Introduction

Strontium (Sr) isotopes are a valuable tool in archaeology for investigating past mobility and migration patterns of humans and animals (Reynard and Balter 2014; Copeland et al. 2016; Copeland et al. 2011; Copeland et al. 2010; Hoppe et al. 1999; Stewart et al. 2020; Willmes 2015; Wooller et al. 2021; Joannes-Boyau et al. 2019; Ericson 1985; Leach et al. 2009). Strontium isotope ratios in tooth enamel reflect the geological composition of the region where an individual lived during tooth formation. Their measurement can allow for the identification of individuals who migrated to a new region during their lifetime, and to estimate the distances and directions of their migrations. Additionally, Sr isotopes can help distinguish between local residents and non-local individuals buried in the same cemetery, shedding light on social and cultural practices (Price et al. 2011; Evans et al. 2006). Ericson (1985) discuss the progress and prospects of strontium isotope analysis in archaeology, while Price and Gestsdóttir (2014) and Frei et al. (2015) present case studies of its application to Icelandic and Bronze Age contexts, respectively. Such studies have used strontium isotope ratios to identify non-local individuals who likely travelled long distances to a region, providing evidence for long-distance exchange networks and inter-regional interactions.

While Sr isotope measurement in enamel requires only a small amount of material and is of limited destructive nature, it is often not possible to obtain specimens, because of their value in collections. When this is the case, an alternative would be to consider fossil dental calculus, which generally has lesser value to collections. Dental calculus is a calcified deposit that forms on teeth due to the build-up of bacteria and other substances in the mouth (Lang and Bartold 2018). Dental calculus has received increasing attention as a source of information about past diets, health, and behaviour (Power et al. 2014). One of the main uses of dental calculus analysis in archaeology is to reconstruct ancient diets, nutritional deficiencies, and other health conditions (e.g. Warinner et al. 2014; Buckley et al. 2014; Hendy et al. 2018; Henry et al. 2011). Dental calculus has also been studied for DNA and other biomolecules preserved within it. This approach has the potential to reveal information about the genetic makeup of past populations, as well as the presence of pathogens and other microorganisms (Weyrich et al. 2017).

In this study, we investigate whether Sr isotopes in dental calculus could be used to assess the origin of an individual. Calculus is much less dense, more porous, and chemically more diverse than enamel. Thus, one potential obvious limitation is that calculus may be prone to diagenesis, thus not providing an accurate record of the individual’s Sr isotope signature. In addition to Sr isotopes, trace element ratios were also measured, as Mn/Ca or U/Ca ratios are commonly used as markers of diagenesis (Reynard and Balter 2014). Measurements in fossil calculus were paired and compared with measurements in enamel and dentine of corresponding teeth, since enamel is more dense and commonly less prone to diagenesis (Budd et al. 2000). We also measured trace element ratios in modern calculus, to provide a comparison with fossil material and assess further diagenetic uptake of trace elements such as Sr.

Material and Methods

Fossil calculus and teeth samples were previously studied in Adler et al. (2013). Specimens are Late Medieval (1200–1300 CE) in age, from Jewbury Cemetery, in York, UK (Lilley et al. 1994). Modern (2022 CE) calculus samples were provided by a dentist practice in Wollongong, NSW, Australia. Bulk and in situ analysis of trace element concentrations and Sr isotopes were conducted at the Wollongong Isotope Geochronology Laboratory (WIGL). Only calculus samples were processed for bulk analysis.

For bulk analysis of trace element concentrations, 5–12 mg of calculus (fossil or modern) was weighed and dissolved in high-purity HNO3 and H2O2 (Seastar Baseline® grade). Solutions were dried, re-dissolved in 2M HNO3, and an aliquot was taken for trace element analysis. Analyses were performed on a ThermoFisher iCAP-Q quadrupole inductively coupled plasma mass spectrometer (ICP-MS), with PFA nebuliser, quartz cyclonic spray chamber, nickel sample and skimmer cone, and high matrix insert. Measured analyte intensities were blank corrected, and concentrations were determined using six calibration standards (Inorganic Ventures 71A) with concentrations ranging from 0.1 to 100 ppb. The following elements were measured: calcium (Ca), manganese (Mn), strontium (Sr), lanthanides, and uranium (U). For the fossil calculus, no internal standard was used; for the modern calculus, 89Y was used as internal standard for Ca, Mn, and Sr, 159Tb for lanthanides and 209Bi for U (fossil and modern calculus were analysed three years apart). Internal analytical uncertainties were better than 0.5% for Ca, better than 1% for Mn, Sr, lanthanides, and U. Accuracy during analysis was better than 10%.

