Investigating the histories of hunter-gatherer plant use in Southwest Asia and the Eastern Mediterranean

Ceren Kabukcu*

University of Algarve, The Interdisciplinary Center for Archaeology and the Evolution of Human Behaviour (ICArEHB)

*with contributions from Chris Hunt (Liverpool John Moores University), Evan Hill (Queen’s University Belfast), Emma Pomeroy (Cambridge University), Tim Reynolds (Birkbeck University of London), Graeme Barker (Cambridge University) and Eleni Asouti (University of Liverpool)

The past few decades have seen mounting evidence of plant use by Palaeolithic hunter-gatherers drawn from plant residues on stone tools (Lippi et al. 2015), ancient tartar on teeth (Henry et al. 2011) and DNA from sediments (ter Schure et al. 2022), as well as plenty of evidence from the burnt remains of plants used as food and fuel. Without a doubt, prehistoric hunter-gatherers in most environments and across a considerable time depth relied on plants for survival and as a means to adapt to changing environments. And yet, research focusing on the protein components of prehistoric diets continues to persistently question the significance of plants, often considered to be ‘inferior’ food choices (Jaouen et al. 2019).

New evidence on prehistoric hunter-gatherer cooking

Our recently published article in Antiquity (Kabukcu et al. 2023) reports on a set of rather unusual charred remains from plant food preparations. The specimens we examined come from two Palaeolithic cave sites: Franchthi Cave on mainland Greece and Shanidar Cave in northern Iraq. Franchthi Cave provided evidence of food remains dating to c. 13-11,000 years ago. Shanidar Cave’s similar finds are considerably older, dating to the Baradostian Upper Palaeolithic (c. 40-50,000 years ago, corresponding to the early modern human occupation of the cave) and the Middle Palaeolithic (c. 55-75,000 years ago) Neanderthal occupation. Charred plant food remains at both sites (see Figure 16) contain several plant species, preserved in a gelatinised and bonded mass not dissimilar to burnt breadcrumbs, charred lumps of porridge or pancakes. Wild pulses (vetch, grass pea, pea, etc.) were the most common ingredients in all specimens examined from both sites. Scanning Electron Microscopy examination revealed that pulse seeds were likely soaked and then mashed in the process of food preparation. There is also evidence for the use of mustard seeds and terebinth (wild pistachio) nuts in Upper Palaeolithic food fragments as well as of grass seeds in Middle Palaeolithic fragments from Shanidar Cave.

Beyond identifying the plant food species and reconstructing the possible Palaeolithic food preparation steps, we considered the broader implications of our findings: why do we find common ingredients (wild pulses) across this remarkable time depth, and at sites located so far apart? Moreover, what might this mean for Palaeolithic plant foraging and food choices? Looking at the existing archaeobotanical evidence from the broader region, we argued that certain plant food species may be characteristic of Palaeolithic hunter-gatherer culinary traditions across Southwest Asia and the Eastern Mediterranean. They include wild almonds and pistachios, pulses, mustards and other plants characterized by tannin-rich, bitter or sharp flavours. The charred food remains we studied contained fragments of most of these plants, while seeds, starches, phytoliths (silica impressions of plant cells) and nutshells of the same species are also commonly found in archaeobotanical samples from these and other sites in the region. Thus, we argue that their processing and cooking as multi-component plant foods represent distinctive culinary practices that enabled their safe consumption, and that they might have been targeted for reasons beyond ease of collection or caloric content.

Figure 16: Scanning Electron Microscope images of charred food fragments. Left panel: Overview of Final Palaeolithic/Epigravettian pulse-rich specimen from Franchthi Cave. Right panel: Close-up of wild pea fragments in specimen from Upper Palaeolithic/Initial Baradostian from Shanidar Cave. SEM micrographs by Ceren Kabukcu.

