A02_01

Compound-specific radiocarbon dating of lipid residues in pottery vessels: a new approach for detecting the exploitation of marine resources 

Casanova E ^1,2, Knowles T², Bayliss A³, Barclay A⁴, Walton-Doyle C², Evershed R

¹Museum national d'histoire naturelle, Paris, France, ²University of Bristol, Bristol, UK, ³Historic England, London, UK, ⁴Costwold Archaeology, Cirencester, UK

Over the last decades, organic residue analysis has been shown to be especially useful in ancient diet reconstruction; however, it is only recently that the direct radiocarbon dating of lipid residues has become a reliable method for dating pottery vessels and food procurement activities. We applied lipid residue analysis to late Bronze Age pottery vessels from the site of Cliffs End Farm (UK), previously dated by their visible charred residues and dated four of these vessels by compound-specific radiocarbon analysis (CSRA) of their absorbed lipids. Lipid biomarker detection was limited to only one class of aquatic biomarker in four vessels, which indicated that the diet relied mainly on terrestrial animals. Four vessels datable from their absorbed lipids produced ¹⁴C measurements significantly older than those on the charred residues, suggesting a reservoir effect affecting the dated lipids. Furthermore, the low abundance of pig remains (1-4%) at the site suggests that enriched δ¹³C16:0 and δ¹³C18:0 values recorded on the absorbed fatty acids were caused by marine contributions. The percentage of marine products in the CSRA dated vessels was quantified using the δ¹³C values of individual fatty acids and from a modern reference database. These percentages were incorporated in mixed-source radiocarbon calibration using OxCal for the correction of a marine reservoir effect. Therefore, the combination of lipid compositions, radiocarbon dates and faunal analyses enabled the identification of marine product exploitation at the site. Finally, this allowed comparisons to be made with the dates of charred residues obtained from the same sherds.

A02_02

Efforts to remove ancient carbon from charred food crust: Successes and Failures

Varney R¹, Scott Cummings L¹

¹PaleoResearch Institute, Inc., Golden, United States

After radiocarbon dating reference foods, and long after some archaeologists had abandoned getting reliable dates from pottery, we conducted experimental pretreatments to remove ancient carbon from the archaeological contexts. We used a mixture of chemicals to mobilize both water-soluble and nonwater-soluble compounds in the food crusts – with many good results.  We used a variety of chemical extraction methods including hydrolysis. We examined placement of the samples on the vessel – with good results. Rims generally yielded dates more congruent with dates on annuals from the same context. The key lay in viewing the cooking process as a proxy for chromatography. Carbohydrates, which are presumed to take in atmospheric carbon, burned or charred first and usually at or near the rim. Proteins charred to a lesser degree, and fats/lipids remained uncharred. Therefore, we sought to remove all of the fats/lipids and any uncharred proteins using these chemicals to be able to obtain more congruent dates, which we did. But some dates remained too old, suggesting the charred food crust had integrated proteins that contained ancient carbon. At this time we focused on a simplistic animals vs. plants model. We read about medical research that could “uncook” proteins, returning them to a mobile rather than a fixed state. We participated in an experimental trial to test whether our charred food crust was a good candidate for this treatment. It was not. We present the multiple radiocarbon dates on these fractions for comparison and discussion.

A02_03

Examining carbon sources and radiocarbon dates on food plants and animals to understand offset dates on charred food crust

Scott Cummings L¹, Varney R¹

¹PaleoResearch Institute, Inc., Golden, United States

When ceramics, charred food crust, or even bone collagen yields ages older than expected, the Freshwater Reservoir Effect is suspected in several parts of Europe and North America, as well as elsewhere in the world. Research has examined whether or not regional offsets may be calculated or estimated and also whether or not some or all of the ancient carbon may be removed, particularly from ceramics or charred food crust. Although we expect that all carbon entering an organism is available for deposition into the cells of that organism, few studies have examined distribution of radiocarbon dates within single organisms by tissue type. In light of the fact that individual vessels may be used to cook or process multiple meals and foods, we have focused a portion of our research on radiocarbon dating several portions of that food prior to the time when it become food for human consumption. After radiocarbon dating bone collagen from four fish caught in 1939 and observing offsets varying from approximately 300 to 1200 radiocarbon years, we have broadened our research to include radiocarbon dating multiple samples (tissue and bone collagen) from reference animals, as well as stomach contents. For plants we examine seeds, stems, and leaves. Stomach contents of aquatic animals have yielded greater offsets than did either the flesh or bone collagen of those animals. We compare this information with radiocarbon dates on charred food crusts from various projects in the US.

