Elemental beginnings

Given their ubiquity in dietary reconstruction, it is fitting that the story of isotopes began with a conversation over dinner. Although coined in scientific literature by Frederick Soddy (Reference Soddy1913), the word ‘isotope’ was first conceived by Margaret Todd, a medical doctor (also known as the novelist ‘Graham Travers’, and an all-round gender-stereotype-smasher of their age). In 1912, Soddy and Todd were attending a supper in Glasgow. When talk turned to work, Soddy described the then nameless concept of elements of different masses that occupy the same place in the periodic table. Todd suggested the term ‘isotope’, from the Greek isos (‘same’) + topos (‘place’), and the name stuck (Nicol Reference Nicol1957; Nagel Reference Nagel1982).

Two decades after Soddy's death, the first stable isotope studies in archaeology were published. Emerging from advances in radiocarbon dating, plant sciences and ecosystem research (e.g. DeNiro & Epstein Reference DeNiro and Epstein1978a; Vogel Reference Vogel1978), initial applications focused on the uptake of maize agriculture in North America, with prehistoric human bones serving as ‘markers’ for maize consumption (Vogel & van der Merwe Reference Vogel and van der Merwe1977; van der Merwe & Vogel Reference van der Merwe and Vogel1978). These seminal studies demonstrated the ground-breaking potential of this new technique in estimating past dietary patterns.

The use of nitrogen isotopes in exploring trophic level relationships in terrestrial and marine ecosystems followed soon after (DeNiro & Epstein 1981; Minagawa & Wada Reference Minagawa and Wada1984; Schoeninger & DeNiro Reference Schoeninger and DeNiro1984), and archaeological investigations of marine resource exploitation using carbon and nitrogen isotope data were published (e.g. Schoeninger et al. Reference Schoeninger, DeNiro and Tauber1983). Research on oxygen isotope ratios of meteoric water (Craig Reference Craig1961; Dansgaard Reference Dansgaard1964) allowed relationships between drinking water, body water and mammalian mineralised tissues to be investigated as a means of reconstructing climate (Longinelli Reference Longinelli1984; Luz et al. Reference Luz, Kolodny and Horowitz1984). In 1985, the first study relating skeletal strontium isotope ratios to geologically sourced strontium to establish lifetime mobility in archaeological individuals was published (Ericson Reference Ericson1985).

Fractionating into a field

Research into the relationship between isotopic inputs and bodily isotope values has been central to the development of isotope bioarchaeology. This has involved probing the complexities of environmental, physiological and metabolic variation in modern plants and animals (e.g. Ambrose 1991; Kohn Reference Kohn1996; Burton et al. Reference Burton, Price and Middleton1999; Heaton Reference Heaton1999). Controlled feeding experiments provided vital perspectives on the systematics, caveats and capabilities of isotopic approaches (e.g. DeNiro & Epstein 1978b; Tieszen et al. Reference Tieszen, Boutton, Tesdahl and Slade1983; Ambrose & Norr Reference Ambrose, Norr, Lambert and Grupe1993; Ambrose 2000; Howland et al. Reference Howland, Corr, Young, Jones, Jim, van der Merwe, Mitchell and Evershed2003; Sponheimer et al. Reference Sponheimer, Robinson, Ayliffe, Roeder, Hammer, Passey, West, Cerling, Dearing and Ehleringer2003). Of equal importance were studies of the ways isotope ratios can be altered after burial (e.g. DeNiro Reference DeNiro1985; Nelson et al. Reference Nelson, DeNiro, Schoeninger, DePaolo and Hare1986; Tuross et al. Reference Tuross, Fogel and Hare1988; Hedges et al. Reference Hedges, Millard and Pike1995; Collins et al. Reference Collins, Nielsen-Marsh, Hiller, Smith, Roberts, Prigodich, Wess, Csapò, Millard and Turner-Walker2002).

Quality criteria for bone collagen have increased confidence in the data produced (Ambrose Reference Ambrose1990; van Klinken Reference van Klinken1999; Nehlich & Richards Reference Nehlich and Richards2009). While similar easily applicable criteria are not available for the assessment of biomineral isotope data, there are emerging consensuses: e.g. that tooth enamel best preserves in vivo signatures, and that preparation protocols can alter isotope measurements substantially, particularly in bone (e.g. Koch et al. Reference Koch, Tuross and Fogel1997; Hoppe et al. Reference Hoppe, Koch and Furutani2003; Grimes & Pellegrini Reference Grimes and Pellegrini2013).

By highlighting the caveats and limitations of isotope approaches in bioarchaeology, these studies have together led to greater certainty that isotope measurements determined from archaeological remains are representative of in vivo values and have widened their applications. With the parallel developments in mass spectrometry (and its increased cost-effectiveness and availability), the 1990s and 2000s saw a huge increase in the output of isotopic data for archaeological case studies (Makarewicz & Sealy Reference Makarewicz and Sealy2015: 147).

