Assessing the preservation of biogenic strontium isotope ratios (87Sr/86Sr) in the pars petrosa ossis temporalis of unburnt human skeletal remains: A case study from Saba

Rationale Strontium isotope (87Sr/86Sr) analysis of skeletal remains has become a powerful tool in archaeological studies of human migration and mobility. Owing to its resistance to post‐mortem alteration, dental enamel is the preferred sampling material used for 87Sr/86Sr analysis in bioarchaeological provenance research, although recent studies have demonstrated that cremated bone is also generally resistant to diagenesis. This paper presents the results of a pilot study exploring the potential of unburnt petrous bone (pars petrosa) as a reservoir of biogenic (diagenetically unaltered) strontium, as the otic capsule or bony labyrinth within the petrous bone is extremely dense and is thought to be unable to remodel after early childhood, potentially providing an alternative for dental enamel. Methods From an individual from a colonial‐era (18th century) site on the island of Saba in the Caribbean for whom previous enamel 87Sr/86Sr results had indicated non‐local origins, multiple locations (n = 4) on the petrous were sampled and measured for strontium isotope composition. Saba (13 km2) has been extensively mapped for baseline strontium isotopes (n = 50) with 87Sr/86Sr varying from ca 0.7065 to 0.7090, whereas enamel 87Sr/86Sr (n = 3) ranged from 0.7104 to 0.7112. Results All four petrous 87Sr/86Sr ratios (0.7111–0.7122) are consistently and considerably higher than the local bioavailable range, and very similar to the enamel 87Sr/86Sr. These results provide initial evidence that unburnt petrous bones may preserve biogenic strontium, at least in this specific burial context. Conclusions While more research in diverse burial conditions is needed to validate this observation, if confirmed, it would have broader implications for sample selection strategies in bioarchaeological studies using the strontium isotope method.


| INTRODUCTION
The human pars petrosa ossis temporalis has become the 'Holy Grail' in ancient DNA studies since pioneering research by Pinhasi et al in 2015 showed that relatively elevated quantities of high-quality endogenous aDNA were preserved in the bony otic capsule of the inner ear. 1 Recently, the otic capsule, or bony labyrinth, has proven to be valuable in archaeological isotope studies as well. 2 Due to their increased crystallinity, cremated or calcined archaeological skeletal remains have proven to exhibit the diagenetically unaltered, i.e. original or biogenic, strontium isotope composition ( 87 Sr/ 86 Sr) at the time of death. [3][4][5] Due to the extremely slow remodelling rate of the otic capsule compared to the surrounding bone of the petrous or other bones in the skeleton, [6][7][8] the bony labyrinth of cremated petrous portions (PP) is nowadays considered the equivalent of dental enamel and thus a reliable proxy for the geochemical environment of infancy. As such, calcined otic capsules can be effectively used to provide insights into the childhood (geographical) provenance of cremated burials for which dental enamel is often not well preserved or recovered. [8][9][10] However, to date, little research has been conducted on the applicability of the pars petrosa or otic capsule in unburnt archaeological skeletal assemblages for provenance studies in cases where dental elements are absent. A major problem in isotopic studies on the apatite component of bone, however, is the process of diagenesis. 11 The post-mortem chemical and physical changes that happen to a bone during burial have a negative and irreversible impact on the biogenic strontium isotope composition of bone bioapatite. In contrast to the porous structure of bone, the otic capsule is the densest bone in the human skeleton 12 and therefore considered to be less prone to diagenetic alterations.
To investigate the effect of diagenesis on the biogenic 87 Sr/ 86 Sr of the otic capsule of unburnt pars petrosae, and its applicability for provenance studies, a unique case study was investigated. In this study the 87 Sr/ 86 Sr 13 and Sr concentrations of enamel and dentine samples from three molars and of one pars petrosa are presented from an inhumed female from a historic period site on the island of Saba, a special municipality of the Netherlands in the northeast Caribbean. Based on the archaeological, osteoarchaeological and isotopic evidence (elevated, non-local 87 Sr/ 86 Sr), this individual was interpreted as an individual of African origin, who was enslaved and forced to migrate to Saba in the late 18th century. 13 The strontium isotope difference between the region of childhood origin (Africa) and the location of death (Saba) is immense. The bioavailable strontium isotope range on Saba (mean: 0.7083; range: 0.7065-0.7090) is much lower than that in much of West and West-Central Africa, which is underlain by diverse geological formations including large areas of continental bedrock (including cratonic formations) with highly variable but generally higher 87 Sr/ 86 Sr (greater than ca 0.710). 14 Consequently, a diagenetic shift in the otic capsule's 87 Sr/ 86 Sr towards the Saba range should be apparent if the otic capsule is affected by diagenesis. This paper presents the results of this pioneering study and provides directions for future research on the use of unburnt pars petrosae in archaeological mobility studies.