A02_04

Permafrost melt as the driver of archaeological and modern freshwater reservoir effect

Hyland C¹, Schulting R¹, Weber A², Styring A¹

¹University Of Oxford, Oxford, United Kingdom, ²University of Alberta, Alberta, Canada

Freshwater reservoir effects (FRE) are extremely varied in their driving factors which makes them challenging to correct with both high accuracy and precision. Identifying the factor(s) that have driven the incorporation of “old carbon” into freshwater environments helps researchers to improve the precision and accuracy of dating archaeological remains from important freshwater resource contexts. New insights from the radiocarbon dating and stable carbon and nitrogen isotope analysis of modern freshwater fish from the rivers surrounding Lake Baikal, Russia provide insight into the driver of this region’s modern FRE. The modern freshwater fish and previous paired dating of archaeological terrestrial faunal in direct association with human remains indicate a negative relationship exists between bulk δ¹³C values and offsets in the ¹⁴C age caused by FRE. This research proposes that both the modern and archaeological FRE is driven primarily by the incorporation of permafrost melt into these freshwater ecosystems. The permafrost melt acts as both a source of “old carbon” and a driver of low bulk δ¹³C values due to the C₃ vegetation that the permafrost contains. Additional research using stable sulfur and hydrogen isotope analysis will further examine the role of permafrost melt in the FRE of the Lake Baikal region.

A02_05

Elite diet and its effects on the ¹⁴C dates of Estonian Bronze Age human remains

Tõrv M¹, Oras E¹, Meadows J², Lang V¹, Kriiska A¹

¹University of Tartu, Tartu, Estonia, ²Zentrum für Baltische und Skandinavische Archäologie, Schleswig, Germany

Human bone is a valuable material for dating ancient events, but these ¹⁴C dates can be misleading if the ingested carbon derives from organisms outside of the atmospheric isotopic equilibrium. Dietary intake of freshwater and marine species may result in markedly older dates than the actual event. Thus, to interpret the ¹⁴C dates obtained from human bone, we firstly need to identify the food sources and their proportions in the diet of the analysed individuals. We will present an exemplary case study about the Bronze Age stone cist grave populations – the first archaeologically visible elite buried into the monumental graves – in Estonia. Recently ca. 100 new radiocarbon dates have been obtained from these human skeletal remains to establish a new Bronze Age chronology in Estonia. To critically evaluate these dates, we revisited the published ¹⁴C dates and present new stable isotope results – carbon (δ¹³C) and nitrogen (δ¹⁵N) – from these graves, focussing on the relationship of the ¹⁴C dates to the individual dietary preferences. 

We reconstructed the elite food habits of the Bronze Age populations in Estonia. We employed quantitative palaeodiet reconstruction and dietary reservoir effect estimation to indicate the complexities behind dating ancient human remains. Statistical modelling of calibrated dates further allowed reconstructing the chronology of specific burial practices. With this study we aim to exemplify how combined analytical approaches allow more refined interpretation of ¹⁴C dates and reconstruct both site-specific and wider chronology of Bronze Age elite burial customs in Estonia.   

A02_06

Dietary reservoir effect correction for mid-Holocene human remains from Sakhtysh, Russia: a novel regression-based approach

Meadows J^1,2, Khramtsova A³, Kostyleva E⁴, Krause-Kyora B⁵, Piezonka H⁶

¹ZBSA (Centre for Baltic and Scandinavian Archaeology), Kiel, Germany, ²Christian-Albrechts-Universität zu Kiel, Leibniz-Labor für Altersbestimmung und Isotopenforschung , Kiel, Germany, ³Excellence Cluster ROOTS, Christian-Albrechts-University Kie, Kiel, Germany, ⁴State University Ivanovo, Ivanov, Ivanovo, Russian Federation, ⁵Institute of Clinical Molecular Biology (IKMB), Christian-Albrechts-University Kiel, Kiel, Germany, ⁶Institute for Pre- and Proto-History , Christian-Albrechts-University Kie, Kiel, Germany

Several small prehistoric cemeteries at Sakhtysh, Ivanovo Oblast, Russia, are attributed to the middle Neolithic Lyalovo and late Neolithic-Eneolithic Volosovo cultures (c.5000-4000 cal BC and c.4000-2500 cal BC respectively). Human bone δ¹³C and δ¹⁵N results confirm that these groups were hunter-gatherer-fishers, with sometimes large dietary differences between individuals (Engovatova et al. 2015). Accurate dating has been challenging, due to variable collagen preservation, uncertain association between human skeletal remains and osseous grave goods, and the unknown magnitude of mid-Holocene freshwater reservoir effects, which must have been much greater than those recorded in modern fish.

