Human mobility amid emerging sedentism in the Central Zagros: laser ablation strontium isotope analysis at Early Neolithic Bestansur, Iraqi Kurdistan

1 Introduction

The development of sedentary lifeways during the Neolithic (c. 10,000–5,200 BCE) profoundly transformed human-environment relationships in Southwest Asia. This process, linked but not always synchronous with the emergence of domestication and agriculture (Dow and Reed, 2015; Zeder, 2024a), saw highly mobile groups progressively commit to select locales, forming settled communities. Yet, its pace and trajectories, along with food procurement and cultural changes, varied across regions. In particular, the balance between sedentary and mobile behaviors is likely to have been contextually dynamic, driven by needs, choices, and factors such as climate and resource availability (Gregoricka, 2021). Over the last 15 years, the role of human mobility in shaping and sustaining settled communities has been increasingly explored through direct analyses of human remains (e.g. Alt et al., 2013; Santana et al., 2021; Knipper et al., 2023; Pearson et al., 2023; Wang et al., 2023a,b; Plug et al., 2025). However, research has primarily focused on Anatolia and the Levant, leaving the eastern part of the Fertile Crescent comparatively understudied.

The Central Zagros, encompassing the uplands and lowlands of Eastern Iraq and Western Iran, have long been recognized as a theater for independent experiments with domestication and sedentism (Matthews et al., 2020a; Daly et al., 2021; Darabi, 2022; Zeder, 2024a). Additionally, autonomous population development has been highlighted by ancient DNA studies, which have identified Early Holocene individuals from the region as the most genetically distinct in Southwest Asia (Broushaki et al., 2016; Lazaridis et al., 2016; Altınışık et al., 2022). Archaeological, zooarchaeological and archaeobotanical evidence suggests that a major transition from seasonal site occupation to permanent, year-round settlements took place in the Central Zagros between 8,500 and 7,000 BCE (Zeder, 2024a). Despite a general increase in residential stability, it has been proposed that logistical movements for resource acquisition retained their importance long after the adoption of sedentary lifeways, influenced by the slow adaptation of food procurement and cultural practices (Zeder, 2024a), and by the expansion of exchange networks (Watkins, 2023). However, aside from two studies reporting human mobility data for the lowland site of Ali Kosh, in the Iranian Zagros (Darabi et al., 2024; Ávila et al., 2026) and Nemrik, in northern Iraq (Ávila et al., 2026), the lack of systematic research into mobility dynamics in the Eastern Fertile Crescent has limited an in-depth understanding of human movement and its contribution to community make-up during this formative period. The availability of a large osteological assemblage from Bestansur, a major 8th-millennium BCE site in the Iraqi Zagros, offers an opportunity to address this gap using a human-centered approach.

Continued advancements in strontium isotope analysis (⁸⁷Sr/⁸⁶Sr), a well-established tool for the study of mobility (Bentley, 2006; Britton, 2020), have considerably enhanced fine-scale reconstructions of past human movement. Laser ablation multi-collector inductively coupled plasma mass spectrometry (LA-MC-ICP-MS) now enables the rapid acquisition of ⁸⁷Sr/⁸⁶Sr profiles measured at high spatial resolution across sequentially formed dental enamel (Pryor et al., 2020; Linscott et al., 2023). Here, we applied this technique to 101 teeth belonging to 41 individuals from Bestansur, with the aim of investigating life-course mobility and its intersections with sociocultural structures at the transition to sedentism. Human ⁸⁷Sr/⁸⁶Sr profiles are contextualized using a multi-proxy baseline constructed for the site. In addition, we report baseline data for four other archaeological sites in the Central Zagros (two in Iraq and two in Iran; Figure 1, Supplementary Text S1 and Figures S1–S4, S7), contributing to the characterization of ⁸⁷Sr/⁸⁶Sr variability in a region with sparse isotope records.

1.1 Archaeological context

Located in the Sulayimaniyah province of Iraqi Kurdistan, the site of Bestansur (35 °22'36.7 N, 45 °38'44.4 E; c. 550 m asl) occupies a low natural rise with access to a perennial karst spring in the Sharizor plain, on the western foothills of the Central Zagros (Figures 2a, b). Since 2012, it has been investigated by the University of Reading and the Slemani Directorate of Antiquities and Heritage as part of the Central Zagros Archaeological Project (CZAP), marking the resumption of Neolithic research in the region after a long archaeological hiatus (Matthews et al., 2020a; Zeder, 2024a,b). Absolute dating places the site in the Early (Pre-pottery or Aceramic) Neolithic, approximately between 7,700 and 7,100 BCE (Flohr et al., 2020), with a recent radiocarbon date indicating occupation into the early 7th millennium BCE.

Archaeological investigations at Bestansur have been conducted through the excavation of 15 trenches (Supplementary Figure S8), revealing evidence of Early Neolithic occupation over four hectares. These include the remains of permanent mudbrick buildings, work and fire installations, and traces of daily activities (Iversen, 2020a; Matthews et al., 2020b; Richardson et al., 2020). Together with zooarchaeological and archaeobotanical data, these findings suggest year-round habitation, mainly sustained by wild resources from a biodiverse landscape (Bendrey et al., 2020; Matthews et al., 2020c; Whitlam et al., 2020; de Groene et al., 2021, 2023;). The artifactual record of the site is dominated by locally sourced materials, such as clays, stone and freshwater shells from the nearby spring and river, used to produce functional objects and adornments (Matthews et al., 2020d; Mudd, 2020; Richardson, 2020). Materials from distant sources, including marine shells, obsidian and carnelian, are less common but attest to involvement in wide-reaching exchange networks, stretching from the Red Sea to East Anatolia and Iran (Richardson, 2020, 2025).