For bulk analysis (thereafter termed ‘solution analysis’) of Sr isotopes, a fraction of the aliquot dissolved for trace element concentrations was isolated, dried down, re-dissolved in 2M HNO3, and processed for ion exchange chromatography. Strontium was isolated from the sample matrix using automated, low-pressure chromatographic system Elemental Scientific prepFAST-MC™ and a 1 mL Sr–Ca column (Eichrom™) (Field and Sullivan 2015; Romaniello et al. 2015). The Sr elutions were re-dissolved in 0.3 M HNO3. Strontium isotope analysis was performed on a ThermoFisher Neptune Plus multi-collector inductively coupled plasma mass spectrometry (MC ICP-MS) at WIGL. The sample introduction system consists of an ESI Apex-ST PFA MicroFlow nebulizer with an uptake rate of ~ 0.1 mL/min, a SSI Quartz dual cyclonic spray chamber, jet sample, and X skimmer cones. Measurements were performed in low-resolution mode. The instrument was tuned at the start of each session with a 20 ppb Sr solution, and sensitivity for 88Sr was typically around 4 V. Masses 88, 87, 86, 85, 84, and 83 were collected on Faraday cups. For samples, intensity on mass 88 was typically ~ 4 V, < 0.0005 V on mass 85, and < 0.0001 V on mass 83. Masses 85 and 83 were used to correct the 87Sr/86Sr ratio for the isobaric interference of 87Rb and 86Kr, respectively, as well as to calculate the 87Rb/86Sr ratio. Instrumental mass bias correction was applied to 87Sr/86Sr and 87Rb/86Sr ratios using measured 88Sr/86Sr (and a true value of 8.375209). NIST SRM987 was measured after every five samples to assess accuracy during analysis. Accuracy of the whole procedure was assessed by processing NIST SRM 1400 and 1486. 87Sr/86Sr values for SRM 1400 and SRM 1486 were 0.71309 ± 2 and 0.70930 ± 2 (2SE), respectively; comparable to values published in Brazier et al. (2020) (0.71308 ± 2 and 0.70931 ± 2, respectively). Total procedure blank was 35 pg Sr (the typical amount of Sr processed for analysis was ~ 100 ng).

In situ analysis of trace element concentrations was performed on fossil calculus and teeth, by laser ablation ICP-MS at WIGL. Laser ablation was performed using a New Wave Research™ 193 nm ArF excimer laser, equipped with a TV2 cell. Samples were ablated for 60 s with a spot size of 50 μm, a laser pulse rate of 20 Hz, and fluence set to 4 J/cm2. Pre-ablation was performed 2 s with a spot size of 60 μm, a laser pulse rate of 5 Hz, and fluence set to 1 J/cm2. Helium (He) was used as a carrier gas at a flow rate of 0.8 L/min. Before entering the ICP-MS, the He carrier gas was mixed with the ICP-MS Ar sample gas in a glass smoothing device. Nickel sample and skimmer cones with a high sensitivity insert were used. At the start of each session, the laser and ICP-MS were tuned using NIST SRM 612 glass for maximum sensitivity and minimum oxide level. USGS MAPS-4 reference material was used as calibration standard, and analysed before and after each sample. The typical analysis sequence consisted of 5 spots on MAPS-4, 10–20 spots on a sample, 5 spots on MAPS-4, and so on, finishing with 5 spots on MAPS-4. On the ICP-MS, the following analytes were collected: 43Ca, 48Ca, 55Mn, 88Sr, 137Ba, lanthanides, and 238U. Dwell time was 0.1 s for each analyte except for 43Ca, 48Ca, and 88Sr (0.01 s). For 48Ca, an interference correction was applied using 47Ti. Data were processed in Iolite 4™ (Paton et al. 2011) performing a baseline correction, using MAPS-4 as external standard and 43Ca as internal standard (applying values of 33.83 wt. % for MAPS-4, 39.89 wt. % for enamel and 28 wt. % for dentine). Calcium concentrations were then derived using 48Ca (and other element concentrations using their respective analytes).

