Chapter 10: Estimating Ancestry in Human Skeletal Remains

BIOLOGICAL PROFILE

Ancestry is another component of the biological profile.² Legal authorities often wish to have as much information for identification as possible.⁵ This is especially true when a burial is found, and it is suspected that the person was the victim of foul play.⁵ Estimation of ancestry is important not only to assist identification directly, but also as a required precursor to estimating age, sex, stature, and other attributes. Prior knowledge of ancestry for sex estimation, for example, can improve the accuracy of the estimation. Furthermore, missing lists do include a mention to ancestry.³ Therefore, the more information a forensic anthropologist can provide regarding the individual’s physical characteristics, the more he or she can help to narrow the search. As an exercise, create a list of all of the women you know who are between the ages of 18 and 24 and approximately 5＇4″ to 5＇9″ tall. You probably have several dozen people on the list. Now, consider how many females you know who are between the ages of 18 and 24, approximately 5＇4＂ to 5＇9″ tall, and are Vietnamese. Your list is going to be significantly shorter. That’s how missing persons searches go as well. The more information you can provide regarding a decedent’s phenotype, the fewer possible matches law enforcement have left to investigate. This is how ancestry has become an indispensable part of the biological profile.²

Biological ancestry today is often incorrectly labeled as race and generally refers to the individual’s phenotype (outward appearance). Biological ancestry refers to the underlying genetic differences between modern populations. In any other organism/living thing, groups divided according to the biological race concept would be defined as a separate species. The major issue with applying the biological race concept to humans is that there are not enough differences between any two populations to separate on a genetic basis. In other words, biological races do not exist in human populations. However, the concept of race has been perpetuated and upheld by sociocultural constructs of race. The conundrum for forensic anthropologists is the fact that while races do not exist on a biological level, we still recognize and categorize others based on their phenotype. Clearly, our phenotype is an important factor in not only how we are viewed by others but also how we identify ourselves.²

History

These analyses, as well as classification systems involved in ancestry evaluation went through a “dark” period, which can be dated back to 18th century where typological and frequently racist attempts to categorize human variation occurred. The Swedish botanist Carolus Linnaeus (1707–1778) included humans in his broad binomial classification system of plants and animals. Linnaeus gave humans the genus and species Homo sapiens. He also indicated that subdivisions based on geographical variation could be recognized. These subdivisions were classified as African (H. afer), American (H. americanus), Asian (H. asiaticus) and European (H. europaeus). Criteria for this typology were primarily based on impressions of behavior and skin color.³

Subsequently, Johann Friedrich Blumenbach (1752–1840) extended the Linnaeus classification and added detail on features of head anatomy. Blumenbach added a new category for Malayan. The classification systems of both Linnaeus and Blumenbach reflected attitudes of that time that such human variation was relatively fixed and static, as well as a religious perspective that a natural order of human variation existed and could be described.³

Later attempts at classification varied extensively in regard to the breadth of group differentiation. Classifiers generally could themselves be classified as either lumpers or splitters, reflecting the number of groups identified. In regard to Native Americans, Morton and Hrdlička tended to agree with Linnaeus in a single group classification. In contrast, others (e.g. Retzius, Meigs, Virchow, Ten Kate, Dixon and Hooton) all recognized various subgroups within the general Native American category.³

With augmented information on human variation, especially in regard to population genetics, the morphological boundaries of the old racial types became elusive. Data revealed a continuum of variation that was not clearly clustered into typological categories. In regards to the original criteria of Linnaeus, skin color appeared to vary extensively in different regions of the world and the behavior variables seem to reflect the attitudes and preconceptions of the classifiers rather than characteristics of populations. While the scientific basis of these original groupings gradually crumbled, the terminology persisted and became ingrained in public/folk classification.³

The old racial concept of groups being static, pure and fixed gradually gave way to more dynamic, realistic views that recognized processes of gene flow and genetic variation within all groups and areas. Besides gene flow, dynamic adaptation (by natural selection), demography, namely population size, sexual selection and generic drift all have played a role in shaping the nowadays variation.³

Although much of the racial terminology continued, embedded in public perceptions of variation, many anthropologists argued that the terms and underlying implied foundations had become toxic and subject to exploitation. The influential T.D. Stewart argued that the need persisted to examine and document human variation but noted the evils of racism and that the word “race” had become problematic since it had different meanings to different people. Past ethnocentric classifications frequently reflected world-view and religious orientation and led to improper value judgements and racism. Race concepts, or even the lack thereof, also varied extensively in different cultures.³

With changing perspectives on the nature of population variation, terminology emerged as a major issue. Many scientists began to refrain from using the term “race” in favor of ethnic group, breeding population and/or cline to describe human variation. Lasker distinguished “biological race” from “social race” based on the extent to which differentiation depended upon biological attributes or groupings defined by ethnic or social factors.³