To date, a minimum of 85 individuals have been recovered from buildings and open spaces across the site (Figure 2c) (Ragazzon, 2025). Mortuary deposits range from single, discrete inhumations to highly commingled, multiple interments, with widespread evidence for the retention, curation and redeposition of skeletal elements (Supplementary Figure S9) (Walsh and Matthews, 2018; Walsh, 2020). The greatest concentration of human remains and objects made from non-local materials has been documented in two elaborate, superimposed buildings, collectively referred to as the Building 5/8 complex (Walsh, 2020). This complex, interpreted as a communal and ceremonial space with mortuary functions, has been linked with social differentiation and the negotiation of collective identities at the site (Matthews et al., 2020d; Richardson et al., 2020)

Variability in mortuary behaviors and the distinctive features of Building 5/8 have led to the hypothesis that Bestansur gathered local and non-local individuals from extended mobility networks. At the same time, zooarchaeological and artifactual evidence suggests that some social segments were periodically mobile, mixing sedentary strategies with logistical forays for resource procurement (Matthews et al., 2020d). The application of laser ablation strontium isotope analysis to human remains from the site offers an avenue for exploring and expanding on these ideas, supporting a more nuanced understanding of community dynamics and mobility in an understudied region.

2 Strontium isotope analysis for mobility reconstruction

Strontium isotope analysis relies on the relationship between ⁸⁷Sr/⁸⁶Sr ratios in dental enamel and those of the local geology at the time of enamel formation (Ericson, 1985). Radiogenic ⁸⁷Sr, a decay product of the parent isotope rubidium-87 (⁸⁷Rb), varies relative to stable ⁸⁶Sr depending on the age and original rubidium content of rocks (Bentley, 2006; Britton, 2020), resulting in different isotopic signatures across different geological substrates. Because of the long half-life of ⁸⁷Rb (48.8 billion years) (Faure and Mensing, 2005), these signatures remain essentially stable over archaeological time scales. With bedrock weathering, soluble strontium is transferred to soils, water reservoirs and the biosphere (Capo et al., 1998; Flockhart et al., 2015; Britton, 2020). Its bioavailable fraction is then incorporated into the forming enamel of animals and humans as a substitute for calcium, mainly via gastrointestinal absorption from diet and drinking water (Johnson et al., 1966; Apostoaei, 2002). Along this pathway, ⁸⁷Sr/⁸⁶Sr ratios do not undergo significant fractionation, remaining mostly unaltered from those of the geological source (Bentley, 2006). Therefore, by comparing ⁸⁷Sr/⁸⁶Sr ratios in dental enamel against baselines for specific areas, it is, in principle, possible to draw inferences about the provenance and mobility of individuals (Montgomery, 2010; Bataille et al., 2020). In practice, however, the level at which movement patterns can be interpreted is contingent upon factors requiring careful consideration. These include the composition of the diet, which may bias ⁸⁷Sr/⁸⁶Sr ratios toward the origin of high-strontium foods (Lahtinen et al., 2021; Linscott et al., 2023), and the degree of isotopic heterogeneity within the study region.

Establishing baselines that adequately reflect ⁸⁷Sr/⁸⁶Sr variation in given environments is essential for mobility studies. While bioavailable strontium pools are primarily determined by local lithology, processes like soil mixing, differential mineral weathering, and atmospheric and hydrological inputs can introduce isotopic variability (Bataille et al., 2020; Holt et al., 2021; Spies et al., 2025), which is then transferred to animals through plants and water. To account for these complexities, the selection of baseline archives should be informed by the environmental context of the investigated area (Holt et al., 2021). In general, it has become increasingly common to integrate multiple environmental and faunal proxies to capture broader and more representative ⁸⁷Sr/⁸⁶Sr ranges (e.g. Al-Shorman et al., 2025; Eckelmann et al., 2025; Goodarzi et al., 2025). Building on these datasets, isotope distribution maps (isoscapes) can further aid mobility tracing and spatial assignment across larger areas. However, the suitability of their use depends on the scale of the study, and is greatest in data-rich regions, where the spatial modeling of ⁸⁷Sr/⁸⁶Sr ratios is supported by dense background sampling (Bataille et al., 2020).

Because dental enamel forms incrementally and does not remodel after the enamel matrix matures into mineralised tissue (Lacruz et al., 2017; Hillson, 2024), it preserves a broadly sequential record of ⁸⁷Sr/⁸⁶Sr ratio variability during tooth growth. Spatially resolved micro-sampling gives insights into these changes, enabling the detection of fine-scale biogeochemical variations and thereby enhancing temporal resolution beyond what is achievable through bulk strontium isotope analyses. However, interpretations are complicated by an incomplete understanding of physiological variables, including strontium residence time, its mobilization and recycling in the body, and the precise chronology of enamel maturation (Pryor et al., 2020; Boethius et al., 2024; Hillson, 2024; Spies et al., 2025). These factors have raised questions about the extent to which ⁸⁷Sr/⁸⁶Sr ratios in teeth may be averaged over time. Recent experimental studies have begun to address these concerns, prompting refinements in micro-sampling strategies. In particular, applications of LA-MC-ICP-MS to animals and humans with known movement histories have demonstrated good alignment between laser ablation data and actual mobility patterns (Lazzerini et al., 2021; Boethius et al., 2022; Le Corre et al., 2023; Yang et al., 2025), supporting its increasing use in mobility studies (e.g. Linscott et al., 2023; Kurila et al., 2025; Vaiglova et al., 2025; Zazzo et al., 2025; Wathen-Avila et al., 2025). While techniques such as Thermal Ionization Mass Spectrometry (TIMS) and solution-based multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) are reported to yield higher precision (Spies et al., 2025), LA-MC-ICP-MS has proven especially effective in capturing short-term movements when enamel is ablated along the growth axis, from cusp to cemento-enamel junction (Boethius et al., 2024). Targeting enamel regions adjacent to the dentine-enamel junction, which mineralise first and are therefore less temporally averaged than other parts of the crown (Müller et al., 2019), has been proposed to further improve temporal resolution. The time- and cost-efficiency of LA-MC-ICP-MS also facilitates the analysis of multiple sequentially forming teeth per individual (Pryor et al., 2020, 2024a,b), supporting life-course approaches to mobility reconstruction.