In situ analysis of Sr isotope ratios was performed on fossil calculus and teeth, by laser ablation MC ICP-MS at WIGL using the same laser ablation system as above and a ThermoFisher Neptune Plus MC ICP-MS. At the start of each session the laser and MC ICP-MS were tuned using NIST SRM 610 glass. Samples were ablated for 50 s with a spot size of 150 μm, a laser pulse rate of 10 Hz, and fluence set to 6 J/cm2. Pre-ablation was performed 2 s with a spot size of 150 μm, a laser pulse rate of 10 Hz, and fluence set to 1 J/cm2. Helium and nitrogen were used as carrier gas at a flow rate of 0.9 and 0.01 L/min, respectively. The carrier gas was mixed with the MC ICP-MS argon sample gas before entering the plasma. Jet sample and x skimmer cones were used. A clam shell with modern seawater Sr isotope ratio (0.709182 ± 0.000004 (Ma et al. 2013)) was used as primary standard, while a seal tooth enamel was ablated as secondary standard. The typical analysis sequence consisted of 3 spots on the clam, 3 spots on the seal tooth enamel, 10–20 spots on a sample, 3 spots on the clam, 3 spots on the seal tooth enamel, and so on, finishing with 3 spots on the clam. Masses 82 to 88 were collected in Faraday cups in static mode. Typical intensity on mass 88 was 1–2 V and < 0.0003 V on mass 82. Data were processed in Iolite 4™ (Paton et al. 2011) using the “Sr isotopes” data reduction scheme. A baseline correction was applied. Signal collected at mass 82 was used to correct 88Sr and 86Sr from isobaric interferences (mainly CaAr2+), and signal collected at mass 85 to correct 87Sr from 87Rb isobaric interference. Mass bias was quantified using the interference-corrected 88Sr/86Sr ratio and applied to the interference-corrected 87Sr/86Sr ratio. Finally, the corrected 87Sr/86Sr ratio of samples and seal tooth underwent an additional correction by standard-sample-standard bracketing using the corrected 87Sr/86Sr ratio of the clam to improve accuracy. The mean 87Sr/86Sr value for seal tooth enamel was 0.709281 ± 0.000004 (2 standard errors, n = 63). This differs from the modern seawater value (0.709182 ± 0.000004 (Ma et al. 2013)) by 0.0001, which is comparable to quality control results obtained by LA MC ICP-MS in other studies (Boethius et al. 2022).

Results

Bulk analysis yields Mn/Ca, Sr/Ca, and U/Ca ratios ranging, respectively, from 5.0 to 52 × 10−4, 12.0 to 15.8, and 0.052 to 0.154 in fossil calculus, and from 0.05 to 0.35 × 10−4, 1.1 to 2.1, and less than 8.3 × 10−6 in modern calculus (Table 1). Rare earth element (REE) concentrations determined by bulk analysis range from 0.005 to 4.8 ppm in fossil calculus, and mostly below detection limits (ranging from 0.002 to 0.005 ppm) in modern calculus. In situ analysis of fossil tooth enamel and dentine yields Mn/Ca ratios ranging from 0.0035 to 3.2 × 10−4 in dentine, 0.0059 to 15 × 10−4 in enamel (Supplementary Table 1; Fig. 1) (fossil or modern calculus was not analysed in situ for trace element concentrations). Mn/Ca ratios are higher in fossil calculus compared to dentine or enamel, or to modern calculus. In most samples, Mn/Ca is similar between dentine and enamel; although in one sample it is greater in enamel compared to dentine (ACAD8326), and lower in another (ACAD8328).

