Human Mobility in the Central and Eastern Mediterranean during Hellenistic and Roman Times: The Potential and Limitations of Bioarchaeological Research

A review of bioarchaeological research on Hellenistic and Roman mobility in Central and Eastern Mediterranean yielded 52 results, including academic papers, book chapters, excavation reports, conference presentations, and graduate theses. The map of Figure 1 depicts the sites that have been examined in the context of the above research, and the following sections will briefly present selected studies to demonstrate the development of research in this region from the 1980s to the present day, as well as highlight areas where further work is needed. 3.1. Early Studies (1980s–1990s) All studies included in this review, which date back to the 1980s, focused on Jewish remains. These studies primarily examined cranial metric and/or cranial and postcranial nonmetric variation and provided detailed descriptive tables and comparisons with other sites. However, an explicit biodistance analysis was not conducted. For example, Arensburg and Smith [43] studied the Jewish population of Jericho from 100 BCE to 70 CE. They provided a comprehensive description of the remains, including information on demography, pathology, stature, and craniometrics. The craniometric data is particularly relevant to paleomobility studies. The authors presented tables with summary statistics for various cranial measurements in Jericho, as well as comparative assemblages from Hellenistic and Roman period sites in Israel. They found that there was a general similarity in cranial shape among Second Temple populations, but differences compared to earlier and later period remains. However, they did not discuss the implications of these patterns for the mobility and connectivity of the Hellenistic-Roman groups. Similarly, Smith and Zias [44] provided craniometric data on a Late Hellenistic assemblage from the French Hill Tomb in Jerusalem, along with demographic and pathological information. Although they did not explicitly conduct a biodistance analysis, they compared their craniometric data descriptively to other contemporary assemblages in Israel and found an overall similarity.The final two studies conducted in the 1980s had a specific focus on metric and nonmetric variability, rather than considering it as part of a broader bioarchaeological study. Arensburg et al. [45] examined the metric variability present in Jewish cranial morphology during the Hellenistic, Roman, and Byzantine periods. While they did not estimate the biodistance between the different groups studied and relied solely on descriptive statistics, the authors explicitly concentrated on the population history of the Jewish people and concluded that there has been a significant level of similarity and stability over several centuries. Similarly, Goldstein et al. [46] directed their attention towards Jews from the Hellenistic and Roman periods, but this time focusing on cranial and postcranial nonmetric traits. Their objective was descriptive in nature, providing extensive tables displaying the frequency of various traits. However, they did not conduct any biodistance analyses, despite acknowledging the potential of such data as indicators of biological affinity. In the 1990s, scholarly research in the field continued to be predominantly descriptive, with a primary emphasis on compiling detailed tables of craniometric data. However, there was a noticeable shift towards incorporating statistical analysis and utilizing various biodistance measures. Furthermore, studies began to encompass a wider geographical scope within the region of interest. A representative example is Billy’s [47] study, wherein a significant resemblance in cranial morphology and associated affinity was observed between the population inhabiting the Roman era Douch in the Kharga Oasis, Egypt, and urban populations residing in the Nile valley. This finding was based not only on a descriptive evaluation of similarities and dissimilarities, but also on the statistical examination of the morphological diversity in Egyptian and Nubian groups. Another study of this period that utilized biodistance data in conjunction with historical and archaeological evidence is by Manzi et al. [48]. The study focused on dental metric and nonmetric traits from two distinct assemblages: one from the rural town of Lucus Feroniae, primarily inhabited by slaves and war veterans, and the other from the Isola Sacra necropolis at Portus Romae, representing an urban “middle class” group. The findings revealed greater variability in dental dimensions and morphology in Lucus Feroniae, indicating a higher level of heterogeneity compared to the assemblage from Portus Romae. This pattern was attributed to the unique population compositions and histories of the two sites, with the authors linking them to human mobility and genetic admixture during both pre-Roman periods and Rome’s military and political