With the non-specialist in mind, this commentary aims to provide a discursive overview of the history of isotope bioarchaeology, explore some highlights and challenges, and speculate a little on the position of isotope analysis in bioarchaeology now and into the future. Given the burgeoning size of the field, and in light of the author's own experiences, case studies explored here are largely focused on (relatively) recent time periods, and on Western Europe, although themes are hopefully universal. With a view to brevity, the reader, and the confines of word count, a limited number of references are given in the main text, and a more extensive (but not exhaustive) list of recommended further reading is provided as online supplementary material for some of the topics explored here. In the companion bibliography I also refer the reader to excellent specialist reviews for detailed information on the background, theory and methods, current status and future directions of isotope bioarchaeology.

Major contributions

Stable isotope approaches have made central contributions to several archaeological debates, and have opened new lines of enquiry. For example, isotope evidence for the diets and movements of our earliest ancestors has had profound implications for our understanding of human ecology and evolution (Lee-Thorp et al. Reference Lee-Thorp, Sponheimer, Passey, de Ruiter and Cerling2010; Copeland et al. Reference Copeland, Sponheimer, de Ruiter, Lee-Thorp, Codron, le Roux, Grimes and Richards2011; Schoeninger Reference Schoeninger2014).

Isotope analysis has also proved to be a valuable tool for investigating major archaeological transitions, such as the shift from Mesolithic hunter-gatherer-fisher communities to Neolithic farmers. Initial bone collagen studies suggested the abandonment of marine foods at the onset of the British Neolithic (Richards et al. Reference Richards, Schulting and Hedges2003). This sparked debate, as researchers sought to reconcile isotope data with other evidence (Milner et al. Reference Milner, Craig, Bailey, Pedersen and Andersen2004; Richards & Schulting Reference Richards and Schulting2006). Dental micro-sampling approaches have somewhat resolved this, confirming not only the dietary predominance of terrestrial foods, but also the intermittent, regional consumption of marine resources (Montgomery et al. Reference Montgomery, Beaumont, Jay, Keefe, Gledhill, Cook, Dockrill and Melton2013). Beyond the issue of fish, the first studies were significant in contributing to a rethinking of the mode of change during the Mesolithic to Neolithic transition in Britain, prompting new analyses that now overwhelmingly support an ‘abruptist’ model (Rowley-Conwy Reference Rowley-Conwy2011). Isotope approaches have, however, demonstrated that the dynamics of ‘Neolithisation’ were non-homogeneous world-wide. Strontium and oxygen isotopes, for example, indicate a more prolonged transition in Thailand, with changes in residency patterns as matrilocality gained predominance alongside agriculture (Bentley et al. Reference Bentley, Pietrusewsky, Douglas and Atkinson2005). In contrast, analyses in Central Europe have concluded that patrilocality very likely prevailed (Haak et al. Reference Haak, Brandt, de Jong, Meyer, Ganslmeier, Heyd, Hawkesworth, Pike, Meller and Alt2008; Bentley Reference Bentley2013).

These studies illustrate perhaps the most valuable contribution that isotope bioarchaeology makes to archaeology: the illumination of past intra-society variation. By providing evidence for individual life-histories, isotope studies reveal differences between individuals and within societies, such as dietary variation by age group (Pearson et al. Reference Pearson, Haddow, Hillson, Knüsel, Larsen and Sadvari2015) or faith (Alexander et al. Reference Alexander, Gerrard, Gutiérrez and Millard2015). Multi-isotope approaches are particularly effective for characterising inter-personal variations linked to socio-cultural identities (Knudson & Stojanowski Reference Knudson and Stojanowski2008); for example, the combination of isotope analyses used to identify fish consumption among the high-status immigrant Bishops of Whithorn (Müldner et al. Reference Müldner, Montgomery, Cook, Ellam, Gledhill and Lowe2009).

Expanding the spectrum

Isotope bioarchaeology is not confined to human mobility and diet; other aspects of nutrition, malnutrition and similar physiological conditions have been investigated. Isotope datasets evidencing the age at onset and subsequent completion of weaning (evidenced by a drop in trophic level; Fogel et al. Reference Fogel, Tuross and Owsley1989) have been published from British sites across multiple periods, from the Iron Age to the eighteenth/nineteenth centuries AD (e.g. Jay et al. Reference Jay, Fuller, Richards, Knüsel and King2008; Nitsch et al. Reference Nitsch, Humphrey and Hedges2011). Despite being from geographically and socially disparate populations, the volume of isotope data allows broad trends to emerge. Within a diachronic framework, the implications of the post-medieval reduction in breastfeeding duration in Britain highlight the potential relationship between breastfeeding, population increase and urbanism (Haydock et al. Reference Haydock, Clarke, Craig-Atkins, Howcroft and Buckberry2013).