At Bestansur, multi-tooth laser ablation strontium isotope analysis was employed to investigate mobility from infancy to late adolescence. In the absence of regional isoscapes or detailed bioavailable strontium isotope records for the Central Zagros, ⁸⁷Sr/⁸⁶Sr data are integrated with multi-proxy baseline characterization for the site, supplemented by archaeological and bioarchaeological evidence to support interpretations.

3 Materials and methods

3.1 Materials – sediment samples

Sediments analyzed for this study originated from archaeological archive samples kept at the University of Reading and collected during the excavation of Neolithic sites investigated by CZAP in 2008 (Jani and Sheikh-e Abad; Kermanshah Province, Iran), and between 2012 and 2023 (Bestansur, Shimshara and Zawi Chemi Rezan; Sulaymaniyah Province, Iraq). Based on material availability, at least two samples were taken from each site. All archive samples were derived from sealed Neolithic deposits, except for one from Jani and one from Shimshara (JAN2 and SHI1, respectively). These had originally been removed as sediment blocks from sections of the archaeological mounds exposed by water erosion. In the other cases, sample selection targeted clay sediments from occupational layers. An overview of sampling locations and context types is provided in Supplementary Table S6.

3.2 Materials – faunal dental samples

Seven archaeological animal teeth were selected from the Iraqi sites of Bestansur, Zawi Chemi Rezan and Shimshara, complementing sediment analysis for the definition of bioavailable strontium baselines (Supplementary Table S5). Sampling was dictated by material availability but preferentially targeted rodents. Indeed, small mammals with restricted feeding ranges and shorter lifespans are recognized to be better proxies for local strontium bioavailability on circumscribed spatiotemporal scales (Bentley, 2006; Holt et al., 2021; Spies et al., 2025). Samples included three mouse (Mus) teeth for Bestansur, and one hare (Lepus) and two rabbit (Oryctolagus) teeth for Zawi Chemi Rezan. For Shimshara, a single sheep/goat (Ovis/Capra) tooth was available for analysis. The seasonal movement and status of ovicaprids at the site are not well established, and evidence for potential management is inconclusive (Bendrey et al., 2020). The inadequate understanding of exploitation regimes and feeding ranges for these herbivores, be they wild, managed or fully domesticated, complicates the a priori interpretation of their ⁸⁷Sr/⁸⁶Sr ranges as local. Indeed, these may reflect averaging of ratios incorporated over large areas (Bataille et al., 2020; Holt et al., 2021). Such interpretative uncertainties were here counterbalanced through the integration of the animal proxy with sediment-derived data.

3.3 Materials – human dental samples

The human dataset included 101 permanent teeth belonging to 41 individuals from Bestansur. To explore variability in provenance and mobility patterns, individuals originating from articulated, disarticulated and mixed, commingled mortuary deposits were sampled from different burial locations across the site. All deposits can be dated to the Neolithic occupation of the site, with most sampled individuals dating back to the mid- to late 8th millennium BCE, and two individuals ascribable to a slightly later chronological horizon (C2387 SK1 and C2387 SK2; early 7th millennium BCE). A plan linking the original location of individuals to available radiocarbon determinations is shown in Figure 3.

Based on previous osteological analyses (see Ragazzon, 2025 for methodological details), sampled individuals ranged from 5 to 45+ years of age and were classified as older children (5.6–10.5 years; n = 4), adolescents (10.6–19 years; n = 9), young/middle adults (20–40 years; n = 20) and mature adults (40+ years; n = 8) according to their midpoint age estimate. Biological sex estimates were available for one older child, two adolescents and 23 adults. The older child (C1812 SK9) and one of the adolescents (C1228 SK1) were genetically determined to be male and female, respectively (Lazaridis et al., 2022), while the sex of the second adolescent, aged 16–18 years, was osteologically estimated to be male (C2387 SK2). Due to poor skeletal preservation, confidence in estimates for adults varied considerably, often exhibiting a probabilistic component. Overall, 14 individuals could be classified as females or probable females (F/F?), and nine as males or probable males (M/M?). In six additional cases, biological sex was marked as indeterminate or unknown (ND). For the purpose of this study, females were pooled with probable females, and males with probable males. It is important to stress that the expression “biological sex” is used here to distinguish aspects of anatomical and/or genetic variation from socially constructed gender identities (British Association for Biological Anthropology and Osteoarchaeology, 2022). While biological sex estimates serve as a proxy for accessing diverse experiences in the past, they are employed with an awareness of the limitations of bioarchaeological methods, which do not fully capture the broad range of human variation.

To support a life-course approach to the exploration of mobility, sample selection fell on the first (M1), second (M2) and third (M3) molars, teeth whose sequential formation spans infancy to adolescence (AlQahtani et al., 2010). Individuals preserving M1, M2 and M3s were preferentially sampled, as this allowed for greater temporal depth in analysis. Indeed, based on median development stages, crown formation takes place approximately between 4.5 months and 3.5 years for M1s, 2.5 and 8.5 years for M2s, and 8.5 and 14.5 years for M3s (AlQahtani et al., 2010), although M3s are reportedly characterized by greater developmental variability (Hillson, 2024). A few individuals presenting only one (M1 or M2) or two (e.g. M1 and M2 or M2 and M3 pairs) of the selected teeth were also included to increase representativeness. Further sampling criteria are detailed in Supplementary Text S3, while a complete list of human dental samples is provided in Supplementary Table S7.

3.4 Preparation and analysis of sediment samples using TIMS

The preparation and analysis of sediment samples were carried out at the National Center for Isotope Geochemistry at University College Dublin (Dublin, Ireland). Eleven sediment samples of approximately 8.5 g each were leached in 30 ml of ultrapure water (18.2 MΩ-cm) (Linscott et al., 2023) and agitated three times a day for 15 days. After filtration of the leach solutions, two 1 ml aliquots were collected for each sample. One aliquot was run on a Thermo Scientific iCAP Q quadrupole ICP-MS instrument to measure strontium concentrations. The second aliquot was further processed to determine ⁸⁷Sr/⁸⁶Sr bioavailability.