| DIAGENESIS IN PALAEOMOBILITY STUDIES
The effects of diagenesis on the isotopic integrity of archaeological bone and dentine have been the subject of research for the past few decades. Nearly four decades ago, when strontium isotope analysis was first introduced to the field of human biomolecular archaeology, 15,16 the complicating effects of diagenesis were already acknowledged. 17,18 In the 1980s, decontamination of (bone) tissue was thought to be accomplished through 'mechanical excoriation, tissue separation, and phase separation by chemical reaction' with acetic acid. 16,19,20 Nevertheless, convincing evidence of diagenetic alteration of bone chemistry and strontium isotope composition was provided shortly after in 1986 by Nelson and colleagues and confirmed by many studies thereafter. [21][22][23][24] Many early studies of human palaeomobility based on strontium isotope analysis focused on sampling of enamel and bone. 16,[25][26][27][28] Over time, more evidence accumulated demonstrating the higher susceptibility of bone and dentine to diagenesis (relative to enamel) generally, and post-mortem alteration of biogenic strontium isotope signals more specifically. 21,22,29 Owing to this, bone has become slowly abandoned as a suitable sampling material in most strontium isotope studies of palaeomobility over the last two decades.
Nevertheless, some studies continued to analyse bone and dentine strontium isotope signals as proxies of the local range of bioavailable 87 Sr/ 86 Sr under the debatable assumption that these sampling materials should be reflective of local diagenetic 87 Sr/ 86 Sr 30,31 or that all diagenetic contaminants were successfully removed. 32 Until several years ago, dental enamel was considered the only suitable sampling material for archaeological palaeomobility and provenance studies owing to the fact that it does not undergo remodelling once fully mineralized and that is it generally resistant to post-mortem isotopic alteration. A study by Jørkov and colleagues, 2 however, highlighted the point that the otic capsule of the petrous portion of the temporal bone is also a suitable proxy for stable isotope studies of the early dietary patterns similar to enamel, as it does not appear to undergo remodelling once fully formed in infancy/ early childhood. After pioneering experimental research by Harbeck et al, 3 another experimental study by Snoeck et al 5 convincingly demonstrated that the process of cremation at high temperatures (calcination) does not alter biogenic 87 Sr/ 86 Sr, and opened up the strontium isotope method to be applied to cremated remains in archaeological contexts. 9,10,[33][34][35] In fact, the first application of strontium isotope analyses to PP by Harvig et al obtained 87 Sr/ 86 Sr in both cremated and inhumed skeletal remains to assess the reliability of the petrous as a sampling element for strontium isotope mobility studies. 4 To our knowledge, this study represents the only study conducted to date to measure 87 Sr/ 86 Sr in unburnt PP. 4 Owing to the high degree of correlation between (premolar) enamel and petrous 87 Sr/ 86 Sr signatures obtained from the same individuals, for both cremated and inhumed remains, the authors concluded that biogenic strontium is retained in archaeological petrous samples and is indicative of childhood origins. 4 However, although their study apparently included two individuals of non-local origins, most of the sampled individuals were of local origins. Additionally, premolar enamel crowns develop and mineralize from roughly 2 to 6 years of age after the remodelling of the otic capsule is believed to cease, and thus it is difficult to assess if differences in 87 Sr/ 86 Sr between petrous bone and premolar enamel samples reflect age-related changes or diagenesis of the former.
Furthermore, their study reported the range of bioavailable 87 Sr/ 86 Sr at the national level (Denmark), whereas the local (site) bioavailable 87 Sr/ 86 Sr ranges were not reported. These issues are interpretively problematic as strontium isotope ratios obtained from non-cremated bones may correlate with those obtained from dental enamel for local individuals for various reasons: e.g. bone diagenesis, absence of migration, migration between isotopically similar locales, etc. For example, similar enamel and bone (PP) 87 Sr/ 86 Sr ratios obtained from the same individual could simply result from the lack of movement (local origins) and/or diagenetic alteration of bone 87 Sr/ 86 Sr.
In this regard, the present study represents a reassessment of the reliability of strontium isotope ratios obtained from inhumed (unburnt/non-cremated) pars petrosa samples for human palaeomobility research. The research focuses on a single individual recovered from an archaeological context where the local and regional bioavailable strontium isotope ranges are well characterized (n = 50 baseline samples) and whose dental enamel 87 Sr/ 86 Sr deviates substantially from the local range indicating distant (non-local) childhood origins. Combining strontium isotope and strontium concentration analyses of multiple tissues formed at different periods of this individual's life allows us to more robustly assess intraindividual differential diagenesis and to test the reliability of noncremated archaeological PP as a proxy for childhood geographic origins.