To date, the Shanidar Cave specimens (some dated from as early as c. 75,000 years ago) represent the oldest known charred food remains containing multiple ingredients currently known anywhere in the world. The Franchthi Cave specimens are the oldest known in Europe. We have every expectation that comparable finds occur and will be studied from other Palaeolithic sites in Western Asia, Europe, and other world regions. Their systematic recovery and analysis are likely to provide further insights into the previously unknown and under-appreciated complexity of Palaeolithic foraging choices, cooking practices and culinary cultures. The study of charred plant macrofossils holds unique potential in this respect. Charred plant food remains preserve combinations of ingredients and the signatures of different food preparation steps that can reveal, for example, whether some plants (e.g., those with sharp, bitter and/or tannin-rich flavours) might have been used as flavouring agents mixed with starch or fibre-rich plants (e.g., tubers and grasses), and/or as components of, and preservatives in, meat, dairy or marine foods.

How did we get here?

The research summarized here began when Ceren Kabukcu embarked on a Leverhulme Trust Early Career Fellowship (2017-2021) researching how plants were used and managed by pre-agricultural societies in Southwest Asia and the Eastern Mediterranean. The ambition was (and continues to be) to capture continuity and change in the long-term histories of hunter-gatherer plant use in the timespan of 75,000 to 10,500 years ago, leading up to the transition from foraging to farming. The main objective was to collect and evaluate evidence of plant use by Mousterian Neanderthals, Upper Palaeolithic and Epipalaeolithic early modern humans of the northwestern Zagros and western Taurus mountains and the earliest sedentary hunter-gatherers of southeast Anatolia at the start of the Holocene. With the discovery of charred food remains in samples from Franchthi Cave during anthracological analyses, the project expanded in scope, incorporating archive samples from Franchthi as well. However, to claim that it all began (or was made possible) when research funding became available would be misleading. In fact, everything began with sampling…

Of samples, archives and novel methodologies

Located on the northeastern coast of the Peloponnese, Franchthi Cave was excavated between 1967–1976 by an international team led by T.W. Jacobsen of Indiana University. Franchthi is unique in the Aegean Basin for preserving a remarkably long archaeological sequence extending from the Upper Palaeolithic through the Mesolithic and the Neolithic periods (c. 38,000–6000 years ago). From an archaeobotanical perspective, Franchthi Cave is an iconic site. It represents one of Europe’s earliest examples of the application of water flotation for the recovery of charred plant macroremains from archaeological sites. In collaboration with Sebastian Payne and David French, the excavation team first employed bucket flotation in 1969. In 1971, they developed a modified machine-assisted water sieve (see Figure 17) for the large-scale recovery of light charred plant remains and heavier animal bone and artefacts from all excavated deposits (Diamant 1979). The archaeobotanical assemblage was previously studied and reported in detail (Hansen 1991). But owing to the rigorous sampling, as well as to the richness and density of the finds, we were able to revisit these materials decades after their original discovery and analysis (Kabukcu et al. 2023; see also Asouti et al. 2018).

Figure 17. Water flotation system used at Franchthi Cave (red reservoir for spring water can be seen in the left panel). Images obtained from the Open Access Repository of the Indiana University Archives.

Shanidar Cave lies on the northwestern Zagros Mountains of Iraqi Kurdistan and was originally excavated between 1951 and 1960 by Ralph and Rose Solecki of Columbia University. Apart from the discovery of some of the earliest Neanderthal remains in Western Asia, the site is famous for the retrieval of flower pollen by Arlette Leroi-Gourhan from sediments associated with the remains of the Shanidar 4 Neanderthal individual (Leroi-Gourhan 1975). At the time, systematic sampling for charred plant macroremains from archaeological sites (particularly Palaeolithic sites) was not common. The discovery of pollen grains invisible to the naked eye led Ralph Solecki to memorably declare that “the importance of the Shanidar IV skeleton was determined not with the physical anthropologist’s calipers, not by the cultural associations, but in a new direction, under the microscope. It was the clinging soil from around the burial that yielded the ancient pollens of flowers” (Solecki 1977: pp. 114). Although Leroi-Gourhan’s findings were questioned in the decades that followed, the impact of this study on the proliferation of archaeobotanical sampling in Palaeolithic sites should not be underestimated. Still, at Shanidar Cave the large-scale sampling and recovery of charred plant remains and microfossils alike would have to wait until 2015, when a team led by Graeme Barker began re-excavating the site and applied systematic sampling (Reynolds et al. 2016). From the start of the project, the team has been using a mechanical water flotation system for the processing of large volumes of excavated sediment and the recovery of charred plant remains, bones, shell, and lithics, etc. See Figures 18-19.