Isotope studies of infant feeding practices are significant in that they illuminate an aspect of the past that is otherwise ‘invisible’, and create a narrative built around the experiences of women and children. These studies also mark an emerging area of bioarchaeology to which isotopes will contribute heavily in the future; namely, the interaction between socio-cultural change and diet, mobility and other life-history events. Culture-mediated biological and ecological change in humans has become highly topical in bioarchaeology in recent years. For example, research on the evolution of the human microbiome (Warinner et al. Reference Warinner, Speller and Collins2015) can be viewed as a response to the shift in archaeological theoretical frameworks to include those from evolutionary ecology, such as gene-culture co-evolution and niche construction theory (Laland & O'Brien Reference Laland and O'Brien2010; Laland et al. Reference Laland, Odling-Smee and Myles2010; Makarewicz Reference Makarewicz, Grupe and McGlynn2016: 200). Urbanism and industrialisation are likely to be focal points of future bioarchaeological research within these new frameworks. The near unique potential of isotope analyses to access the cultural, biological and environmental (White & Longstaffe Reference White, Longstaffe, Zuckerman and Martin2016), and to broaden our understanding of past activities and culturally mediated ecological changes, places these approaches at the centre of future bioarchaeological research.

If isotope bioarchaeology is to rise to these and other challenges, new or refined analytical approaches must be a priority. While bulk bone collagen studies are the current mainstay of palaeodietary reconstruction, carbon isotope analysis of single amino acids can provide more nuanced insights into protein sources, particularly in complex foodwebs (Fogel & Tuross Reference Fogel and Tuross2003; Corr et al. Reference Corr, Sealy, Horton and Evershed2005; McCullagh et al. Reference McCullagh, Tripp and Hedges2005; Webb et al. Reference Webb, Honch, Dunn, Eriksson, Lidén and Evershed2015). Single amino acid nitrogen isotope analysis, along with the compound-specific carbon isotope analysis of lipids and bone mineral, also have the potential to provide more nuanced insights into diet (Jim et al. Reference Jim, Ambrose and Evershed2004; Styring et al. Reference Styring, Sealy and Evershed2010; Colonese et al. Reference Colonese, Farrell, Lucquin, Firth, Charlton, Robson, Alexander and Craig2015). ‘Non-traditional’ elements, such as calcium (Reynard et al. Reference Reynard, Pearson, Henderson and Hedges2013) and zinc (Jaouen et al. Reference Jaouen, Szpak and Richards2016), will prove increasingly useful in palaeodietary studies, particularly where organics are not preserved (Jaouen & Pons Reference Jaouen and Pons2016). Underused since initial research, lead isotope applications will also become more widely used, corroborating strontium isotope evidence for human and animal mobility (e.g. Shaw et al. Reference Shaw, Montgomery, Redfern, Gowland and Evans2016). Other isotopes, such as neodymium (Tütken et al. Reference Tütken, Vennemann and Pfretzschner2011), may also prove to be useful provenance proxies. Finally, recent advances in micro-sampling approaches to human dentition (Beaumont & Montgomery Reference Beaumont and Montgomery2015; Willmes et al. Reference Willmes, Kinsley, Moncel, Armstrong, Aubert, Eggins and Grün2016) will make their application more routine, thereby enhancing temporal resolution in isotope studies by providing time-series dietary or mobility data for archaeological individuals.

Compound-specific approaches and new sampling strategies will help in taking stable isotope analyses ‘beyond diet’, and will aid in the identification of nutritional stress and disease (Reitsema Reference Reitsema2013). Similarly, as our understanding of the impact of culinary preparation on the isotope ratios of food and drink improves, other applications will arise. Boiling and fermentation, for example, can alter the oxygen isotope values of drinking water and therefore potentially alter human tissues when consumed (Brettell et al. Reference Brettell, Montgomery and Evans2012). While this poses problems for mobility studies, it could throw light on past drinking habits (Lamb et al. Reference Lamb, Evans, Buckley and Appleby2014). If experimental studies can improve our understanding of the isotope systematics of post-procurement foodways, then the use of established or ‘non-traditional’ isotope approaches in investigating cooking, or other preparation/culinary practices, may move to the fore.

