Use of strontium isotope ratios in geolocation of Guatemalan population: Potential role in identification of remains

Abstract Within post‐conflict communities, attempts to identify and repatriate unidentified and missing individuals poses a difficult task. As current forensic strategies commonly lack the capacity to provide region of origin assessments, forensic anthropologists/investigators are often unable to identify sources of DNA for kinship analysis. Using Thermal Ionization Mass Spectrometry (TIMS), hair samples from 10 volunteers were used to assess the variation in strontium isotope ratios (87Sr/86Sr) between extant people in Guatemala City and Coban; with a leach (external) and digest (dietary) signal analyzed for each sample. A two‐way anova demonstrated that the difference between 87Sr/86Sr of Guatemala City and Coban was statistically significant (F [1, 16] = 259.839, p < 0.05), with no statistically significant differences observed between leach and digest 87Sr/86Sr (F [1,16] = 4.319, p = 0.054). Overall, individuals from Coban demonstrate 87Sr/86Sr comparable to previously recorded baseline values, demonstrating a minimal change in diet which is reflected in associated surveys. Volunteers from Guatemala City, however, show a marked shift in 87Sr/86Sr away from predicted values highlighting the potential influence of imported goods. The results here highlight the applicability of 87Sr/86Sr in hair to serve as a potential tool to support the identification of unknown individuals in Guatemala in a forensic context.