The limited variation among humans at the genomic level has been discussed since Lewontin. It is clear that the majority of the variation exists inside all human populations and only limited variation between populations. In all, there is a limited amount of variation in humans that can be used to assist in generating an estimate of ancestry to assist with identification. But, above all, current efforts to discern ancestry from the skeleton aim to improve the likelihood of making a positive identification.³

Forensic Terminology

Forensic anthropology has struggled with terminology related to the evaluation of ancestry. The goals of forensic anthropology include providing information (the biological profile) regarding an unidentified skeleton to assist authorities in attempts at identification. Since missing persons are frequently described using racial terminology, forensic anthropologists are guided to use that terminology as well. This effort is challenging in that the anthropologist needs to use terms that will be recognized and be useful in the search but also needs to avoid being labeled by colleagues as a 19th-century taxonomist.³

For these reasons, the use of the term “race” has diminished markedly in forensic anthropology publications, discussions, and forensic reports. Most forensic anthropologists prefer instead to discuss likely ancestry. Reports should focus on the likely ancestry of the examined individual or, alternatively, suggesting how this person likely would have characterized himself/herself or have been socially classified by the communities they lived in. Such an approach provides the needed, useful information but avoids any suggestion of use of an outdated racial typology by the investigator. The biological information generated from anthropological analysis of a skeleton must be considered in judging how a person was regarded in terms of community definitions of race and ancestry. Nowadays, the recommended terms relating to the three main geographic groups are African, European and Asian. Specific cases may use targeted groups terminology as defined locally. ³ Certain terminology may be more useful than others though. Case in point, the term Hispanic. Hispanic is a term used for Spanish speaking populations. However, these populations very widely physically and genetically, so not all Spanish speaking populations should be lumped together as one ancestral group.¹

Cultural/Temporal Approaches

It is useful to keep in mind that the goal of providing information on ancestry is to facilitate identification and allow searches of/for missing persons. As noted in the discussion above, even direct ancestry evaluation of human remains involves cultural/historical factors and local folk racial classifications. The categories themselves contain a social/historical component shaped by local culture and community standards of communication. The language employed in discussion and report writing should reflect local standards, i.e. guidelines, and be oriented to facilitate identification and not mislead.³

Cultural data found on the skeleton also can provide direct evidence of ancestry. In most parts of the world cultural information can provide clues of the deep past and indicate that the recovered remains reflect archaeological contexts rather than modern forensic ones. Such information usually takes the form of associated artifacts that reveal culturally and temporally specific mechanisms of dating the remains.³

In some regions, cultural factors as can assist group identification. Such evaluation may prove forensically valuable in consideration of local population history. Similarly, information on medical procedures and related technology may facilitate population identification. This information is especially critical in the evaluation of unidentified migrants.³

Anthropological Assessment

In relation to ancestry in the anthropological assessment of the skeleton, there are two main approaches to be considered, non-metric and metric. The benefits and drawbacks of each of these approaches are worthwhile to mention: the less objective nature of the non-metric assessment is obvious. Non-metric analysis also requires more personal experience. For example, there is ambiguity in the evaluation of the nasal bridge, which can be classified as medium for one person and projected for another. Yet, it is also true that non-metric observations can capture much more information. We strongly recommend the use of both approaches, as follows.³

No matter the type of approach, the accuracy of techniques is always dependent on validation studies. Until the method developed on the basis of a certain sample is applied in a different sample, the results cannot be validated. When choosing the methods one should always pay attention to double check whereas that method is recommended for the sample in question. Above all, it is important to ensure that there is an appropriate tool to distinguish ancestry that includes references from the population that the individual derives.³

To infer about ancestry, the skull, in particular the mid-region of the face, is unanimously accepted as the most informative part of skeletal anatomy. Therefore, we will mainly focus on cranial examination. All the procedures to allocate ancestry, no matter what statistical treatment is followed, focus on craniometrics or on non-metric traits based on the assumption that there is a significant cranial diversity. However, it is obvious that there is a considerable overlap of the features, no matter their type. This is particularly true in the globalized world we are now living in due to the unprecedented admixture which can lead to very complex cases when attempting to distinguish ancestry from a skeleton, especially if we do not have a genomic assessment of predicted ancestry. Hence, the accuracy of ancestry estimation is hampered. It is important to bear in mind that cranial traits and measurements are always phenotypic features, partially determined by hereditability and influenced by the environment. Although there are polymorphisms that are quite distinctive of geographic regions, there isn’t a single trait that can be found only in a single population. The pattern of multiple traits offers a guide only to the most probable group of origin.³