Figure 18. New excavation team at Shanidar Cave (left), led by Graeme Barker (right). Photos courtesy of Chris Hunt and Graeme Barker.

Figure 19. Flotation at Shanidar Cave (left), archaeobotanical samples drying in the shade (right). Photos courtesy of Chris Hunt and Graeme Barker.

At both sites, the charred plant food remains were discovered during the routine sorting and scanning of archaeobotanical flotation samples judiciously collected and processed, which represent hundreds of hours of work in the field and in the laboratory. While excavating and floating sediment samples in the field, it is impossible to know the density of charred plant remains as well as whether anything worth studying might be recovered. As it turns out, it was a more than worthwhile investment!

It is clear that, in addition to work on materials from new excavations, several researchers working on Palaeolithic plant use are successfully revisiting decades of archived samples and materials from old excavations. Going back to Ralph Solecki’s lecture in 1976: “this is what the barrel of a microscope directed at some grains of sand on a slide has revealed”. Sometimes under the barrel of a conventional microscope, sometimes under the beam of a SEM, archaeobotany will continue to reveal new and exciting discoveries in years to come.


  • Asouti, E., Ntinou, M. and Kabukcu, C., 2018. The impact of environmental change on Palaeolithic and Mesolithic plant use and the transition to agriculture at Franchthi Cave, Greece. PloS one, 13(11), p.e0207805.
  • Diamant, S., 1979. A short history of archaeological sieving at Franchthi Cave, Greece. Journal of Field Archaeology, 6(2), pp.203–217.
  • Hansen, J.M., 1991. The Palaeoethnobotany of Franchthi Cave. Excavations at Franchthi Cave, Greece Fascicle 7. Bloomington and Indianapolis, Indiana University Press.
  • Henry, A.G., Brooks, A.S. and Piperno, D.R., 2011. Microfossils in calculus demonstrate consumption of plants and cooked foods in Neanderthal diets (Shanidar III, Iraq; Spy I and II, Belgium). Proceedings of the National Academy of Sciences, 108(2), pp.486-491.
  • Jaouen, K., Richards, M.P., Le Cabec, A., Welker, F., Rendu, W., Hublin, J.J., Soressi, M. and Talamo, S., 2019. Exceptionally high δ15N values in collagen single amino acids confirm Neandertals as high-trophic level carnivores. Proceedings of the National Academy of Sciences, 116(11), pp.4928-4933.
  • Kabukcu, C., Hunt, C., Hill, E., Pomeroy, E., Reynolds, T., Barker, G. and Asouti, E. (2023) “Cooking in caves: Palaeolithic carbonised plant food remains from Franchthi and Shanidar,” Antiquity. Cambridge University Press, 97(391), pp. 12–28.
  • Leroi-Gourhan, A., 1975. The flowers found with Shanidar IV, a Neanderthal burial in Iraq. Science, 190(4214), pp.562-564.
  • Mariotti Lippi, M., Foggi, B., Aranguren, B., Ronchitelli, A. and Revedin, A., 2015. Multistep food plant processing at Grotta Paglicci (Southern Italy) around 32,600 cal BP. Proceedings of the national Academy of Sciences, 112(39), pp.12075-12080.
  • Reynolds, T., Boismier, W., Farr, L., Hunt, C., Abdulmutalb, D. and Barker, G., 2016. New investigations at Shanidar Cave, Iraqi Kurdistan. The Archaeology of the Kurdistan Region of Iraq and Adjacent Regions; Kopanias, K., MacGinnis, J., Eds, pp.369-372.
  • Solecki, R.S. 1977. The implications of the Shanidar Cave Neanderthal flower burial. Annals of the New York Academy of Sciences, 293, pp.114-124.
    ter Schure, A.T., Bruch, A.A., Kandel, A.W., Gasparyan, B., Bussmann, R.W., Brysting, A.K., de Boer, H.J. and Boessenkool, S., 2022. Sedimentary ancient DNA metabarcoding as a tool for assessing prehistoric plant use at the Upper Paleolithic cave site Aghitu-3, Armenia. Journal of Human Evolution, 172, p.103258.