Aliquots to be analyzed for ⁸⁷Sr/⁸⁶Sr were transferred to columns with ~100 μl of SR-Spec resin (Triskem, France), and strontium was separated following procedures modified after Pin et al. (1994). Collected strontium samples and blanks were dried on a hot plate until full evaporation of the liquid fraction (~24 hours), before dilution in 10 μl of 0.05 M HNO3 and loading of 1 μl onto a rhenium filament (99.99%, H.Cross) with 1 μl of TaCl5 activator solution. Strontium isotope measurements were performed on a Thermo Scientific Triton Thermal Ionization Mass Spectrometer on Faraday cups in static mode with amplifier rotation. Instrument performance and analytical accuracy were monitored with the SRM987 isotopic standard, which yields ⁸⁷Sr/⁸⁶Sr = 0.710266 ± 0.000022 (2σ, n = 21). The ratio recorded on the day of the analytical session was 0.710264. Mass fractionation was corrected with an exponential law normalized to ⁸⁶Sr/⁸⁸Sr = 0.1194. Interference from ⁸⁷Rb on ⁸⁷Sr was monitored using ⁸⁵Rb and was negligible in all cases. Contribution from a procedural blank of ~2,200 pg was also negligible.

3.5 Preparation and analysis of dental samples using LA-MC-ICP-MS

Sample collection and preparation for human and animal teeth was carried out in the chemistry laboratory at the Department of Archaeology and History, University of Exeter (Exeter, UK). Following the protocol described in Pryor et al. (2020), slices of enamel and adhering dentine of approximately 2–3 mm in thickness were cut from dental crowns, from the cusp to the cemento-enamel junction, using a diamond-tipped circular saw. This approach keeps destruction to a minimum and enables the preservation of most of the tooth for future analyses. Rodent teeth were cut either longitudinally or, for very small teeth, transversely, to create slices exposing a cross-section through the enamel. Both human and faunal dental slices were embedded in Epofix© epoxy resin then sanded with 35 μm and 15.3 μm grit papers until attainment of a flat, polished surface.

The measurement of strontium isotope ratios took place at the National Oceanography Center, University of Southampton (Southampton, UK), using LA-MC-ICP-MS, with a Thermo Scientific Neptune MC-ICP mass spectrometer coupled to a New Wave 193-nm Ar-F excimer laser ablation system (UP193FX). The laser track was set to target enamel close to the enamel-dentine junction, from cusp to cementoenamel junction. Additionally, dentine was analyzed along randomly placed 1–4 mm lines in a subsample of five human teeth to cross-reference ratios with those obtained from sediments and faunal samples. The enamel and dentine targeted for analysis were cleaned with the laser, and data were then collected using the following analytical settings: laser beam diameter at 110 μm, 15 Hz repetition rate and either 10, 15 or 20 μm/s⁻¹ translation rate depending on the length of the profile being measured. The ablated material was transferred from the laser cell to the MC-ICP-MS using a He carrier gas which was then mixed with Ar and N₂ before injection into the plasma ion source. ⁸⁷Sr/⁸⁶Sr measurements were taken in static collection mode with a set integration time of 1.049 s, and the instrumental setup was tuned for the minimization of oxide production, monitored as ²⁵⁴(UO)+/²³⁸U+, following Lewis et al. (2014).

After the analysis, masses were corrected to an on-peak gas blank for ⁸⁶Kr interference. Subsequently, ratios underwent correction for mass fractionation using an exponential law (Russell et al., 1978) normalized to ⁸⁶Sr/⁸⁸Sr = 0.1194 (Nier, 1938). Rare earth element contamination was assessed using yttrium-89 (⁸⁹Y) as a proxy, with datapoints presenting anomalously high ⁸⁹Y concentrations discarded as diagenetically altered (Woodhead et al., 2005). The isobaric interference of ⁸⁷Rb on ⁸⁷Sr was also corrected to the natural ⁸⁷Rb/⁸⁵Rb ratio of 0.385617 (de Laeter et al., 2003). Calcium dimers and argides in carbonate materials are known to have only a negligible effect on measured ⁸⁷Sr/⁸⁶Sr ratios, however their impact on ⁸⁴Sr/⁸⁶Sr ratios is significant (Woodhead et al., 2005). We therefore corrected ⁸⁴Sr/⁸⁶Sr ratios using mass 82 for the ⁴⁰Ca (or ⁴⁰Ar)⁴²Ca+ molecular ion, followed by a mass balance correction using the natural abundance ratios for calcium (Wieser et al., 2004). This enabled the use of the reported ⁸⁴Sr/⁸⁶Sr ratios as a quality control against the accepted value of 0.0565 (Woodhead et al., 2005). For each analytical session, accuracy and reproducibility were tested through repeat measurement of an in-house enamel standard obtained from pigs fed a controlled marine diet (Lewis et al., 2017), with ⁸⁷Sr/⁸⁶Sr pre-characterized to 0.709076 using thermal ionization mass spectrometry (TIMS). Repeat analyses of the standard across three analytical sessions (n = 29) gave a mean of 0.709127 ± 0.000193 (2σ), equating to a mean positive offset of 71 ± 273 parts per million (ppm; 2σ) for LA-MC-ICP-MS over TIMS.

3.6 Sampling rights and permissions

Sediment, animal and human samples from Iraq and Iran were exported and analyzed with permission from the Slemani Directorate of Antiquities and Heritage (Kurdistan Regional Government, Iraq), and the Iranian Center for Archaeological Research (Iranian Culture, Heritage and Tourism Organization, Iran), respectively. All research activities on organic and inorganic materials from the sites included in the study were conducted with the authorization of the relevant excavation directors and approved by the host institutions.