| INTRODUC TI ON
The discourse of identification and repatriation in Guatemala has been a reoccurring theme for the country's forensic expertise, due to a history of the displacement of individuals through migration or enforced disappearances. This trend has been observed since the 1960s whereby Guatemala's internal conflict, fought until 1996 between governmental forces and guerrilla groups mainly drawn from indigenous populations, saw approximately 500,000-1.5 million individuals fleeing the country or internally displaced. Despite estimates that 200,000 people were killed or disappeared, of those identified, the Mayan population represents more than 80% of documented victims with the Ladino population (those of mixed European and indigenous heritage) making up ~17%, who have become collectively known as 'Los Desaparecidos' or 'The Missing' [1][2][3].
Today, Guatemala's modern 'Missing' are characterized predominantly by individuals who disappeared as a result of migration efforts, spurred by political and socioeconomic marginalization with individuals often seeking refuge, by migrating to the United States.
Despite individuals from Guatemala (in addition to El Salvador and Honduras) making up more than 52% of apprehensions at the US-Mexico border [4], many individuals do not reach their intended destination, as a result of criminal interference or through succumbing to the harsh environment during their journey. When the remains of 'Missing' individuals are found, forensic intervention is required to help provide an identification.
Forensic strategy dictates that identification efforts are focused on the utilization of DNA. However, the use of DNA relies on collecting appropriate familial reference samples, whereas region of origin assessments can prove pivotal in both focusing DNA efforts and excluding potential alternative identities. In Guatemala, while region of origin assessments may be possible by visual assessment of traditional indigenous clothing that is distinct amongst different regions within the country [5], such assessment is subjective and the success and indeed presence of such artifacts relies heavily on burial conditions [6]. The use of region of origin assessments not only serves as a strategy to stratify potential identities and to focus DNA efforts but also has a cultural significance, with proper burial ensuring a restful transition into the afterlife in accordance with traditional Mayan beliefs [7].
Isotopes offer the potential to track the migration of individuals, through tissues such as hair, teeth, and bone, by representing region of origin at various stages during life. In particular, the strontium isotope ratio 87 Sr/ 86 Sr is used frequently in geolocation studies as it differs geographically due to heterogeneity in bedrock geology and atmospherically derived sources [8]. These environmentally controlled isotopes are readily available in the body's hard tissues where the resulting strontium isotope ratio represents an accumulation of the ratios from areas where dietary consumables are obtained [9,10]. In the body, intestinal absorption of strontium results in the manifestation of strontium in the body's hard tissues (ie bones and teeth) where it readily substitutes for calcium in the apatite of both structures [11,12]. Teeth, in particular the enamel, provides region of origin 87 Sr/ 86 Sr during childhood, with the first permanent molar often chosen for 87 Sr/ 86 Sr analysis, where the mineral formation, and therefore strontium incorporation into the enamel of that tissue, ceases approximately within the first 3.5 years of childhood [13]. Alternatively, bone samples such as femoral cortical bone, through remodeling, reflect the last 10-15 years of life [14]. While the comparison of 87 Sr/ 86 Sr in bone and teeth offers the ability to identify migration from childhood residency, the inability to remodel and resistance to diagenesis means tooth enamel is the preferred tissue for region of origin assessments [15].
Ideally teeth would be the focus for sample collection and tissue of choice when analyzing the remains of missing individuals.
However, sampling an extant population is not often feasible, due to its invasive nature and issues regarding donation, particularly in countries such as Guatemala, where Mayan regions often rely on traditional 'teeth pullers' [16] or international teams/volunteers working remotely to provide dental care [17], rather than a central dental practice. In this scenario, hair can serve as a proxy to identify regional differences in 87 Sr/ 86 Sr from human populations.
Unlike strontium in bone and teeth, which is found within the crystalline hydroxyapatite structure, strontium ions (Sr 2+ ) in hair are bound as a result of the negative charge associated with peptides through ionic bonds [18]. As a tissue which exists externally, hair is influenced by strontium from two sources firstly, an internal source, the 87 Sr/ 86 Sr in blood where trace metal exchange occurs until subcutaneous formation prior to exposure from the scalp [19].
Secondly, it is influenced by external strontium sources from the environment, such as bathing water, aerosols and dust [20]. Strontium is found particularly in the internal portion of hair, such as the cortex and medulla but also in the external structures, such as the cuticle, where metals, such as calcium, adhere to the assumption that the chemical similarity between strontium and calcium is the cause [21].
The exogenous 87 Sr/ 86 Sr influence on hair is demonstrated through an increasing concentration in 87 Sr/ 86 Sr from the proximal to distal portions and through transverse cross sections [19], highlighting the permeability of hair as a result of the narrow overlap between cells in the cuticle layer [22].
Hair provides a more recent indicator of provenance or travel history, with 1 month of residency represented by ~1 cm of hair growth [23]. Such physiological factors allow researchers to establish spatial and temporal information through the sequential analysis of hair segments from the tip to the root, tracking migration [24]. Alternatively, bulk analysis of hair is often used in studies when such distinction is non-essential and provides an accumulative 87 Sr/ 86 Sr signal [19,25,26]. The use of hair provides a number of additional benefits in the determination of bioavailable strontium from an extant population, in that it is readily available, collection/sampling is non-invasive, and it offers reportedly, the potential to discern between environmental and dietary sources of strontium by leaching and digesting the hair in a suitable acid and analyzing the resulting (leachate and digest) solutions separately [19]. With strontium isotopes demonstrating a lack of fractionation, unlike light isotopes such as oxygen, hydrogen, nitrogen and carbon [8], comparative strontium isotope ratios should be reflected in the hair, teeth and bone of a sedentary individual.
Despite this, there is a consensus that samples analyzed for forensic purposes should be compared to isotopic reference values of the same type, ie, bone to bone, teeth to teeth [27]. However, understandably unlike hair, such samples are often unobtainable or impractical to collect [27].
Much of the current understanding on bioavailable strontium in Guatemala is based on archeological populations [28][29][30][31][32][33][34][35]. With self-sufficient populations such as the Maya, dietary strontium, and subsequently bioavailable strontium in the community, would be expected to reflect the 87 Sr/ 86 Sr ratios of the underlying geology in which they reside. However, dietary preferences and the availability of imported goods can alter the expected bioavailable 87 Sr/ 86 Sr outside those reported in their archeological ancestors [25,36]. As a result, for this study, the use of modern hair samples for the geolocation of a more contemporary population was chosen given the complexity of the modern globalized diet.
This study therefore represented an exploration of the potential to use hair samples as a means of geolocation. Two populations were chosen for this study with distinct underlying geology, which may in turn be reflected in strontium isotope intake through consumption of locally grown produce.   Table 1.

| Hair preparation -digest
The cleaned hair was digested using 10 ml of Teflon distilled 8 M HNO 3 on a hotplate at 100°C and allowed to evaporate to dryness.
The sample was converted to chloride form by adding 5 ml of 6 M HCl and evaporating to dryness.
Once dissolved and taken up in calibrated 2.5 M HCl, the Sr was separated from both the leachate and the hair samples using an Eichrom© AG50-X8 cation resin. The strontium isotope composition was determined by Thermal Ionization Mass Spectrometry (TIMS) using a Thermo Triton© multi-collector mass spectrometer.
Samples were loaded on to single rhenium filaments using TaF activator following the method of Birck (1986) [39]. Samples were run in peak jumping mode for 100 scans and to an internal precision of ≤0.00001 (2 SE). Strontium procedural blanks were ~50 pg. The reproducibility of the international standard NBS987 was ±0.00006 (2SD, n = 46) over the period of analysis. All data were corrected to an NBS 987 standard 87 Sr/ 86 Sr value of 0.710250.