expansion. In the same direction of using biodistance analyses to examine specific and testable research hypotheses but focusing on issues with a much broader time span, Rubini et al. [49] examined the population history of ancient Sicily spanning the 2nd and 1st millennium BCE, with a focus on the impact of Greek colonization. By studying cranial nonmetric traits in archaeological assemblages from eastern Sicily, their findings indicated that the initial biological influence of Greek settlers in the region may have originated in the Bronze Age. Even during the 1990s, there have been studies that have provided significant descriptive data that could be used in biodistance analysis, but have not been utilized for this purpose. For instance, Brin et al. [50] presented craniometric data for Hellenistic Jews from Ein Gedi and compared them to other groups, yet they did not conduct biodistance analysis and instead interpreted their findings in relation to environmental factors rather than issues of mobility and ancestry. Similarly, Henneberg et al. [51] provided craniometric, odontometric, and osteometric data for a skeletal collection from the House of C Iulius Polybius in Pompeii, dated to 79 CE, but did not further analyze or discuss them. 3.2. Later Studies (2000s–2010s) The number of Hellenistic-Roman biodistance studies increased notably after 2000. Similarly, the geographic span covered by paleomobility studies expanded, as did the methodological repertoire used, with an explosion in the use of isotopic analysis, and a limited use of aDNA analysis, which was much more common for prehistoric assemblages. Biodistance studies continued to be utilized extensively in these more recent periods, incorporating both traditional and advanced statistical methods. Joel Irish has been among the pioneers in this direction and his study of population affinities in ancient Egypt is a representative example [52]. By examining dental morphological traits in Egyptian populations spanning from the Neolithic era to the Roman period, Irish calculated phenotypic similarities using various biodistance measures. The findings of his study indicated a population continuity over time. Besides Irish’s work, North Africa received a lot of attention in terms of biodistance analyses during that period. In their study, Nikita and colleagues [53,54] utilized cranial nonmetric traits and geometric morphometrics to investigate patterns of gene flow between the Garamantes, a civilization that thrived in the Sahara Desert from 900 BCE to 500 CE, and other populations in North Africa. The comparative groups included, among others, individuals from the Ptolemaic and Roman periods in Alexandria, as well as the Punic and early Roman inhabitants of Carthage. The findings indicated that the Garamantes displayed distinct physical characteristics, suggesting a degree of biological isolation despite their participation in trans-Saharan trade networks. On the other hand, the individuals from Alexandria and Carthage were found to be closely related, along with Soleb and Algerians, indicating movement along the Mediterranean coast and the Nile River. Haddow’s [55] thesis is another example of a biodistance study focused on North Africa, examining the population structure of Ismant el-Kharab (ancient Kellis, Dakhleh Oasis, Egypt) during the late Ptolemaic to late Roman period. By recording dental nonmetric traits on individuals from Kellis and comparing them to groups from various regions in Africa, Haddow’s study challenged previous anthropological and archaeological interpretations and suggested that the Kellis population had limited genetic exchange with neighboring communities. Another doctoral dissertation focusing on Egyptian biodistances was by Kindschuh [56] and investigated the population structure of the late Roman to early Christian el-Hesa community. Located on the border between Egypt and Nubia, el-Hesa was expected to exhibit significant gene flow and thus a diverse population. The craniometric analysis supported this hypothesis, revealing a phenotypically diverse population with connections to both Egyptians and Nubians. A final biodistance study in North Africa examined Nubians from the Mesolithic era to the Christian period [57]. This study identified extensive interactions between Romans and Nubians, which the authors believed contributed to the high levels of gene flow observed in the biodistance patterns. Beyond North Africa, the Near East has also been the focus of biodistance studies during this period. One study examined the affinities of the Roman population in Datça-Burgaz to Anatolian groups dating from the Early Bronze Age to the early 20th century. This study utilized nonmetric cranial traits and revealed complex patterns of population continuity and discontinuity over a significant period of time [58]. Another study with a wide temporal scope investigated population affinities in the middle Euphrates valley from the Early Bronze Age