Table 1 Element concentrations and Sr isotope ratios in tooth calculus (bulk analysis)
Full size table
Fig. 1
figure 1

Boxplot of log10 values of Mn/Ca ratios in fossil enamel, dentine (both measured by in situ analysis), fossil, and modern calculus (measured by bulk analysis)

Full size image

Bulk analysis yields Sr/Ca ratios ranging from 12 to 25 × 10−4 in fossil calculus and from 1.1 to 2.1 × 10−4 in modern calculus (Table 1). In situ analysis of tooth enamel and dentine yields Sr/Ca ranging from 3.3 to 9.9 × 10−4 in dentine, 2.7 to 5.8 × 10−4 in enamel (Fig. 2). Bulk analysis of modern calculus yields Ba/Ca ratios ranging from 0.09 to 0.43 × 10−4 (Ba was not measured in fossil calculus). In situ analysis of tooth enamel and dentine yields Ba/Ca ranging from 0.11 to 2.5 × 10−4 in dentine, 0.21 to 4.6 × 10−4 in enamel. Ba/Ca values are generally lower in enamel than dentine, except for one enamel analysis (Fig. 3). Bulk analysis of modern calculus yields U/Ca ratios ranging from below detection limit to 0.083 × 10−8. In situ analysis of tooth enamel and dentine yields U/Ca ranging from below detection limit to 166 × 10−8 in dentine, below detection limit to 381 × 10−8 in enamel (not shown).

Fig. 2
figure 2

Boxplot of log10 values of Sr/Ca ratios in fossil enamel, dentine (both measured by in situ analysis), fossil, and modern calculus (measured by bulk analysis)

Full size image
Fig. 3
figure 3

Boxplot of log10 values of Ba/Ca ratios in fossil enamel and dentine (both measured by in situ analysis), and modern calculus (measured by bulk analysis)

Full size image

Solution analysis of 87Sr/86Sr ratios in fossil dental calculus yields values between 0.710068 ± 0.000021 and 0.710858 ± 0.000021 (Table 1), while in situ analysis yields values between 0.710313 ± 0.000075 and 0.71174 ± 0.00019 (Supplementary Table 2). In situ analysis of fossil dentine and enamel yields 87Sr/86Sr values between 0.70676 ± 0.00022 and 0.718470 ± 0.000096, and between 0.70637 ± 0.00025 and 0.710783 ± 0.000078, respectively (Supplementary Table 2). Modern calculus was not measured for Sr isotopes, since its origin is unrelated to that of fossil specimens and thus is of little value for comparison of Sr isotope ratios (while comparison of trace element ratios does provide useful insights into diagenetic uptake of such trace elements).

Discussion

Diagenetic Uptake of Trace Elements

In fossil calculus, Mn/Ca ratios are one to three orders of magnitude greater than in dentine or enamel (Fig. 1). Because calculus is much more porous than dentine and enamel, it is more prone to diagenetic uptake of various elements. Alternatively, since the mineralogy of dental calculus is more diverse than that of enamel or dentine (Grøn et al. 1967), it is possible that its Mn/Ca is higher than that of dentine and enamel during formation, independent of diagenetic alteration. However, fossil calculus displays Mn/Ca ratios 14 to 1100 times higher than modern calculus (Fig. 1). Similarly, fossil calculus shows Sr/Ca ratios 0.25 to 2 times higher than those of dentine or enamel, and ~ 10 times higher than modern calculus (Fig. 2). Here, too, this likely reflects diagenetic alteration and uptake of Sr. The same is observed for REE, which are enriched by several orders of magnitude in fossil calculus compared to modern calculus (Table 1). Finally, U/Ca in fossil calculus is more than 5000 times greater than in modern calculus (Table 1), illustrating that U uptake occurs within a few centuries of burial. Overall, it is clear that fossil calculus is diagenetically altered, with a significant uptake of Sr and other trace elements such as Mn and U, especially when compared to modern calculus. Below we explore how diagenetic Sr may impact the Sr isotope composition of fossil calculus.

Mn/Ca ratios in dentine are similar, lower, or greater than those in enamel (Fig. 1), suggesting that it is difficult to argue for an unequivocal greater uptake of Mn in dentine compared to enamel, during diagenesis (except perhaps for sample ACAD8328). Conversely, Sr/Ca and Ba/Ca ratios in dentine are systematically greater than in enamel (Figs. 2 and 3), illustrating the diagenetic uptake of Sr and Ba in dentine.