| Isotope analysis
Strontium column extraction and sample loading were performed following previously published protocols. 42 The isotope compositions were measured using a Thermo Scientific™ Triton Plus™ instrument housed at the Vrije Universiteit Amsterdam, the Netherlands. The strontium ratios were determined using a static routine and were corrected for mass fractionation to 86 Sr/ 88 Sr of 0.1194. 43 The NIST ® SRM ® 987 Strontium Carbonate reference material gave an average of 0.710258 ± 0.000008 (2s) during the course of this study (n = 77).

| RESULTS
The results of the strontium isotope analysis and concentration measurements are provided in Table 1 and

| DISCUSSION
Since the strontium isotope ratios of all three dental enamel samples are significantly higher than the local range, this suggests that the proposed transatlantic migration occurred after childhood (after formation of the third molar enamel which is completed around 11-16 years of age 45,46 ). However, the 87 Sr/ 86 Sr of the third molar enamel is somewhat elevated relative to that of the first and second molar enamel samples. In a previous study this was tentatively suggested to indicate a possible prior migration event in adolescence, perhaps within Africa. 13 The strontium concentrations in the dentine sample increased between 9% in the M2 (27)  Sr/ 86 Sr. This suggests that these dentine samples contain a mix of biogenic and diagenetic strontium. A slightly different pattern is observed for the third molar dentine sample, which has a Sr concentration that is slightly higher than that of the corresponding enamel sample (27%), but exhibits an 87 Sr/ 86 Sr that is actually more radiogenic than the third molar enamel 87 Sr/ 86 Sr. If a diagenetic shift towards the local signature is assumed, the biogenic third molar dentine 87 Sr/ 86 Sr must have been even more radiogenic than 0.7116. This pattern suggests that the third molar dentine is not highly impacted by diagenesis, and lends further support to the notion that this individual may have undergone a previous migration that occurred subsequent to the mineralization of the third molar enamel and during the formation of the third molar root dentine. As third molar roots generally start to develop during late childhood and finish forming at approximately 20 years of age, this may indicate that this purported earlier migration event occurred as an adolescent or young adult. As both the M3 enamel and dentine 87 Sr/ 86 Sr are highly elevated relative to the bioavailable range for the Antilles, this also suggests that this earlier (forced) migration within the African continent occurred prior to the later forced migration to the Caribbean.
The 87 Sr/ 86 Sr of the three samples collected from the adult female's otic capsule are substantially higher than the bioavailable strontium signature at the location of burial, and broadly comparable to the 87 Sr/ 86 Sr of the enamel samples. In fact, the strontium isotope composition of the three otic capsule samples is nearly identical to that of the third molar enamel (38) and not to that of the first molar enamel (26) as was expected based on the assumed age at which the PP is thought to stop remodelling (Figure 3). A possible explanation T A B L E 1 Strontium isotope ratios ( 87 Sr/ 86 Sr) of enamel, dentine and petrous samples; and strontium concentrations of enamel and dentine samples a To be clear, we are not advocating that the otic capsule, or the pars petrosa, of inhumed remains are ideal sampling targets for strontium isotope analyses of archaeological skeletal remains.
However, the results of this study support the main conclusions of a previous study 4 indicating that biogenic 87 Sr/ 86 Sr may be retained in inhumed pars petrosa in certain archaeological contexts. Although we expect that dental enamel will remain the preferred sampling material for most isotope studies of human palaeomobility, the pars petrosa may provide an alternative sampling element in cases where teeth are not preserved or are otherwise unavailable for analyses, and can potentially be sampled in conjunction with aDNA studies which increasingly target this skeletal element anyway. Although the results of the present study and those of the previous study from Harvig and colleagues 4 seem promising, it is necessary to conduct more research into differential diagenesis in a wider range of archaeological burial contexts before strontium isotope analysis of inhumed pars petrosa can be more broadly applied in archaeological palaeomobility studies.