We present a regression-based approach, which instead of comparing ¹⁴C ages of human bones to those of associated grave goods, relies on ¹⁴C-age differences between different skeletal elements of the same individual. The resulting dietary reservoir effect estimates for 53 samples from 39 individuals are compatible with estimates produced by a diet-reconstruction model based on realistic stable isotope and freshwater reservoir effect baseline values, and with a handful of ¹⁴C dates of osseous grave goods (Macāne et al. 2019). Calibrated dates of individual burials remain imprecise, but we can still observe temporal trends in human diets, which parallel trends elsewhere in northeastern Europe over the same date range.

A02_P01

Direct and Indirect Attempts at Diachronic Quantification of the Marine Reservoir Effect

Niedospial J¹, Sevink J², Maurer A, Mazzini I⁴, Arienzo I⁵, Kuijper W⁶, van Hall R², Dee M¹

¹Centre for Isotope Research, University of Groningen, Groningen, Netherlands, ²Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Amsterdam, Netherlands, ³Groningen Institute of Archaeology, University of Groningen, Groningen, Netherlands, ⁴Institute of Environmental Geology and Geoengineering, National Research Council of Italy, Rome, Italy, ⁵National Institute of Geophysics and Volcanology, Naples, Italy, ⁶Faculty of Archaeology, Leiden University, Leiden, Netherlands

The marine reservoir effect (MRE) is fundamental to understanding the carbon cycle and for correction of radiocarbon dates on samples either partially or wholly derived from marine carbon. Estimates of the offset commonly make use of paired samples of contemporary marine and terrestrial material, with the resultant offset then applied to the time period of interest. Paradoxically, it is widely accepted that the MRE at any one location is likely to have fluctuated over time, calling into question the validity of such retrojections. In this

study, we set out to both calculate the MRE for a specific location and then to understand the dynamics behind its variation over time using an array of geochemical and palaeoecological proxies. Previous research at our study site of Puntone, on the Tyrrhenian Sea, revealed that the area had transitioned between open marine, lagoonal and terrestrial regimes, and hence the local MRE was likely to have altered over time. This expectation was confirmed by the new radiocarbon data we obtained on marine shells and terrestrial plant remains. First attempts at using metal ion and isotope ratios through the core to elucidate, and indeed model, the MRE fluctuations have shown signs of promise but additional research is required to develop these approaches.

A02_P02

Variability of Radiocarbon reservoir age effects in lakes and rivers of Eastern Anatolia and Lesser Caucasus

FONTUGNE M^1,2, HATTÉ C^2,3, TISNÉRAT-LABORDE N², OLLIVIER V¹, KUZUCUOGLU C⁴

¹LAMPEA, Aix-en-Provence, France, ²LSCE - CEA, Gif-sur-Yvette, France, ³Silesian University of Technology, Gliwice, Poland, ⁴Laboratory of Physical Geography , Meudon, France

Multiproxy sedimentary sequence analysis constitutes the basis for reconstructions of past paleoenvironments and climate evolution. These sequences are, for the most part, obtained by coring in lakes (maar or crater) whose waters can record volcanic activity or karstic contributions, especially in Eastern Anatolia and the Lesser Caucasus. The main consequence is to generate a reservoir age effect and to bias the radiocarbon dates of sedimentary records from these lakes. In the same way, the halieutic resources bordering these lakes also record this reservoir effect. And, as they constitute the food resources of the local populations, this reservoir effect is also reflected in the skeleton of the lake population. We present, here, some results obtained from eastern Anatolian lakes, Van and Sevan lakes and from archaeological sites along the Kura river and its tributaries from Lesser Caucasus.

