.
The data used to validate the transmission Pb isotope compositions between the geogenic source and over lying biosphere for the ore field for Scotland comes from sites on the Hebrides and Orkney and both human and animal samples have been used to try and characterize this data field. All the samples have low Pb concentrations consistent with only geogenic lead exposure and/or predate anthropogenic lead release. The data scatter and human samples show considerable variation in μ and Pb model ages. Unlike the previous three data guidance fields for Chalk, England and Wales ores, the Pb isotope composition of the biosphere samples does not plot neatly within the ore field for Scotland, however, these samples show two distinguishing features. All samples have low μ values, between 9.6 and 8.9, and they also have elevated κ values >4.0 which are typical of ores from Scotland [14]. It is hoped that future studies can develop this validation of the dataset from Scotland with well provenanced samples, free from modern anthropogenic contamination, however, preservation is a problem in large parts of Scotland. We recommend that, while this method can be used to distinguish between sources in northern and southern Britain, the transmission of the high resolution Pb isotope zones seen in the mineral data (Fig 3) cannot currently be demonstrated in biosphere samples.
In summary, reference data fields for the Pb isotope composition of geographical areas of Britain, have been defined with geological data and the transfer of these signatures into the biosphere has been validated with well-provenanced faunal data. It should be noted that as most of the fields are defined at 1SD, which represent c. 66% of the reference data, plotting outside the field only excludes samples at this level of certainty. With this validation in place, the data from the pigs can be examined.

Test case using pig enamel samples
The Pb isotope compositions of the pig enamel are presented in Table 1 and plotted in the form of μ versus T in Fig 4. The data points are colour-coded to represent 87 Sr/ 86 Sr isotope composition groupings of the data: <0.709, 0.709-0.713, and > 0.713. These Sr groups broadly represent a) carbonates terrains, b) most silicate terrains and c) Rb-rich/very old terrains, respectively.
The Pb isotope composition from the pig enamel samples, have model ages that range between the negative values of the southern Britain Chalk field to the Palaeozoic model ages as old as 482 Ma, and all have μ values above 9.62. The μ values excludes them from the field of ore data from Scotland and hence the Pb isotope data indicate that these animals which represent a 36% subset of those in Madgwick et al. 2019 [30], are inconsistent with an origin in Scotland.
Around half the samples have model ages that plot between 195 Ma and 482 Ma, indicating Pb sources from a Palaeozoic/Late Precambrian rocks substrates and three of the samples (DW31, DW40 and DW60) with radiogenic values (> 0.713) plot in the data field for Southwest and central England and hence could be sourced from these areas. Nine of the samples with 87 Sr/ 86 Sr < 0.709 have both Pb and Sr isotope characteristics consistent with an upbringing on a Chalk terrain (Fig 4) Having established that many of the samples record combined Sr and Pb isotope compositions that can be accommodated by known tectonic and lithological domains found in Britain, we are left with a few samples for which the data are more complex. These samples record elevated radiogenic composition for both Sr and Pb. They are referred to as the "anomalous group" and are discussed below.

The relationship between the Sr and Pb isotope compositions in tooth enamel
Although both biosphere Sr and Pb are ultimately sourced from a geological origin it should not be expected that there will be a direct correlation between the two systems; they will reflect different aspects of the source biosphere. A carbonate terrain has restricted and relatively low Sr isotope compositions, due to the very low rubidium (Rb) content of limestone, whereas the Pb isotopes may range in values considerably but will tend to have elevated 204 Pb/ 206 Pb values as a consequence of high and variable time-integrated U/Pb in the limestones [4]. This will lead to high μ values and negative model ages [21].
In silicate terrains the relationships between Sr and Pb isotopes are the inverse of carbonates terrains. The variable lithologies have a wide range of Rb/Sr ratios leading to transmission of wide-ranging 87 Sr/ 86 Sr values into the biosphere whereas the Pb isotopes will reflect the overall tectonic environment.
There are six samples (three animals from Marden and two animals ( (Fig 4). This combination of radiogenic Pb and Sr isotope features is very difficult to reconcile with a single lithological source and essentially suggests a Tertiary source with highly radiogenic Sr composition. Such a rock type does occur in the form of the small island of Lundy (4.5km 2 ) off the north Devon coast [48]. While we cannot exclude this granite zone as a possible origin, and it is known to have been inhabited by Neolithic communities [49], it would not be parsimonious to assign origins for the anomalous pig samples to Lundy, based on the island's small size, and relative remoteness Several possible alternative explanations can be considered. The isotope signal could derive from an upbringing on a sedimentary rock composed of old/Rb rich silicate minerals in a carbonate matrix where the Sr was predominantly derived from the silicate component whereas the Pb was predominantly from the carbonate matrix. However, this isotope mismatch could also reflect metabolic discrimination rather than a simple foddering source. It is known that Pb can be reactivated into the blood stream during times of stress [50], so it is possible that this isotope signal could be caused by mobility, foddering or stress that disconnects the sources of Pb and Sr from different body reserves during the formation of enamel in the animals. Understanding these possible processes is beyond the remit of this paper and will involve detailed investigation on the nature and relationship of Pb and Sr uptake during enamel formation.

Conclusions
It has been frequently stated, and needs to be re-iterated, that isotope methods can only exclude possible areas of origin and, added to that, can only exclude options for which reference data are available, reliable and appropriate. This paper develops the use of Pb isotopes as a further isotope discriminant for biosphere samples that have not been affected by anthropogenic pollution; either pre-metallurgical populations or those with enamel samples where Pb concentrations below the threshold for natural exposure in skeletal studies of 0.5 [11] or 0.7 mg/kg [51] in [human] enamel.
We have produced the first Pb isoscape map for Great Britain which provides a geographic distribution of Pb mineralization events and also characterizes the Chalk, which is an important lithology for archaeological studies. These main subdivisions are defined in T-μ space and have been validated for biosphere use with human and faunal data.
This study highlights the way in which Pb isotope domains are associated with large tectonic terranes and, because to the tectonic make-up of Great Britain, provides a powerful tool for discrimination between biosphere in Scotland (the Laurentian basement) and southern Britain (the Avalonia basement).
We have tested this application using a sample of Neolithic pig enamel from sites in southern England, some of which, because of Sr isotope composition, could not be excluded from an origin in northern Britain. Pb isotope data from the teeth excludes Scotland as a source but the diverse range of Pb isotope results, combined with other isotope proxies, are consistent with the animals being raised on a variety of lithologies of diverse age and from variable environments.
The study also highlights the fact that the uptake of both Sr and Pb may not always be consanguineous as some samples display a combination of Sr and Pb isotope signatures that are currently difficult to reconcile with a single geographic/geological origin.
This study demonstrates the diagnostic potential of Pb isotope compositions as a provenance tool in samples unaffected by anthropogenic Pb exposure and highlights the different scale of discrimination between tectonic terrains which Pb isotopes provide in comparison to the lithological scale of Sr isotope data.