Go back to top 

Thrown away - intentionally or by mistake? A 12th century enamelled reliquary pendant from an ancient dump in the historic district of Mainz – element analysis and content visualization, using Prompt Gamma Activation Analysis and Neutron Tomography (PGAI-NT)

Matthias Heinzel1, Eschly Kluge2, Dorothee Kemper3, Burkhard Schillinger4 and Christian Stieghorst4

1Leibniz Zentrum für Archäologie
2Institut für Kernphysik (IKP), Universität zu Köln
3Deutscher Verein für Kunstwissenschaft e.V. Berlin
4Heinz Maier-Leibnitz Zentrum (MLZ), Technische Universität München

During my time as freelance conservator for the state archaeology in Mainz, Germany, I often received ‘fresh’ interesting objects from current excavations in Mainz and the surrounding area. Most of these came from the Roman Period and the Middle Ages. One late afternoon during the excavation campaign 2008-2009, my colleague Klaus Soukoup brought me a small, severely corroded, but interesting looking metal object from an excavation he was working on in the historic district of Mainz. The excavation had taken place in the courtyard of a baroque palace. The object was found in a trash pit. The pit mostly contained pottery from the middle of the 14th century. But, this curious piece of corroded metal was also found. Over the following months I restored the small object, slowly uncovering that it was probably a pendant in the shape of a quatrefoil. My colleague, Stephan Patscher from the Leibniz Zentrum für Archäologie (LEIZA, before 2023 known as RGZM), initially made some radiographs of the object. The X-rays revealed the exact shape of the quatrefoil and some figural representations: we did indeed have a pendant from the Middle Ages. See Figure 20.

When I began a new job at LEIZA in 2016, I was able to explore the pendant in more detail. On a conference in Hildesheim (Lower Saxony) in 2017, I got to know art historian Dorothee Kemper. When I showed her photos of the pendant, she was able to classify it clearly in terms of art history. She also directly recognized the function as a relic pendant and encouraged me to look into it more closely.


Figure 20. Pendant before restoration and radiograph of side and front. Photograph by M. Heinzel, Radiographs by S. Patscher, LEIZA.

Figure 21. Front, side and reverse of the pendant after restoration. Photographs by S. Steidl, LEIZA.


Also known as a phylacterion, the pendant is quatrefoil in shape with a central square. It is made of gilded copper. An eyelet is visible at the top, suggesting that the object could have been worn around the neck on a ribbon or chain. It consists of two parts: the container and the closure. These two parts are held together by a visible rivet. Front and back are enameled using the émail champlevé technique and are decorated with several figural representations. See Figure 21. All figures carry a nimbus and are raised from the background, engraved and gilded. The background fields, halos, the frame of the central field and the internal drawings of the figures are all enameled. The colour spectrum of the enameled areas is quite narrow and varies between the front and back. The representations in the upper semicircle segment of both sides (where the riveting of the two components is located) were constructed taking into consideration the placement of the rivet, so the figures on the front and back tilt their heads to the right, thereby providing a place for the rivet. The side section of the pendant is cross-hatched (with the exception of the closure). The lozenges are gilded, and the lines are enameled. The pendant has the following dimensions: length 67mm (with bail), 57mm without the bail; width 57mm and thickness 11-12mm. The handle is about 10 mm in diameter. The object’s total mass is 70.5g.