to modern times using dental nonmetric traits [59]. The findings indicated continuity until recent times, when northern Mesopotamia was depopulated due to the Mongolian invasion in the 13th century and later repopulated by Bedouin tribes in the 17th century, leading to changes in the genetic makeup of the region. A final biodistance study focused on dental nonmetric traits from the Hellenistic and Roman French Qumran collection. The Qumran collection was compared to Near Eastern groups from various time periods, as well as a sample of Native Americans. The biodistance analysis identified the Native American sample as an outlier, as expected, while the samples from the Near East showed relatively close relationships. However, the authors of the study correctly noted that the sample size from Qumran was too small to draw definitive conclusions [60]. In Italy, Hens and Ross [61] examined the biodistance among three imperial Roman assemblages: Isola Sacra and Velia, primarily inhabited by middle-class tradesmen and merchants, and Castel Malnome, where freed slaves and heavy laborers resided. Using geometric morphometrics to analyze cranial morphology, the researchers found significant differences in cranial shape between Isola Sacra and Velia, which they attributed to the strong Greek influence in southern Italy. In contrast, Castel Malnome did not show significant differentiation from the other sites, likely due to the diverse geographic origins of the freed individuals in this community. Another noteworthy biodistance study on Italian archaeological remains from this era focused on the skeletons from the Passo Marinaro (Kamarina) Classical-Hellenistic cemetery, as part of a larger biocultural investigation. Sulosky Weaver [62] compared the frequency of cranial and dental nonmetric traits between the Passo Marinaro site and other skeletal collections from Greece, Sicily, and Southern Italy. While the dental traits did not provide conclusive evidence of gene flow at Passo Marinaro, as they were common in both Italian and mainland Greek populations, the cranial nonmetric traits indicated a close connection between Passo Marinaro and other Greek populations in Sicily. Additionally, two crania from the sample displayed morphological characteristics consistent with individuals of sub-Saharan African descent. Two biodistance studies conducted in Cyprus and Greece on a regional level merit attention. Harper and Tung [63] estimated the biodistance between Hellenistic–Roman individuals at Mağara Tepeşi and Frankish–Venetian individuals at Athienou-Malloura and found a lack of population continuity in the region. The authors associated this discontinuity with natural selection, genetic drift, and the influx of a new population into the Malloura Valley between the Roman and Venetian periods. In Greece, Nikita and colleagues [64] examined the Archaic to post-Roman population in Akraiphia, Boeotia. By analyzing biodistances using dental nonmetric traits, the researchers identified a biological affinity among Archaic, Classical and Hellenistic individuals, as well as between Roman and post-Roman individuals, but divergence between these two groups. The discontinuity observed between the earlier (Archaic, Classical, Hellenistic) and later (Roman, post-Roman) phases of the cemetery aligns with the Pax Romana program initiated by Augustus, which promoted unrestricted movement, thus increased gene flow during Roman times. Despite advancements in biodistance studies with more sophisticated methods and specific research hypotheses, there are still instances in the literature where descriptive studies using single traits are used to draw conclusions about past population history and structure based on relative frequencies [65]. Additionally, there are studies that highlight “exceptional” traits within broader bioarchaeological analyses without discussing their significance [66]. Some publications also present frequencies of nonmetric traits, recognizing their limited utility for biodistance analysis currently but suggesting potential for future research [67]. In addition to biodistance analyses, paleomobility studies conducted in the 2000s and 2010s increasingly utilized isotopic data. The methods employed varied in complexity, including the use of single or multiple chemical elements, bulk or incremental analyses, and the examination of different skeletal or dental tissues with varying rates of remodeling. For example, one study focused on strontium isotopes in third molar enamel to investigate human mobility in Late Roman/Byzantine Barsinia, Northern Jordan, finding that all individuals were likely local to the site [68]. Similarly, strontium isotopes were used to investigate the presence of political refugees at Khirbet edh-Dharih after the Roman annexation of the Nabataean Kingdom in the 2nd century CE, with only one definite non-local individual identified [69]. Bulk strontium isotope analysis was also used to study mobility in Roman