A02_P03

Identification of marine reservoir effect in the Holocene sediments from the Nobi Plain, central Japan

Nakanishi T¹, Hori K², Nakashima R³, Hong W⁴

¹Museum of Natural and Environmental History, Shizuoka, Japan, ²Department of Earth Science, Tohoku University, Sendai, Japan, ³Geological Survey of Japan (GSJ), Advanced Industrial Science & Technology (AIST), Tsukuba, Japan, ⁴Korea Institute of Geoscience and Mineral Resources (KIGAM), Daejeon, Republic of Korea

To investigate the relationship between paleoenvironmental changes and marine reservoir effects, the radiocarbon ages of marine shells and terrestrial plants were measured from the same horizons of the Holocene sediments. Two sediment cores, HN1 and OG, were obtained from the western margin of the Nobi Plain, which faces the Kuroshio warm current. This plain is a fluvial–coastal lowland formed mainly by Kiso, Nagara, and Ibi rivers (Hori et al., 2019). These drilling sites are located in subsidence area associated with the Yoro fault (Kuwahara, 1968; Ishiyama et al., 2007). Based on analyses of lithology, molluscan assemblages, and radiocarbon dating, we interpreted five sedimentary units in order of older age: estuary, prodelta, delta front, delta plain, and artificial soil. These paleoenvironmental changes had been mainly associated with the sea-level rise during the deglacial period. Terrestrial accumulation curve was consistent with the Kikai-Akahoya volcanic ash (K-Ah: ca. 7,300 cal BP; Machida and Arai, 2003). The reservoir ages during the period from 9,300 to 2,800 cal BP of 13 pairs obtained from the estuary to delta front facies were evaluated. The average from 140 ± 70 to 900 ± 50 was 350 ± 180 years. The chronological change in the reservoir effect will be compared with the previous results from the other coastal area in Southwest Japan (Nakanishi et al., 2017ab, 2019). Drilling program of the sediment cores were supported by the Grants-in-Aid for Scientific Research, Kakenhi JP17K18526. Radiocarbon dating was funded by the JSPS Kakenhi grant number JP18H01310.

A02_P04

Time varying Local Marine Reservoir Effect in Coastal Systems

Macario K¹, Alves E¹, Oliveira F¹, Chanca I², Scheel-Ybert R³, Gaspar M³, Tenorio C³, Dias F¹, Aguilera O¹, Bianchini G³, Vitorino B¹, Cardoso R¹, Anjos R¹, Muniz M¹

¹Universidade Federal Fluminense, Niteroi, Brazil, ²Max Planck Institute, Jena, Germany, ³Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil

The radiocarbon (¹⁴C) Marine Reservoir Effect (MRE) is known to vary with time due to global environmental changes, now better represented by the Marine20 calibration curve. In coastal regions, however, the scenario is considerably more complex and the MRE is influenced by carbon sources that exhibit a wide range of isotopic signatures. Indeed, the complexity of the carbon cycle in coastal systems demands extensive research for an accurate quantification of the MRE and its variability, which has the effect of hindering accurate ¹⁴C chronologies on the coast. The local MRE offset (∆R) in estuaries, is subject of the interplay of the hydrography with factors such as the regional geology, upwelling and sea-level variations. Using paired archaeological samples from the Southeastern coast of Brazil, we show a highly variable MRE and a possible correlation with the sea level regression.

A02_P05

Freshwater reservoir effects in charred ‘food crusts’ on pottery: frequency, magnitude, risk factors and prospects for correction

Meadows J^1,2, Lucquin A³, Gonzalez Carretero L^3,4, Dolbunova K^4,5, Craig O³, Heron C⁵

¹ZBSA (Centre for Baltic and Scandinavian Archaeology), Kiel, Germany, ²Christian-Albrechts-Universität zu Kiel, Leibniz-Labor für Altersbestimmung und Isotopenforschung , Kiel, Germany, ³BioArCh, University of York, York, United Kingdom, ⁴Department of Scientific Research, The British Museum, London, United Kingdom, ⁵State Hermitage Museum, St Petersburg, Russia

The ERC-funded INDUCE project is concerned with the adoption of pottery by hunter-gatherer-fishers in northeastern Europe. In this region and period, it is often difficult to date pottery find contexts, but potsherds often have carbonised deposits (‘food crusts’, FCs). Thus ¹⁴C dating of FCs is desirable, but these FCs probably contain carbon from aquatic organisms with unknown ¹⁴C reservoir effects. The problem is more salient at inland sites, given the potential scale of freshwater reservoir effects (FREs).