Technological analyses

Two different techniques were applied for the analysis of the pendant’s surface. Elemental analysis was done using μ-XRF. Results found that the substrate metal is copper (Cu 99.2% / Pb 0.4% / Ag 0.2% / Sn 0.2%) and the gilding is fire-gilding, evidenced by a high content of mercury (Au 83% / Hg 17%). The glass and pigments of the enamels were investigated using Raman spectroscopy. There are four different opaque enamel colours on the object: white, blue, green and turquoise green. All colours were made from a soda glass; the green glass is a leaded glass. The other three enamel colours contain lead with a small percent of lead oxide (PbO). Calcium antimonate was found to be the opacifying agent in the opaque enamels (white, blue, and turquoise). The antimony content of the green glass is too low to create an opaque enamel, but the colours of the other enamels were created as follows: blue by copper and cobalt, green by copper and lead and finally turquoise by copper.

After the restoration, it was clear that the pendant still contains small reliquary packages of some sort. To visualize the contents, neutron tomography was performed at the Research Neutron Source Heinz Maier-Leibnitz (FRM II), Technical University of Munich in Garching. A combination of a position-sensitive Prompt Gamma-ray Neutron Activation Analysis and Neutron Tomography (PGAI-NT) was used to reveal the reliquary’s internal structures and content. It was surprising to see that the carbonaceous content was revealed to have five intact reliquary packages. See Figure 22.

Figure 22. Neutron tomography of the pendant with five reliquary packages. Neutrontomography by B. Schillinger, FRM II.

I was more than happy to have a look inside the pendant without having the possibility to open it! The small packages consist of tiny particles of bone covered by textile and wrapped by a thin thread to keep them together. Usually, such packages are labelled with a specific saint’s name written on attached parchment strips known as authentic or cedula. However, in the present case, no signs of parchment strips can be found in the NT images. Thanks to the neutron imaging, the exact measurements of the textiles and bones could be taken by Eschly Kluge. See Figure 23. I would never have expected that such a thing could be possible. Through the PGAI-NT, the content of the pendant could be determined by elemental analysis. Six of 29 positions stand out. See Figure 24. In comparison to reference positions and surface materials, these positions contain significant abundances of hydrogen, calcium, and potassium. See Figure 25.

Figure 23. Detailed structures of the reliquary packages (textile and splinters) on a NT-slice. Neutrontomography by B. Schillinger, FRM II.

Figure 24. Reliquary PGAI-NT positions with 29 measurement points. Neutrontomography by B. Schillinger, FRM II.

Figure 25. PGAI-NT for the elements a) calcium, b) hydrogen and c) potassium/iron. Neutrontomography by B. Schillinger, FRM II.

The additional hydrogen content along the vertical axis supports the presence of biological material. The relatively high hydrogen content at the closure fits well with the identification of beeswax within this area. The increased calcium content along the vertical axis in combination with the amounts of hydrogen is a strong indication for the presence of bone. Some of the hydrogen likely also stems from the textiles. Based on PGAI-NT’s structural and elemental analysis combination, it is therefore extremely likely that the objects within bags 2, 4 and 5 are indeed bone material. The plate in container 2 further identifies as bone by the visually prominent spongy structures in the cortex tissue. The same certainty cannot be applied for the contents of bags 1 and 3, so these objects can simply be labelled as ‘fragments’. When it comes to the components of the pendant’s frontal and back plates, it is of no surprise that the results of the PGAA coincide with those of the X-ray fluorescence spectroscopy.