Estimating Ancestry: Non-Metric Approaches

Because human populations vary in their phenotype due to environmental forces, forensic anthropologists are able to use non-metric traits to predict the ancestral origins of an unidentified individual. In general, anthropologists are able to divide humans into broad geographically discrete groups, including (but not limited to) the following: European, African, Asian, Native American, and Hispanic. ² Within the non-metric approaches, two types of traits should be considered: morphoscopic ones, which evaluate the shape, size and discrete traits, that are recorded as present or absent. The list of non-metric traits for the skull is particularly large; more than 200 have been described. Among the morphoscopic features, suture shape, as well as palate shape are good examples; for the discrete traits, wormian bones and the metopic suture are among the most known ones.³

Table 1 Differentiating features and points between ancestral groups.⁴ 

Scoring systems like OSSA, Optimized Summed Scoring Attributes, can also be used. Since no single trait or suite of traits accurately defines a population, there was a need to find a way to evaluate the threshold from which an individual could be considered as a member of a certain geographic group. OSSA quantifies the probability of a certain individual belonging to a given population.³ However, OSSA can only be used to classify skeletons as being of African or European ancestry. OSSA involves the assessment of six traits: Post-bregmatic depression (present or absent), Nasal bone structure (based off the shape), Nasal aperture width (broad, narrow, or intermediate), Interorbital Breadth (broad, narrow, or intermediate), Inferior Nasal aperture (based off the shape), and Anterior nasal spine (absent, blunt, or sharp).¹ A score sheet is provided where each trait is scored. The sum of all scores gives the ancestral group.³

The analysis of dental traits should always be considered as well in particular when the experts do not have a complete cranium or the morphological and/or metric analyses provide ambiguous results. Basically, non-metric dental approaches can be useful when only teeth are available for analysis and to support other methods. However, it should be bear in mind that dental studies are limited by teeth present and can have large error rates within their classification equations. Dental morphological characteristics have always been recognized as very informative in relation to ancestry and have a long history within dental anthropology. Two of the most known traits are Carabelli’s cusps and shoveling of incisors.³

To increase the power of non-metric traits to evaluate ancestry, the frequency of many more traits in additional populations around the world is needed. More information is required about the frequencies of well-known traits such as metopic suture, which is thought to be more common among Europeans, or the “Inca” bone, which is purportedly more frequent among South Americans. A better knowledge on traits distribution should definitely be a priority in order to avoid problematic reliance on the experience of the observer.³

Estimating Ancestry: Metric Approaches

Ancestry can also be determined through metric analyses.² The metric approach is more traditional and has more ancient roots. Craniometry has a long practice in both physical and forensic anthropology. It has the advantage of being more objective since each cranial measurement is well-defined on the basis of well-defined craniometrics points. Nevertheless, an index or simple ratio should not be used in forensic anthropology analysis for ancestry estimation since more complex craniometrics methods offer greater accuracy.³

What becomes critical is the way to combine the several cranial measurements and the selection of those that are more relevant. Linear discriminant analysis is one of the more used statistical approaches. This procedure was used in the North-American software FORDISC, which, on the basis of a maximum of 34 cranial and 39 postcranial measurements calculates discriminant function (DF). Essentially, an unknown individual will be compared with those represented in the database, that is, measurements of an unknown individual are compared with measurements of individuals with known ancestry in the database. This means that if the geographic region of the individual under analysis is not represented, the ancestral group cannot be found. Classification of an unknown individual is based on overall similarity. The accuracy is increased when sex estimation is performed by other means than the skull. The FORDISC database includes a forensic database as well as the famous Howells craniometrics series. Currently, it is largely used in the US and because the database is composed largely of North American forensic cases, it works better there than in other geographic regions.³

Cranid is another software that enables the assessment of the skull’s probable biological ancestry (in the broad geographical sense). On the basis of 29 measurements, the skulls are classified after comparison with 74 samples that include 3,163 skulls from around the world.³

Lately, techniques of geometric morphometrics (GM) have allowed a good analysis of cranial shape through three-dimensional (3D) coordinate data. GM is a statistical analysis based on Cartesian landmark coordinates. GM software developed at North Carolina State University (NCSU) called 3 D-iD is freely available software which, besides ancestry, also allows sex assessment. To use 3D-iD, however, a digitizer is needed since most of the 3D data are collected using digitizers that record the location of particular points in three dimensions. Digital morphometrics is also being used to assess the shape of particular features, such as suture shape. Linear discriminant analyses using craniometrics and geometric morphometrics, have been able to identify group differences with high cross-validated accuracies of 89%.³

Regarding postcranial methods, they are not only less investigated but also less used. Although several postcranial bones are being searched, the femur is, by far, the most examined one. But new research is being published with promising results not only for the femur but also for the tibia.³

Other approaches

Ancestry analysis is one of the areas where the cooperation between forensic anthropologists and geneticists can prove fruitful. When ancestry estimation proves difficult or impossible with other methods, DNA can be extracted and specific molecular markers, known as ancestry informative markers (AIMs) can provide an answer. Retrieving DNA from the skeleton has been incredibly improved during the last decades and nowadays it is possible to assess ancestry through targeted sequencing of very small quantities of DNA, including degraded samples.³