Four legs good

Isotope zooarchaeology has recently proved valuable in disentangling complex aspects of human-animal subsistence, and economic and socio-cultural relationships (Makarewicz Reference Makarewicz, Grupe and McGlynn2016). Often using intra-tooth sampling (which provides time-series isotope records; Balasse Reference Balasse2003; Zazzo et al. Reference Zazzo, Balasse and Patterson2006), aspects of animal husbandry such as foddering and birth seasonality have been investigated, illuminating the experiences of animals and the decisions of herders (Balasse et al. Reference Balasse, Tresset and Ambrose2006; Towers et al. Reference Towers, Gledhill, Bond and Montgomery2014). Strontium isotope studies on archaeological fauna have explored aspects of past human lifeways such as trade and transhumance (Bentley & Knipper Reference Bentley and Knipper2005; Thornton Reference Thornton2011). The need to provide food for omnivorous domesticates brings additional considerations to human foodways, and isotope studies have explored these in a range of contexts (e.g. McManus-Fry et al. Reference McManus-Fry, Knecht, Dobney, Richards and Britton2016).

Remains of wild animals from Pleistocene sites have also been analysed to explore ancient foodwebs and environmental change (Bocherens Reference Bocherens2003; Richards & Hedges Reference Richards and Hedges2003; Stevens et al. Reference Stevens, Jacobi, Street, Germonpré, Conard, Münzel and Hedges2008; Feranec et al. Reference Feranec, García, Díez and Arsuaga2010). While this kind of isotope palaeoecology is well-established, its potential for archaeological studies is only now being realised. Archaeological faunal remains are significant in that they are often the product of human activity, and can provide insights into that activity. Herbivore intra-tooth oxygen isotope data can, for example, provide evidence of seasonal temperature variations and, when applied to anthropogenically derived assemblages, generate terrestrial palaeoclimate proxy data near-synchronous to human site-use (e.g. Bernard et al. Reference Bernard, Daux, Lécuyer, Brugal, Genty, Wainer, Gardien, Fourel and Jaubert2009). Strontium isotope evidence for migratory behaviours of important prey species, such as reindeer, can provide useful insights into human behaviours and decisions, such as landscape-use and hunting strategies (Britton et al. Reference Britton, Grimes, Niven, Steele, McPherron, Soressi, Kelly, Jaubert, Hublin and Richards2011; Price et al. Reference Price, Meiggs, Weber and Pike-Tay2017).

Connected lines of enquiry in isotope ecology and archaeology are now set to emerge; for example, in investigating the relationship between past climatic change, faunal migrations and human societies. Cross-disciplinary research projects will require specialists from both fields, and the combination of methodological and theoretical frameworks. Site-formation processes, and site-specific physical and chemical taphonomy, must all be understood in order to assess the potential and limitations of any particular assemblage or research question. New approaches to sampling must be sought, since the large sample sizes and controls typical in isotope ecology are difficult for archaeological materials to meet. To utilise isotope evidence for climatic change or faunal mobility (and through this, human behaviours and experiences), novel methods of analysing isotope data are also required. Among these, GIS tools and computational models that are commonly used in spatial ecology could allow landscape-level approaches to isotope data. This could take intra-tooth strontium data beyond the identification of migratory or non-migratory individuals, to the modelling of seasonal movements across ‘iscoscapes’. First, however, it will be essential to develop these tools within a proof-of-concept framework using modern materials.

An unnatural abundance

A growing strength of isotope bioarchaeology is the sheer quantity of data generated from ‘routine’ applications. The publication of full isotope datasets, along with %C, %N, C:N or strontium concentration data, is increasingly common and ensures that other scientists can properly access the data. Although practices in data reporting still require improvement (Szpak et al. Reference Szpak, Metcalfe and Macdonald2017), it is now possible to conduct original research using datasets combined from previously published studies. Such syntheses reveal diachronic and population-level trends, and allow enhanced critique of the capabilities of the techniques to, for example, identify immigrants from oxygen isotopes (Lightfoot & O'Connell Reference Lightfoot and O'Connell2016). New methods of analysing large datasets, such as Bayesian mixing models or GIS tools, will prove increasingly valuable (e.g. Fernandes et al. Reference Fernandes, Millard, Brabec, Nadeau and Grootes2014; Willmes et al. Reference Willmes, McMorrow, Kinsley, Armstrong, Aubert, Eggins, Falguères, Maureille, Moffat and Grün2014).

User-contributed datasets, such as ‘GenBank’ for DNA, present a promising new method of collating large quantities of bioarchaeological isotope data ready to be ‘excavated’ later. A recent call for a similar ‘IsoBank’ from the ecological community (Pauli et al. Reference Pauli, Steffan and Newsome2015) was met with enthusiasm from isotope archaeologists (Pilaar Birch & Graham Reference Pilaar Birch and Graham2015). The framework, management and costs of such a database, among other issues, still need to be addressed by the broader community (Pauli et al. Reference Pauli, Steffan and Newsome2015). Once these hurdles are overcome, however, ‘IsoBank’ will represent not only a sound curatorial move for a field with burgeoning data, but (and more significantly) will yield fantastic new opportunities to generate new knowledge.