4 Results

4.1 Local baselines for Bestansur and other Central Zagros sites

In this study, the concept of “local” was defined as encompassing Bestansur and its primary source area for water, food, and construction materials, namely the karst spring and river catchment in which the site is located, as well as the surrounding alluvial plain up to the Zagros foothills, within a radius of approximately 2.5–3 km (Figure 2b). While this spatial definition is relatively narrow, it reflects the current distribution of baseline samples and, based on contextual evidence, it should be regarded as a conservative minimum estimate of the area regularly used by individuals for basic resource procurement. A local baseline of bioavailable strontium was established by combining ⁸⁷Sr/⁸⁶Sr ratios from archaeological sediment leachates from the site (0.70794–0.70812; n = 3), rodent enamel (0.70800–0.70819; n = 2) and dentine also from the site (0.70798–0.70815; n = 2) (Supplementary Tables S2, S5–S6), and published modern plants collected from the mound and surrounding fields (0.70815–0.70816; n = 3; Supplementary Table S1) (Elliott et al., 2015). Human dentine samples yielded consistently higher ⁸⁷Sr/⁸⁶Sr ratios (0.70832–0.70854; n = 4) and were thus excluded from baseline calculations (see expanded discussion in Supplementary Text S2, 3, Figures S5, 6 and Table S3).

Based on the minimum and maximum ratios covered by all baseline proxies, the lower and upper bounds of the local ⁸⁷Sr/⁸⁶Sr range were set at 0.70794 and 0.70819, respectively. In terms of surficial geology, all microfaunal and environmental samples originated from the archaeological mound and the surrounding fields, which lie on a Cretaceous clastic substrate, overlain by Quaternary alluvial deposits in places. The baseline range is therefore likely to reflect mixed inputs from Quaternary alluvium, Cretaceous clastic rocks and pelagic limestone (Saeed Ali, 2007; Elliott et al., 2015), with groundwater from the spring contributing averaged ⁸⁷Sr/⁸⁶Sr signals from multiple Cretaceous-age geological formations (Shiranish, Tanjero, Kometan and Balambo) (Saeed Ali, 2007). Ratios fall largely within the 0.7076–0.7082 range reported for groundwater samples from 16 water wells within a maximum radius of 70 km from the site (Figure 2a), where small ⁸⁷Sr/⁸⁶Sr variability has been attributed to the dominant influence of Cretaceous carbonate dissolution (Mahmmud et al., 2022). Strontium isotope ratios reported for two modern plant specimens sampled on the limestone foothills of the Zagros, approximately 3 km away from Bestansur, overlap with the site baseline, corresponding to 0.70795 and 0.70803 (Elliott et al., 2015).

Fewer archival proxies were available for the other Central Zagros sites (Supplementary Tables S5, 6). In the Iraqi Zagros, sediment leachates (n = 2) and sheep/goat dentine (n = 1) from Shimshara, which is located on a conglomerate outcrop capped by Quaternary alluvium, yielded a narrow range of 0.70790–0.70801. This expands slightly to 0.70781–0.70824 when enamel ⁸⁷Sr/⁸⁶Sr ratios from same ovicaprine tooth are considered. At Zawi Chemi Rezan, which sits on an Eocene carbonate bedrock, sediment samples (n = 2), and rodent enamel (n = 3) and dentine (n = 1) produced ratios between 0.70803 and 0.70832. In the Iranian Zagros, sediment leachates yielded ⁸⁷Sr/⁸⁶Sr ratios of 0.70822 ± 0.00003 and 0.70825 ± 0.00002 (n = 2) for Jani and 0.70829 ± 0.00002 for Sheikh-e Abad (n = 2), both located on Quaternary alluvial deposits (Supplementary Figure S7). Overall, variation is low across sites, and ratios fall within indicative ⁸⁷Sr/⁸⁶Sr ranges for the isotopic zones proposed by Lü et al. (2023), namely the Iranian plateau (0.7075–0.7090) and the Iraqi foothills of the Zagros (0.7080–0.7085).

4.2 Laser ablation ⁸⁷Sr/⁸⁶Sr data for Bestansur human enamel

Sequential ⁸⁷Sr/⁸⁶Sr profiles for 101 human teeth (38 first, 38 s and 25 third molars) representing 41 individuals from Bestansur are presented in Supplementary Figures S10–13. Contextual information about sampled teeth and summary statistics for the entire dataset are provided in Supplementary Tables S4 and S7, respectively. Strontium isotope ratios for dental enamel range from 0.70773 to 0.70856.

Enamel ⁸⁷Sr/⁸⁶Sr profiles show strong alignment with the site baseline across the dataset. Intra-tooth ⁸⁷Sr/⁸⁶Sr variability, calculated as the range between maximum and minimum ratios, is low across the sample, spanning 0.00012 to 0.00068 (Figure 4). The most homogenous profiles are observed in the molars of two mature adult (probable) females, C2286 SK1 and C3002 SK1, while the largest variation is noted in the second molar of a young/middle adult female, C1788 SK1 (Figure 5). Comparisons by biological sex and burial location reveal similar intra-tooth ⁸⁷Sr/⁸⁶Sr variability between the teeth of females (n = 34; 0.00012–0.00068) and males (n = 25; 0.00021–0.00058), and those of individuals interred within Building 5/8 (n = 59; 0.00021–0.00068) and at other locations (n = 42; 0.00012–0.00052).

Despite general consistency with the baseline, repeated intra-tooth fluctuations toward more and less radiogenic ratios are observed across several ⁸⁷Sr/⁸⁶Sr profiles (Figure 5). At the same time, a few teeth fall completely or largely above the local range. These include the first and second molars of a mature adult of indeterminate sex (C2632 SK1) and the second molar of a young/middle adult female (C2909 SK3). In individual C2909 SK3, a shift toward higher ratios is already apparent in the last third of the ⁸⁷Sr/⁸⁶Sr profile for the first molar. Two additional individuals, C2286 SK1 and C3002 SK1, have profiles plotting mostly above the baseline, though they cluster very close to its upper bound. Except for the first molar of C2632 SK1, all teeth display at least some overlap with the local range. Although mobility remains a plausible interpretation for individuals showing the highest and lowest ratios, profile deviations from the range defined as “local” fall largely within the 2σ analytical uncertainty of laser ablation measurements. Their interpretation as non-local signals therefore cannot be made with confidence.