| Statistical analysis
Statistical analysis was completed using SPSS v25. Data were tested for conformity of requirements of parametric analysis prior to use of two-way anova to assess the effects of location, nature of sample and their interaction.

| RE SULTS
When assessing the data in Table 2, it can be seen that individuals from Coban exhibit comparatively higher 87 Sr/ 86 Sr in leach (0.7069-0.7076, mean = 0.7073 ± 0.0003) and digest F I G U R E 1 A map of the Maya region from Thornton [35] adapted from Hodell et al with associated data [28] (Elsevier©), Guatemala is highlighted using a bold outline from the southern lowlands to the apex of the Pacific coast. Reproduced with permission from Elsevier [28,35]. From the ratios provided in Table 2, a two-way anova demon-

| DISCUSS ION
Overall, this research has focused on ten hair samples from Guatemala City and Coban with five samples collected from each location. Using 87 Sr/ 86 Sr, Guatemala City and Coban are distinguishable (p < 0.05); however, crossover exists between leach and digest values for the same location (p < 0.054), which is reflected in Figure 2. It is important to note that whilst the statistical analysis TA B L E 1 Details of isotopic interest from individuals whose samples were analyzed for 87 here demonstrates the potential of how strontium isotopes can be used and exploited, the statistical analysis is based on a limited dataset, with a minimum of 30 samples often suggested [15]. Although such differences in 87 Sr/ 86 Sr may seem negligible numerically, within the context of the instrumental capabilities, such differences are large isotopically, with the TIMS instrument demonstrating a precision of ±0.000006 for strontium in this study [15].
The ability to distinguish between Coban and Guatemala City on the basis of 87 Sr/ 86 Sr would be expected given the underlying geology. In terms of consumables, whilst heterogeneity existed in the water consumed during the day, the majority of respondents in both populations confirmed that municipal water was utilized in the home.
However, contrary to expectations, the majority of individuals from Guatemala City recorded that locally grown food was consumed (100%), when compared to Coban where the majority of individuals were unable to confirm if local food was consumed (with bioavailable 87 Sr/ 86 Sr in individuals influenced, as a key factor, by the origin of foodstuffs) [40].

Based on the values provided by Hodell et al, individuals from
Coban as part of the Alta Verapaz region in the Southern Lowlands, consuming a diet predominantly consisting of local produce, would be expected to exhibit an internal 87 Sr/ 86 Sr spanning from 0.7070-0.7085 [28] with the city founded on Cretaceous limestone deposits.
Guatemala City located within the traditional Southern Highlands, founded on Tertiary and Quaternary volcanic deposits, would be expected to exhibit an 87 Sr/ 86 Sr range between 0.7038 and 0.7050 [28]. Overall, when the leach and digest 87 Sr/ 86 Sr for Coban are considered, the results gained in this study fall within the range offered in the literature [28][29][30][31][32][33][34][35]. that modern indigenous diets show 76% similarity to the diets eaten by individuals during the 1960, from foods which could be sourced locally [43]. Comparatively, it was found that only 72% of such foods contribute to the modern Ladino diet [43]. While the impact of population shows an interesting tendency, the capability to distinguish between Mayan and Ladino diets at the same location would offer greater clarity in assessing the impact of population for this study.
However, despite apparent population differences, the presence of elevated 87 Sr/ 86 Sr, away from locally defined ratios in Guatemala City, is most likely due to a combination of access to imported goods along with socio-economic factors [44,45]. The higher proportion of Ladino individuals in the capital, often characterized as more economically prosperous, resulted in supermarkets such as Walmart, who import 85% of produce, originating from within the countries' capital to grow exponentially from the 1980s (with the supermarket sector commanding 32% of Guatemala's food market as early as 2002) [44]. Importantly, therefore, whilst respondents from Guatemala stated that food was local in origin, it is likely that this was assumed based on geographical proximity to its area of purchase.
As the data collected in this study represent a modern popula-  [44,46].
In this assessment, indigenous communities are often only afforded access to small informal retailers, with the majority of foods either being sourced locally, via supplementation from subsistence farming or through markets [44,47]. Although in regards to the latter, modern food networks increasingly mean food at such markets originate from across the country [48]. However, with a diet more

| CON CLUS ION
Overall, this study has demonstrated that the use of 87 Sr/ 86 Sr in modern human hair has the potential to discern between areas