Furthermore, with the aid of infrared spectroscopy, a small sample of wax from the area of the closing part was indeed identified as beeswax. The analysis of a small fragment of fibre (which had survived in the pendant’s suspension eyelet) revealed that it was made of silk. All in all, we were ecstatic to get such a great deal of information about the components of this pendant through the use of different – and importantly – non-destructive methods!

For example, here are the details of just one of the five reliquary packages numbered along the vertical symmetry axis of the pendant from top to bottom:

1. Packaging: a very fine, lose lying outer textile (vertical and horizontal thread spacing of 0.33mm and 0.26mm; 2. vertical and horizontal thread densities of 30 and 39 per cm) with an estimated spatial volume of about 2cm; 3 encloses a relatively tight-fitting and coarse-threaded textile, whose nominal thread characteristics could not be determined. Contents: an irregular tetrahedral-like splinter of osseus material with maximum extents of 7.2mm to 5.3mm to 5.1mm.

An exact description of the manufacturing technique of the pendant was possible following the NT. See Figure 26.

Figure 26. Exploded drawing of the construction components. Drawing by V. Kassühlke, LEIZA.


This quatrefoil-shaped reliquary pendant dates to the last third of the 12th century and most likely comes from a workshop in Hildesheim, Lower Saxony. This is supported by the general shape of the item as well as other characteristics which can be found in comparable objects from the region. The object presented here is one of only four phylacteria of this type known today from the Hildesheim workshop; the other three are currently located in museums in Halberstadt, Boston, and Rome. The artefact was made to contain relic particles. After completion of the container, the reliquary packages were inserted, and the reliquary was tightly closed. Through different examination methods, the various materials used could be determined. The pendant was made of gilded copper and enameled using the émail champlevé technique. The four different enamel colours were successfully analysed and determined by μ-XRF and Raman spectroscopy. The examination of a thread fragment from the area of the suspension eyelet revealed silk. Importantly, this is also the first evidence for the material on which such pendants were most likely carried. A sample of a waxy substance was identified as beeswax using infrared spectroscopy. Neutron tomography was used to visualize the contents of five relic packages. In addition, measurements of the textiles and individual splinters could be taken. The material thickness of the metal parts and the enamel fields were also determined in this way. With the additional PS-PGAA, the content could also be determined by elemental analyses.

This reliquary pendant is a rare example of such an object from a controlled excavation. It was only through the collaboration of archaeologists, art historians, chemists, neutron researchers, a nuclear physicist, restorers and goldsmiths that it was possible to arrive at these detailed results. Although this short article describes a fruitful interdisciplinary research endeavor concentrating on an exciting, once-in-a-lifetime find, one question that has not been answered after 500 hours of restoration and six years of research is why. Why was it thrown away? Was the discard intentional? Was it by mistake? Both scenarios are of course possible. Was the object hidden in a pot and thrown away without knowledge of what the pot contained? Or, consider a more entertaining option—perhaps our reliquary was lost in the dump while the archbishop of Mainz was taking a…well… Regardless, what we do know is that this reliquary is a further example of the surprising treasures that can be unearthed in old rubbish pits.

Bibliography and further reading

  • Brepohl, Erhard 1999. Theophilus Presbyter und das mittelalterliche Handwerk, Buch III, Kap. LIIII, 136-140. Köln
  • Heinzel, M., Kluge, E., Kemper, D., Schillinger, B., Stieghorst, C. 2022. New discovery of a 12th century enameled reliquary pendant – element analysis and content visualization, using Prompt Gamma Activation Analysis and Neutron Tomography (PGAI-NT). Metal 2022 – Proceedings of the Interim Meeting of the ICOM-CC Metals Working Group. Edited by Paul Mardikian, Lisa Näsänen and Ari Arponen, 184-191. Helsinki
  • Kemper, Dorothee 2020. Die Hildesheimer Emailarbeiten des 12. und 13. Jahrhunderts. Regensburg

Go back to top