All individuals with unworn first molars (n = 6), where enamel loss is negligible, and which are therefore expected to best capture the earliest phase of cusp mineralisation, exhibit ratios that are broadly consistent with, or only slightly more radiogenic than the local range in the initial section of their ⁸⁷Sr/⁸⁶Sr profiles. The terminal sections of profiles from non-adult teeth whose crowns were still forming at the time of death and are thus likely to reflect their last place of residence (n = 7), also fall within the baseline or cluster close to its upper bound. Beyond C2909 SK3, there is no evidence for marked changes in ⁸⁷Sr/⁸⁶Sr catchments between the mineralisation of the first, second and third molars, that is, between infancy, childhood and adolescence, in individuals with multiple teeth available for analysis.

5 Discussion

5.1 Strontium isotope variability and the definition of local baselines in the Central Zagros

Environmental and faunal baseline proxies from five Neolithic sites in the Central Zagros yielded ⁸⁷Sr/⁸⁶Sr ratios between 0.70794 and 0.70832, largely within the predictive range proposed by Lü et al. (2023) for the Zagros foothills. While samples for Shimshara, Zawi Chemi Rezan, Jani and Sheikh-e-Abad were limited in number, offering only a preliminary characterization of local strontium bioavailability, their analysis indicates low inter-site ⁸⁷Sr/⁸⁶Sr variability. This is partly due to similarities in geolithological substrates within a landscape dominated by carbonates. However, the small range of ⁸⁷Sr/⁸⁶Sr ratios even where surficial geologies apparently differ, as at the Iraqi sites, suggests an additional influence of specific environmental factors and geomorphic processes. Proximity to water and extensive alluviation are correlated with differential weathering and the mixing of strontium from multiple sources (Bentley, 2006), potentially resulting in dampened ⁸⁷Sr/⁸⁶Sr variability. Although groundwater is generally expected to have a limited impact on bioavailable strontium pools (Holt et al., 2021), in the Zagros, where karst aquifers provide one of the main water sources (Mustafa et al., 2016; Mahmmud et al., 2022), it likely plays a non-negligible role. At Bestansur, for example, local isotopic averaging may have been shaped by the position of the site within the catchment of the perennial spring and spring-fed river, which supplied water, clay for construction and fertile soils for plants exploited during the Neolithic. This apparent homogeneity, however, should not be assumed to characterize the whole region, and particularly non-settled areas where mixing processes may be less pronounced. In the Sulaymaniyah province, geolithological variation is seen within a minimum radius of 20 to 50 km from the sites analyzed (Figure 2a), and some degree of ⁸⁷Sr/⁸⁶Sr variability may be expected across yet-unsampled substrates.

Based on the marine strontium isotope curve, which reflects ⁸⁷Sr/⁸⁶Sr variation in seawater over the geologic time scale (McArthur et al., 2020), ratios dropped to minima below 0.7070 during the Middle/Upper Jurassic and steadily increased above 0.7085 from the Middle Miocene. The ⁸⁷Sr/⁸⁶Sr ratios of marine carbonates dating to these periods can thus be expected to mirror negative and positive shifts in the curve slope, influencing bioavailable ⁸⁷Sr/⁸⁶Sr ranges. When taking Bestansur as a reference point, Middle/Upper Jurassic and Neogene geologies are found ~20 km to the north-east and ~25 km to the west, respectively. Further distinctive ⁸⁷Sr/⁸⁶Sr ranges may be found in areas where carbonates interact with volcanic rocks. One such area is the Mawat district, ~100 km to the north, where whole-rock ⁸⁷Sr/⁸⁶Sr ratios of 0.7057–0.7067 are reported for mafic rocks and andesites (Koyi, 2009; Aswad et al., 2014). The Penjween district, ~40 km north-east of the site, may also exhibit different ⁸⁷Sr/⁸⁶Sr ranges due to the contribution of mafic and felsic rocks (⁸⁷Sr/⁸⁶Sr = 0.7031–0.7074) (Abdulzahra et al., 2018; Azizi et al., 2025) to bioavailable strontium pools. While this study had to rely on existing archival collections for the analysis of environmental proxies, additional fine-scale sampling will be required to refine the characterization of ⁸⁷Sr/⁸⁶Sr distribution across the region.

5.2 Community make-up, residential stability and logistical mobility at Bestansur

Laser ablation strontium isotope analysis at Bestansur provides new insights into human mobility and the emergence of sedentism in the Central Zagros. When considering the first molars (n = 38), which mineralise during infancy, a single mature adult of indeterminate sex (C2632 SK1) had a ⁸⁷Sr/⁸⁶Sr profile plotting entirely above the baseline (0.70825–0.70840), consistent with a possible non-local origin. Two mature adult females (C2286 SK1 and C3002 SK1) also exhibited ⁸⁷Sr/⁸⁶Sr profiles largely above the baseline, though repeatedly overlapping with the local range. For all three individuals, deviations from the baseline fall within the 2σ analytical uncertainty for laser ablation measurements, making their identification as possible non-locals tenuous. All other teeth displayed broad alignment with the baseline, indicating a predominantly local make-up of the Bestansur community. Because the sample includes individuals spanning childhood to mature adulthood, this pattern of locality may have extended across multiple generations.

Adherence to the baseline from the first through the third molars further suggests commitment to the site throughout non-adult life. A single young/middle adult female (C2909 SK3) may have spent part of their childhood in a different geological setting, as suggested by an extended shift in strontium catchments between the first and second molar. However, the deviation from the baseline in the second molar is inconsistent across the ⁸⁷Sr/⁸⁶Sr profile, indicating either intermittent interactions with the site or movement within areas characterized by similar bioavailable strontium pools. Overall, the close correspondence between ⁸⁷Sr/⁸⁶Sr ratios and the local range in unworn first molars and molars still forming at the time of death is compatible with the individuals having been born, lived, and died at the site or in its immediate surroundings. The paucity of mobility data for the Central Zagros and the Eastern Fertile Crescent limits the contextualization of Bestansur within broader regional frameworks. At the contemporary site of Ali Kosh, in the Iranian Zagros, strontium isotope data suggest greater mobility (Darabi et al., 2024). Conversely, individuals from Nemrik, in the Upper Tigris valley, have provisionally been interpreted as local based on strontium and oxygen isotope analyses (Ávila et al., 2026). Mobility reconstructions at these sites are currently constrained by the preliminary nature of investigations and small sample sizes (seven individuals at Ali Kosh and two at Nemrik; Darabi et al., 2024; Ávila et al., 2026). Nevertheless, they point to some contextual variability in mobility patterns during the 8th millennium BCE, highlighting the need for larger datasets to support robust interpretations of regional trends in human movement.