and Byzantine Hierapolis, Turkey, revealing the presence of non-local individuals in the Byzantine population linked to the site’s history of pilgrimage, while the preceding Roman population showed more homogeneity and lacked distinct nonlocals [70]. An early paper of this period examined oxygen and nitrogen isotope ratios to study migration patterns at the Kellis 2 cemetery in the Dakhleh Oasis, Egypt, around 250 CE [71]. The findings revealed the presence of at least two individuals who were not from the local area, possibly originating from the Nile Valley or Nubia. Notably, one individual, who exhibited distinct isotopic values, was found to have suffered from leprosy, leading the authors to suggest that this individual may have been an exile. Additionally, all individuals with higher oxygen isotopic values were male, indicating their likely involvement in caravan trade. A more complex isotopic study from this period focused on oxygen isotope ratios in first and third molar teeth of individuals buried in the necropolis of Isola Sacra near Rome, Italy, during the 1st to 3rd centuries CE [72]. Analysis of the sample revealed that approximately one-third of the individuals had first molar values that fell outside the range observed in the modern data, suggesting a significant number of nonlocal individuals, which aligns with historical records. Furthermore, the study showed that many individuals had migrated during their childhood, before the third molar crown had completely formed. This finding indicates that migration involved entire families rather than solely adult males. Analysis of the isotopic values from the third molars also indicated that a quarter of the sample had moved to Portus sometime after their childhood or adolescence. The nonlocal individuals were likely from regions located to the East and North of Rome, with one individual possibly originating from North Africa. This pattern of a substantial number of nonlocal individuals originating from various regions is consistent with the multinational nature of port cities like Portus. Another important rather early isotopic study from Italy was Killgrove’s [73] PhD thesis. Among other analyses, the author examined the mobility patterns in the two Imperial assemblages from Casal Bertone and Castellaccio Europarco. Strontium isotope analysis applied to 105 individuals identified two nonlocals in Castellaccio Europarco and five at Casal Bertone. Oxygen isotope analysis was performed on a subsample of 55 individuals and identified 12 nonlocals at Casal Bertone and four in Castellaccio Europarco. The nonlocal individuals identified in the study exhibited isotopic values indicative of various regions within and outside of Italy. More recently, oxygen and strontium analyses of a small sample (n = 6 for strontium and 4 for oxygen) from Castellaccio Europarco dating to the Republican period revealed the presence of one individual identified as a likely nonlocal through strontium isotopes, as well as two individuals potentially categorized as nonlocals based on oxygen isotopes [74]. Also in Italy, Milella et al. [75] utilized a multi-isotope methodology incorporating strontium, oxygen, nitrogen, and carbon to examine the relationship between funerary practices, geographical provenance, and dietary patterns in Bologna from the 1st to the 4th centuries CE. The findings indicated a lack of discernible correlation between these factors, implying that the diversity in funerary practices was influenced by multiple heterogeneous elements and did not correspond to varying geographic origins. Finally, this period also witnessed the combined use of different bioarchaeological proxies of paleomobility. Stark’s [76] thesis examined mobility patterns at the imperial Roman sites of Isola Sacra and Velia, and the Rue Jacques Brel Necropolis in Saintes, France. The study involved the analysis of strontium and oxygen isotope ratios alongside nonmetric characteristics. The isotopic data revealed the presence of several individuals who were not local to the sites, including females and children [77,78]. Additionally, biodistance analysis indicated a general similarity in the biological background of the populations at the three sites, likely due to the widespread movement of individuals to and from Rome. However, there were also distinct biological populations present at each site, underscoring the diversity of the Roman population. Another interesting multi-proxy study also focused on Italian material. Prowse et al. [79] utilized a combination of isotopic and DNA analysis on Roman Vagnari to ascertain the origins of the inhabitants at the site. Results from oxygen isotope analysis on 23 individuals indicated the presence of six nonlocals, while genetic analysis on a subset of ten individuals revealed ancestral connections to populations from sub-Saharan Africa and Asia in two cases. In a subsequent investigation conducted by Prowse [9], an additional 30 individuals were studied