Numerous molecular analyses using combinations of single nucleotide polymorphisms (SNPs), short tandem repeats (STRs), variable number of tandem repeats (VNTRs) or even certain insertions/deletions (INDELS) indicate strong molecular patterning in worldwide samples, allowing an accurate classification of groups, despite large amounts of within region variation. Furthermore, SNIPs have been allowing a differentiation, without error among African, Asian and European groups proving that a good panel of ancestry informative SNPs can provide very good estimates. Also, noteworthy that DNA analysis of AIMs and physical trait markers from biological stains can also help provide investigative leads in cases without suspects. Some specific markers can suggest physical traits.³

Above all, the assessment of the geographic origin is a holistic approach in which perspective from different disciplines and datasets are important. For example, clothes, labels of clothing and personal belongings can provide some guidance. Within these secondary ancestry indicators, epidemiological data on some bone diseases may provide some clues since some pathological conditions are more frequent in some regions of the world.³

Reporting Ancestry

As the report is an important component in a forensic case, the way ancestry is reported is likewise paramount. It is important to clearly state how ancestry was evaluated, namely, which bones/anatomical areas were used. The methods applied should also be clearly indicated. Equally important when reporting ancestry is to include the accuracy of the method.³

Anthropologists should not “overstep” with their ancestral classification of the remains (e.g. classifying a skeleton as being of European ancestry when only postcranial remains are available for analysis), and should use “probable” when appropriate. Furthermore, there are always a considerable number of cases in which the final result is indeterminate. In the use of FORDISC, Ossa, genomic assessment and 3D-iD, it is possible to quantify the statistical probability of belonging to a certain population group. Otherwise, terms like possible, probable, compatible and consistent, or indeterminate are usually employed.³

Since databases used by law enforcement organizations and missing persons lists make reference to ancestry, this parameter may allow an exclusion. If the recovered skeleton suggests an African origin and all the individuals in a given data basis are European, an exclusion can be hypothesized, that is, we can state that most probably there is no match with any of the missing persons listed. However, it should be recognized that many populations are poorly represented in the published literature regarding skeletal morphology. Hence, caution is called for. More comprehensive databases are needed for missing persons to strengthen identification efforts.³ 

Final Considerations

Among the major four parameters of the biological profile, ancestry is the least applied since there are still some practitioners in some countries who don’t do it. Indeed, there are still many forensic anthropologists who simply do not evaluate ancestry despite the missing persons lists always making reference to the geographic origin of the disappeared. Unquestionably, this parameter remains controversial, despite objective guidelines and software now being available. However, with respect to the use of software and mathematical formulae in general, the expert should always verify whether the appropriate geographic region is represented in the databases. The quantification of the results and establishment of statistical probability are of utmost importance for forensic anthropology. Moreover, the validation of the methods worldwide is also critical. Above all, a holistic approach, including anthropology, genetics, genomic, and other disciplines, offers opportunities not only for exclusion but also to generate useful information to assess ancestry from recovered human remains.³

References:

1. Angi M. Christensen, Nicholas V. Passalacqua, and Eric J. Bartelink, Forensic Anthropology: Current Methods and Practice, 2nd ed. (London: Academic Press, 2019): 272-277.

2. Ashley Kendell, Alex Perrone, and Colleen Milligan, “Bioarcheology and Forensic Anthropology” In Explorations, ed. Beth Shook, Katie Nelson, Kelsie Aguilera and Lara Braff (Arlington: American Anthropological Association, 2019). https://pressbooksdev.oer.hawaii.edu/explorationsbioanth/chapter/osteology/.

3. Eugénia Cunha and Douglas H. Ubelaker, “Evaluation of Ancestry from Human Skeletal Remains: A Concise Review,” Evaluation of Ancestry from Human Skeletal Remains: A Concise Review,” Forensic Sciences Research 5 (2020): 89–97. https://doi.org/10.1080/20961790.2019.1697060.

4. Purva Wagisha Upadhyay and Amarnath Mishra, “Forensic Anthropology” in Biological Anthropology – Applications and Case Studies, ed. Alessio Vovlas (London: IntechOpen, 2021). https://www.intechopen.com/chapters/73372.

5. Roberta Hall, Kenneth Beals, Holm Neumann, Georg Neumann, and Gwyn Madden, Introduction to Human Osteology (Michigan: Grand Valley State University, 2010). https://pressbooks.gvsu.edu/introhumanosteology/.

Figure Attributions:

Figure 10.1 Skeletal traits commonly used in a morphological assessment of ancestry by Colleen Milligan original to Explorations: An Open Invitation to Biological Anthropology is under a CC BY-NC 4.0 License.