Strontium Isotope Composition of Enamel, Dentine, and Calculus

The Sr isotope values measured in different sample materials are ordered as follows: fossil calculus (in situ) > fossil calculus (solution) > dentine ≥ enamel (Fig. 4). The shift between fossil calculus (in situ) and enamel ranges from 0.0002 to 0.0016, and between fossil calculus (solution) and enamel ranges from 0.0002 to 0.0014. These ranges are comparable to differences between ratios in enamel determined by in situ vs. solution (up to 0.0017; Copeland et al. 2010). Note that the difference between fossil calculus determined in situ vs. solution (0.00002–0.0007) is less than the difference for enamel analysed in situ vs. solution in Copeland et al. (2010). Greater 87Sr/86Sr values determined in situ compared to solution analysis of fossil calculus could potentially result from the isobaric interference of 87Rb onto 87Sr. Rubidium is not removed during in situ analysis, while it is removed by chromatography prior to solution analysis. This is illustrated by the 87Rb/86Sr ratios in fossil calculus analysed in solution several orders of magnitude lower than during in situ analysis (Fig. 5). While there are no clear positive relationships between 87Sr/86Sr and 87Rb/87Sr for in situ and solution calculus analyses, when each sample is considered separately, in situ analyses generally show higher 87Sr/86Sr and 87Rb/87Sr values than solution analyses (except for sample ACAD8477; Fig. 6).

Fig. 4
figure 4

Boxplot of 87Sr/86Sr ratios in fossil enamel, dentine (both measured by in situ analysis), and calculus (measured by solution analysis—in navy—or in situ analysis—in red)

Full size image
Fig. 5
figure 5

Boxplot of the log10 values of 87Rb/86Sr ratios in fossil enamel, dentine (both measured by in situ analysis) and calculus (measured by solution analysis—in navy—or in situ analysis—in red)

Full size image
Fig. 6
figure 6

87Sr/86Sr ratios as a function of the log10 values of 87Rb/86Sr ratios for in situ and solution calculus analyses

Full size image

The shift between dentine and enamel ranges from 0.00004 to 0.001. This is less than what Copeland et al. (2010) found in rodents (generally < 0.001 but up to 0.023), suggesting that diagenetic alteration of dentine did not significantly affect its Sr isotope composition. Greater 87Sr/86Sr values in dentine compared to enamel could result from the isobaric interference of 87Rb onto 87Sr, since dentine is more porous than enamel and thus more prone to take up metals such as Rb, following burial (and Rb is not removed during in situ analysis). However, there is no systematic variation of the Rb/Sr ratio between dentine and enamel. Rb/Sr ratios in dentine are greater, similar, or lower than those in enamel, depending on the sample (Fig. 5). For instance, in samples ACAD8328 and ACAD8817 where 87Sr/86Sr values in dentine are greater than in enamel, 87Rb/87Sr are lower in dentine compared to enamel. Alternate hypotheses are (i) that primary dentine records a different diet origin than enamel, or (ii) Sr with a high 87Sr/86Sr ratio is incorporated post-burial. The former is unlikely since dentine starts mineralising only shortly before enamel (Cate and Richard 1980). For each sample, solution analysis of fossil calculus yields 87Sr/86Sr values greater than those in dentine or enamel. This observation would also support to the incorporation of Sr with a high 87Sr/86Sr ratio post-burial (since Rb is removed prior to solution analysis and high values cannot be explained by isobaric interference from 87Rb).

For many analyses of enamel and dentine in sample ACAD8327, there is a negative relationship between 87Sr/86Sr and 87Rb/86Sr (Fig. 7). This is interesting as more 87Rb should result in an overestimation of the 87Sr/86Sr ratio and thus produce a positive relationship. One of the corollaries of this observation is that correction of in situ 87Sr/86Sr analyses for the presence of 87Rb seems to be effective. Samples ACAD8328 and ACAD8817 show the same negative relationship but enamel analyses clearly have higher Rb/Sr and lower 87Sr/86Sr values than the dentine (Fig. 7). This would imply that enamel is more diagenetically altered than dentine, which is unlikely since it is denser and it is generally accepted that enamel is more resistant to diagenesis (e.g. Reynard and Balter 2014). However, while Ba/Ca and Sr/Ca are lower in enamel compared to dentine, this is not systematically the case for Mn/Ca ratios.