and the relationship between migrant status and burial offerings was examined in order to identify any potential disparities in the treatment of nonlocal individuals. The results of oxygen isotope analysis showed that 91% of the sample were likely native to the Vagnari region, confirming the findings of the initial study in 2010 that the labor force at this imperial estate primarily consisted of local residents rather than imported slaves or freedmen from other regions of the Roman Empire. The nonlocal individuals were of varying age and sex. Significantly, the nonlocals were interred with similar grave goods as those found in the tombs of local residents, and the positioning of their graves in the cemetery indicated no segregation from the rest of the community. Emery’s [80] thesis and subsequent publications [81,82] performed additional isotopic and ancient DNA research at Vagnari. The mitochondrial DNA data suggested a closer genetic affinity of Roman Italians with Neolithic, Bronze Age, and Armenian Iron Age populations in western and central Eurasia, compared to Iron Age Italians, Ptolemaic and Roman Egyptians. Most of the Vagnari population appeared to be local, with only two individuals belonging to an eastern Eurasian haplogroup, indicating maternal migration. Additionally, isotopic data revealed a small percentage (7%) of non-local individuals at the site. 3.3. The Past Five Years (2020–2024) In recent years, starting from 2020, there has been a sustained focus on human mobility in the Hellenistic-Roman period, with ongoing utilization of diverse methodologies and increased exploration of geographical areas that had previously been overlooked. Biodistance studies have increasingly focused on addressing specific research inquiries rooted in historical data. For instance, Cilli [83] conducted a study on the influence of the Roman road network on facilitating connectivity and gene flow in Abruzzo, Italy. By analyzing biodistances among five populations spanning from the 9th century BCE to the 12th century CE, the author revealed a close biological affinity among these groups, indicating a shared ancestry among the local populations in Abruzzo. Furthermore, the study demonstrated that populations residing near the primary Roman communication routes exhibited higher levels of admixture compared to those situated farther away from the road system, suggesting greater genetic isolation. Thus, the research underscored the significance of the Roman road network in enhancing connectivity in a region characterized by challenging terrain, such as the Apennines mountain range. Another notable biodistance study examined the influence of political integration of Phoenician coastal cities into the Roman Empire. Mardini et al. [84] utilized dental metric and nonmetric traits to estimate biodistances at both individual and group levels among the cities of Tyre, Beirut, and Byblos. The objective of the study was to determine whether the colonial status of Tyre and Beirut, in contrast to Byblos, was associated with increased population diversity and mobility. The findings demonstrated that Tyre and Beirut exhibited greater internal diversity, as indicated by greater intra-assemblage biodistances. Furthermore, the inter-assemblage biodistances revealed a close relationship between Tyre and Beirut, while Byblos appeared as an outlier in the dendrogram. This lend further support to the connectivity between the two coloniae and/or suggested that the nonlocals in these cities shared a similar origin. In contrast, Byblos appeared to have a predominantly local population, which distinguished it from the colonies. These results align with literary sources, which state that Beirut and Tyre, as prominent Roman colonies and commercial centers in the Eastern Mediterranean, experienced higher levels of mobility compared to other coastal cities, like Byblos. The use of isotopic studies has continued to expand. A recent study also focused on Beirut, and combining strontium and oxygen isotope analysis, was by Kalenderian and colleagues [85]. The authors analyzed nineteen individuals. Despite Beirut’s status as the initial Roman colony in the Near East, the isotopic findings identified only two potential newcomers and two definite migrants, including males and females, who came from various regions within the Empire. Additionally, an intriguing discovery was made regarding the absence of correlation between the graves’ material culture and the geographic origin (local or nonlocal) of the individuals, highlighting the fact that the identity of migrants was not expressed through their burial practices. In Greece, a recent isotopic analysis investigated the mobility of 16 individuals from Roman Edessa by examining oxygen isotopes [86]. Through the analysis of various skeletal and dental tissues, the researchers identified intricate mobility patterns in nine individuals who had moved to and from Edessa at different stages of their lives. In nearby