At Bestansur, no clear patterns of ⁸⁷Sr/⁸⁶Sr variability can be identified based on biological sex, burial location or type of mortuary deposit (e.g. articulated vs. disarticulated; single vs. multiple individuals). Comparable isotope studies focusing on Neolithic Southwest Asia report no clear associations between provenance and mortuary treatment, attributing this to processes of integration and constructed affiliation that rendered non-locals and locals archaeologically indistinguishable (Pearson et al., 2023; Wang et al., 2023b; Plug et al., 2025). The picture emerging from the site broadly agrees with this interpretation, suggesting that mortuary practices were shaped by roles, statuses and identities acquired during life, and/or by mourning behaviors negotiated by the living (Croucher, 2012, 2018), rather than by the individuals' origins.

At present, enamel ⁸⁷Sr/⁸⁶Sr data is not sufficient to support the identification of Bestansur as a supra-local aggregation center drawing people from the wider region, although convergence from isotopically indistinguishable landscapes cannot be excluded. When integrated with archaeological evidence, ⁸⁷Sr/⁸⁶Sr data indicate ties between the analyzed individuals and the site locales that transcend short-term place connections. Hallmarks of sedentism include the construction, maintenance, and renovation of permanent buildings, as well as continued mortuary activities binding the sphere of the living to that of the dead (Matthews et al., 2020b; Richardson et al., 2020). The presence of synanthropic rodents linked to food storage and waste-management practices, and the seasonality of exploited animal taxa, including birds, can be interpreted as additional indicators of year-round habitation (Tchernov, 1991; Bendrey et al., 2020). Residential stability at Bestansur was likely sustained by increased water availability during the Early Holocene (Regattieri et al., 2023; Rostami et al., 2024) and by resource accessibility in a favorable environment. In this context, ⁸⁷Sr/⁸⁶Sr data from the site align with archaeological evidence for the establishment of increasingly permanent settlements in the region during the 8th millennium BCE (Zeder, 2024a). Beyond the Zagros, the results of this study echo isotope analyses from other sites across Southwest Asia, which also reveal predominantly local community make-ups and strong commitment to places during the Early Neolithic (Alt et al., 2013; Benz et al., 2016a,b; Santana et al., 2021; Pearson et al., 2023; Wang et al., 2023b).

Alongside the increasing adoption of sedentary lifeways, the period spanning the 9th to the 7th millennium BCE is associated with an intensification of cross-regional interactions in Southwest Asia (Watkins, 2023). The material culture of Bestansur reflects such connections: objects made of obsidian, carnelian and marine shells point to contacts stretching from the Sinai to Anatolia and Iran, and from the Persian Gulf to the Red Sea and the Eastern Mediterranean (Richardson, 2020, 2025). The exact modalities through which these networks were maintained, however, are difficult to reconstruct. One reason is that extended mobility undertaken by adults who grew up at Bestansur, left the site and later returned to be buried would remain invisible in the strontium isotope record, which only captures non-adult residence. Although some children may have participated in extended mobile behaviors, as possibly suggested for individual C2909 SK3, evidence for this is weak. Interpretations are also complicated by overlaps between the ⁸⁷Sr/⁸⁶Sr range for the site and those of many contemporary contexts across Southwest Asia (Santana et al., 2021; Pearson et al., 2023; Wang et al., 2023a,b; Darabi et al., 2024; Giaccari et al., 2025; Plug et al., 2025), limiting the ability of biogeochemical analyses to confidently detect episodes of long-distance mobility. Notably, however, non-local materials are far less common than utilitarian and decorative objects made with resources from the nearby springhead and Zagros limestone outcrops (Matthews et al., 2020d; Mudd, 2020; Richardson, 2020). Estimates of obsidian movement into the site likewise indicate relatively modest volumes of traffic, not supporting intensive or highly organized exchanges (Matthews et al., 2020d). Instead, these data appear more consistent with episodic interactions, in which materials reached the site through multi-step circulation and/or mobility by select individuals, rather than large-scale engagement in long-distance movements.

While cross-regional mobility in Southwest Asia may be challenging to explore using ⁸⁷Sr/⁸⁶Sr data, repeated intra-individual fluctuations in ⁸⁷Sr/⁸⁶Sr ratios within and across teeth require consideration. Except for two mature adult females with highly homogeneous profiles (C2286 SK1 and C3002 SK1), reflecting consistent strontium sourcing during infancy and childhood, several individuals exhibit ⁸⁷Sr/⁸⁶Sr ratios that oscillate above and below the local baseline, spanning approximately 0.70773–0.70856. Because the timing of enamel mineralisation, strontium residence in the body and dietary equilibration all act to average ⁸⁷Sr/⁸⁶Sr ratios (Pryor et al., 2020; Boethius et al., 2022), these fluctuations are indicative of marked and/or prolonged shifts in dietary strontium sourcing, sufficient to produce detectable variation in enamel profiles. These may be linked with the exploitation of bioavailable strontium pools within and, possibly, outside of the local catchment, as consistent with episodes of short-term mobility away from the site. Lower ⁸⁷Sr/⁸⁶Sr ratios are compatible with contributions from ubiquitous Upper Cretaceous carbonates, which have reportedly low whole-rock ratios (0.70764–0.70796 for dolomites in the Zagros basin) (Mansurbeg et al., 2021), as well as from Middle/Upper Jurassic carbonates, which dominate the landscape 20 km north-east of Bestansur. Conversely, endmembers with higher ⁸⁷Sr/⁸⁶Sr ratios are more difficult to pinpoint without more extensive characterization of the regional isoscape. These may tentatively be identified with Middle Miocene and Pliocene substrates located ~25 km to the west, in the biodiverse area of the Qara Dagh mountains. It must be emphasized, however, that the observed fluctuations in ⁸⁷Sr/⁸⁶Sr profiles are small, limiting confidence in inferring mobility beyond the site locales. Interpretations are further constrained by the lack of dense background sampling in the area surrounding Bestansur and the broader region, which makes it difficult to exclude the impact of minor isotopic heterogeneity within homogenous geological settings. Additional fine-scale baseline mapping will be required to better understand micro-scale strontium isotope variability and refine the provisional interpretations presented here.