Fig. 7
figure 7

87Sr/86Sr ratios as a function of the log10 values of 87Rb/86Sr ratios in samples a ACAD8327, b ACAD8328, and c ACAD8817, for enamel, dentine, and calculus

Full size image

Since higher Rb/Sr ratios would generally be interpreted as pointing towards the composition of a diagenetic end member (after ruling out isobaric interference from 87Rb), the negative relationships observed between 87Sr/86Sr and 87Rb/86Sr ratios could hint that the diagenetic end-member has a low 87Sr/86Sr value, 0.7085 or lower (Fig. 7). In absence of calculus data, this could be how fossil teeth data would have been interpreted. Nevertheless, this is somewhat lower that the expected modelled value for the York area: 0.7096–0.7117 (IQR) (Evans et al. 2018). The closest region with values between 0.708 and 0.7085 would be about 20 km east of York (Evans et al. 2018). If the diagenetic end member does have a 87Sr/86Sr ratio that low, a possible explanation would be that teeth were buried 20 km east of York for a period of time (during which Sr would be uptaken) before being re-located to York.

Where the negative relationship between 87Sr/86Sr and 87Rb/86Sr is observed, it seems to converge towards that measured in the calculus by solution analysis (~ 0.7100–0.7108) (Fig. 7). Perhaps this indicates that the diagenetic end member has a high 87Sr/86Sr ratio and, surprisingly, a low Rb/Sr ratio. Values between 0.7100 and 0.7108 could fall within the range of values expected for bioavailable Sr around York (Evans et al. 2018). Overall, the implications are that (i) where only fossil teeth data are available, using the Rb/Sr ratio could be misleading when trying to identify a diagenetic end member; (ii) the 87Sr/86Sr ratio of the fossil calculus is most likely dominated by Sr uptaken post-burial and is unlikely to yield useful information about the past location of an individual (unless it has been relocated post-burial); (iii) fossil calculus can be helpful in determining the Sr isotope composition of the diagenetic end member.

A corollary from this work is that the Sr isotope composition of fossil enamel for the different samples suggests that the individuals studied were not originally from York. To assess this, we used the range of 87Sr/86Sr values measured in enamel for each sample in combination with the Sr isoscape of the UK (using their 90% central data) (Evans et al. 2018). For samples ACAD8326, ACAD8477, and ACAD8817, results suggest that these individuals would have most likely originated at a minimum > 20 km east from York, or from southeastern England (Fig. 8). Interestingly, for samples ACAD8327 and ACAD8328, enamel 87Sr/86Sr ratios suggest these individuals would have originated from the Glasgow area, at the nearest, 100s of km away from York (Fig. 8).

Fig. 8
figure 8

Possible areas of origin (in orange) for the individuals studied here, using enamel 87Sr/86Sr ratios and the UK Biosphere Isotope Domains website (https://mapapps.bgs.ac.uk/biosphereisotopedomains/index.html) (Evans, et al. 2018) for samples A ACAD8326, B ACAD8327, C ACAD8328, D ACAD8477, and E ACAD8817

Full size image

Conclusions

In this study, we have assessed the potential to use Sr isotope ratios in fossil dental calculus as a way to investigate the geographical origin of individuals. When comparing modern and fossil dental calculus, it is clear that most metals present in fossil calculus were acquired post-burial, including Sr. Sr isotope and Rb/Sr ratios in calculus, dentine, and enamel suggest that the Sr isotope composition of calculus is dominated by that of diagenetic Sr. Combining these results, the Sr isotope composition of calculus seems to reflect that of the soil in which remains were buried, and not that at the time of calculus formation.

Interestingly, measuring Sr isotopes in calculus helped correctly identify the composition of the diagenetic end member. Without those measurements, a low 87Sr/86Sr ratio would probably have been assigned to the diagenetic end member, using high Rb/Sr ratios as a guide. This would have resulted in erroneous interpretations on the origin of those individuals and their burial. Thus, while Sr isotopes in fossil dental calculus cannot help constraining the geographical origin of humans and other animals, they could be helpful in correctly defining the composition of the diagenetic component. This could ensure correct interpretation of the Sr isotope composition of enamel, and constraining of geographical origin and migration.