To investigate pottery function, INDUCE has acquired EA-IRMS data (%C, %N, δ¹³C, δ¹⁵N) on FCs on more than 400 sherds, from over 50 sites. In most cases we have also obtained biomarker (GC-MS) and compound-specific δ¹³C (δ¹³C16:0, δ¹³C18:0) data on soluble lipids extracted from the same FC. Some FCs have also been examined by SEM-EDX microscopy. These analyses confirm that FCs are composed primarily of food remains, which are dominated by aquatic species.

We have dated more than 100 FCs, often when their context date, and/or local FRE, is well-constrained. We have tested whether any combination of isotopic, biomolecular and microscopic data predicts apparent ¹⁴C age offsets in FCs. Contrasting results from sites with >10 dated FCs illustrate the scope for using FC ¹⁴C to date hunter-gatherer-fisher pottery. Where FC ingredients are more varied, a multi-proxy approach appears to account for most of the variation in FC offsets, but at other sites, FC composition is too uniform, and FREs are too varied, to justify attempts at FC ¹⁴C-age ‘correction’.

A02_P06

Estimating freshwater reservoir ages using Bayesian models for the Mesolithic to neolithic transition

Olsen J¹, Maaring R¹, Mannino M¹

¹Aarhus University, Aarhus, Denmark

The chronology of the Mesolithic to Neolithic transition is difficult to date accurately using radiocarbon analysis on human or animal bone remains. This is because that in particular the hunter-gather culture of the Mesolithic involves diets from multiple sources. When both marine and freshwater food webs are exploited, it becomes difficult to correct the radiocarbon ages for possible reservoir effects. In particular, the freshwater reservoir effect is unknown and expected to vary substantially. We have collected a large dataset of dog bones across the Mesolithic to Neolithic transition from which we have deduced the percentage of terrestrial, marine and freshwater diets using a FRUITS model. Further, we have included a dataset of domesticated cattle. All from multiple sites across Denmark. We have constructed a Bayesian model with phases based on the archaeological typology, i.e. belonging to either Mesolithic or Mesolithic. In the model we have allowed for estimating the total reservoir age using a wide uniform prior. If we assume that the marine reservoir age is known then the freshwater reservoir age can be estimated from the posterior total reservoir age probability distribution. The calculated freshwater reservoir age vary from a few hundred ¹⁴C years and up 1200 ¹⁴C years. Further, we have tested the robustness using different Bayesian model of the Mesolithic to Neolithic transition and found our freshwater reservoir estimates are similar and independent of model choice.

A02_P07

Reservoir Effect determination in marine shells from Mexico

Rodriguez-Ceja M¹, Díaz-Castro M², Solis C¹, Álvarez-Lajonchere L³, Méndez-García C^1,4, Chávez-Lomelí E¹

¹Universidad Nacional Autónoma de México, Mexico City, Mexico, ²2 Instituto Superior de Tecnologías y Ciencias Aplicadas. Universidad de La Habana, Cuba., La Habana, Cuba, ³Museo de Historia Natural Felipe Poey. Universidad de la Habana, Cuba. , LA Habana, Cuba, ⁴Cátedras Conacyt. Instituto de Física. Universidad Nacional Autónoma de México. , Mexico City, Mexico

When trying to establish the chronology of a site through radiocarbon dating, terrestrial samples are usually preferred. However, since they are not always available, an alternative is the use of malacological material from marine organisms as mollusks.

Given that radiocarbon activity in aquatic environments is usually different from that of the atmosphere, organisms that grow in both environments have different apparent ages, even though they are contemporary. This apparent age is known as “reservoir effect” and varies depending on latitude and other local factors.

Mexico is a country with a littoral of more than 11,000 Km, and a great interest in dating malacological samples from coastal contexts.   However, reservoir effect studies are still scare. Must of the available data come from studies done in the 60s and 90s decades, of the XX century.

We present new reservoir effect data from coastal Mexican sites in the Pacific Ocean and the Caribbean Sea, obtained from a Collection of Mexican shells samples belonging to the Natural History Museum Felipe Poey, from the Universidad de La Habana, Cuba and to the Institute of Biology of the Universidad Nacional Autónoma de México.