Ethnoarchaeological research suggests that when residential mobility decreases in foraging communities, logistical mobility increases to ensure continued access to a wide pool of resources (Binford, 1980; Kelly, 2013). At Bestansur, zooarchaeological and archaeobotanical evidence indicates strong reliance on wild foods within a hybrid, low-level food production economy (Bendrey et al., 2020; Iversen, 2020b; Whitlam et al., 2020; de Groene et al., 2021, 2023), implying that logistical movements may have been periodically required to meet the needs of the community. Preliminary carbon and nitrogen stable isotope analyses of human bone collagen delineate a C₃ terrestrial diet (n = 7; δ¹³C = −19.9 ± 0.2‰; δ¹⁵N = 7.1 ± 0.5‰). When compared to available baseline values for pigs (n = 4; δ¹³C = −20.0 ± 0.6‰; δ¹⁵N = 6.1 ± 0.6‰), as well as medium-large herbivores, including sheep/goats (n = 3; δ¹³C = −19.3‰; δ¹⁵N = 5.4‰), deer (n = 1; δ¹³C = −19.7‰; δ¹⁵N = 6.8‰) and cattle (n = 1; δ¹³C = −20.3‰; δ¹⁵N = 5.9‰), human δ¹⁵N values reflect trophic offsets well below the 3–5‰ expected for the regular consumption of these animals (Bocherens and Drucker, 2003). This suggests that animal protein from these species contributed only minimally to human diet. Conversely, the high frequencies of caries recorded at the site are consistent with reliance on plant foods rich in fermentable carbohydrates and, possibly, starches (Ragazzon et al., in preparation). In terrestrial ecosystems, plants contain higher strontium concentrations than meat (Sillen and Kavanagh, 1982; González-Weller, 2013), meaning that the ⁸⁷Sr/⁸⁶Sr ratios of the consumer will be biased toward the ratios of plant foods. Zooarchaeological, micromorphological and artifactual evidence also demonstrates the consumption of molluscs and freshwater fish at Bestansur (Bendrey et al., 2020; Iversen, 2020b; Mudd, 2020), both resources that can be rich in strontium (Schroeder et al., 1972; Schoeninger and Peebles, 1981; Lahtinen et al., 2021). Although plant foods may have offered lower caloric returns than large game, their accessibility and storability likely made them dietary staples, complemented by the regular exploitation of aquatic resources. As major contributors to human ⁸⁷Sr/⁸⁶Sr ratios, these foods would have made enamel ratios particularly sensitive to shifts in foraging locales, especially since plants more closely reflect local strontium bioavailability (Bentley, 2006).

While strontium isotope data are not, by themselves, conclusive, their integration with contextual information about the site and its economy enables cautious interpretations. In particular, it can be speculated that intra-individual fluctuations in ⁸⁷Sr/⁸⁶Sr profiles may reflect non-adults periodically accessing resources from different foraging locales, possibly through active involvement in logistical forays. Although their spatial extent cannot be determined, these activities could have reasonably expanded the resource base available at Bestansur and in its immediate catchment. Participation may have included children accompanying caregivers on foraging activities during the breastfeeding period and early childhood, and later acting as independent food collectors (Lew-Levy et al., 2017). In modern hunter-gatherer groups, adults and children often operate within coordinated and partially overlapping foraging radii, with non-adults fulfilling a substantial portion of their own food requirements (Kelly, 2013). This scenario is broadly consistent with bioarchaeological evidence for the participation of children in subsistence and occupational activities at the site (Walsh, 2022; Ragazzon, 2025). Taken together, biogeochemical and contextual data may therefore be compatible with the inhabitants of the site engaging in targeted logistical movements to secure resources, leaving subtle isotopic traces of a mobile foraging component within otherwise sedentary lifeways. Given the analytical limitations of the dataset, these reconstructions remain tentative; nevertheless, they offer a plausible perspective on patterns of mobility and resource use.

In summary, laser ablation strontium isotope analysis delineates the Bestansur community as largely local, with multi-generational ties to the site. Since excavations are still ongoing, earlier, more mobile phases may remain unidentified. Nevertheless, biogeochemical data, architectural continuity and markers of year-round habitation indicate residential stability during the 8th millennium BCE. While a strong commitment to the site may have been maintained at least throughout non-adult life, it is likely that mobility retained a key role. Small quantities of imported materials suggest episodic cross-regional exchanges, while subtle shifts in ⁸⁷Sr/⁸⁶Sr profiles across several analyzed teeth might reflect the exploitation of different foraging locales. Based on current knowledge of the economy of Bestansur, foraging activities would have contributed to sustaining an early settled community that was still largely reliant on wild plants and animals. By integrating isotopic data with contextual evidence and ethnoarchaeological parallels, this study offers new insights into emerging sedentism and community structure in the Early Neolithic Central Zagros. Beyond Bestansur, it contributes to the characterization of regional bioavailable strontium isotope variability, offering a framework for further discussions of mobility, landscape use and associated interpretative challenges in Early Neolithic Southwest Asia.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fearc.2026.1818214/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Michelle S. Eusebio, University of the Philippines Diliman, Philippines

Reviewed by: George Kamenov, University of Florida, United States

Berta Morell, University of Barcelona, Spain