modern day Albania, a multi-isotopic study (carbon, nitrogen, oxygen) focused on diet and mobility patterns from the Archaic-Hellenistic period to the Medieval period at two Greek colony sites: Epidamnus/Dyrrachion and Apollonia in Illyria [87]. The study revealed that while individuals from cemeteries outside the walls of Apollonia were local, two thirds of individuals from Epidamnus/Dyrrachion had likely spent their childhood elsewhere, highlighting the latter as a cosmopolitan port city founded by foreigners that attracted immigrants throughout its history. A final isotopic study focused on reconstructing past osteobiographies using a multi-proxy bioarchaeological approach [88]. The authors combined isotopic analyses with dental calculus microscopy on eleven individuals from the Roman cemetery of San Donato and six individuals from the medieval cemetery of San Pietro, both in Lamon. The results indicated a diet consisting of a combination of C3 and C4 plants and high levels of animal protein consumption, with no significant differences between the two populations. The analysis of paleomobility indicated that the individuals likely resided locally or regionally, suggesting a continuity between the Roman and medieval populations. The past five years have also witnessed the more generalized use of ancient DNA analysis in the study of ancestry, and by extension gene flow and mobility, in Hellenistic and Roman assemblages in the region. Some recent studies have focused on mitochondrial DNA, while others have used whole-genome sequencing. For example, Tepgec and Gorgulu [89] analyzed the mtDNA of three Hellenistic individuals from Ayasuluk Hill in Ephesus and found that they belonged to the T2b maternal haplogroup. This haplogroup is believed to have originated in Anatolia during the last Ice Age and spread during the Neolithic period. Italy has been a significant area of interest for aDNA analysis. A study investigating the mitogenome of Umbria in Central Italy examined both contemporary inhabitants of the city and 28 individuals from the necropolis of Plestia, Colfiorito, dating from the early 9th to the late 3rd century BCE. The results suggested a combination of local genetic continuity and inputs from various population sources, including gene flows from central-eastern Europe [90]. De Angelis et al. [91] successfully conducted a whole-genome study on seven individuals from Quarto Cappello del Prete, a rural site in Imperial Rome. The findings indicated a genetic similarity with present-day populations in the Southern Mediterranean and Southern Near East. In a larger-scale study, Antonio et al. [92] examined the whole genomes of 127 individuals from 29 archaeological sites in Rome and neighboring areas, covering a period of 12,000 years. The results supported two major prehistoric ancestral changes, while by the establishment of Rome, the genetic makeup of the region resembled that of modern Mediterranean groups. The Imperial period witnessed immigration as Rome attracted individuals from the Near East and later from Europe, aligning with its geopolitical affiliations and contributing to genetic diversity. In another large-scale study by Lazaridis and his colleagues [93], the genetic history of populations in the Southern Arc from the end of the Bronze Age until modern times was investigated. The research delved into various topics, but with a focus on the Roman era, it was discovered that individuals from the Imperial period of Rome shared genetic ancestry with individuals from Anatolia during the Roman/Byzantine period, whereas earlier Italians had distinct genetic backgrounds. Furthermore, Imperial Romans displayed genetic similarities with pre-Roman Anatolians, while individuals from Rome and its vicinity prior to the Imperial era were genetically distinct. These findings suggest a degree of population continuity in Anatolia, a region that played a significant demographic role within the expansive Roman Empire. This study also recognized the presence of migrants in the Southern Arc during various historical periods, including two individuals from the Roman era discovered in Samsun on the Black Sea. Finally, a study by Salesse and colleagues [94] integrated isotopic analyses, dental morphology and ancient DNA to identify the first documented instance of an African-born migrant buried in a mass grave located in the catacombs of Saints Peter and Marcellinus in Rome. Analysis of the dental morphology data revealed variability within the overall collection, with two individuals exhibiting distinctive characteristics. Isotopic and genetic data subsequently verified the African ancestry of one of these individuals. While the precise circumstances of this individual’s arrival in Rome, whether as a free person or an enslaved individual, remain unclear, this case study provides further evidence of long-distance migration that occurred throughout the Roman Empire and underscores Rome’s cosmopolitan nature and appeal.