The early chronology of broomcorn millet (Panicum miliaceum) in Europe.

Motuzaite-Matuzeviciute, Giedre ; Staff, Richard A. ; Hunt, Harriet V. 等

Introduction

One of the most economically important plants in prehistory was broomcom millet (Panicum miliaceum), a cereal in the same grass subfamily as maize, sorghum and foxtail millet. Today, P. miliaceum is grown as an economic plant mainly in eastern and central Asia, India, Africa, the Middle East, Eastern Europe (Russia and Ukraine) and North America. It is mostly consumed as porridge or placed in meat stews, or roasted grains are eaten with milk. In North America, it is grown primarily for animal feed, although its importance for the human food market is increasing (Graybosch & Baltensperger 2009). Broomcorn millet completes its life cycle in a very short (40-90 day) period (Nesbitt & Summers 1988), has the lowest water requirement of any cereal (Baltensperger 2002) and its grains are nutritionally more valuable than wheat, barley or rice (Rachie 1975; Baltensperger 2002; Weber & Fuller 2007). This crop therefore was and still is highly suitable for growing by semi-nomadic societies in central Asia inhabiting arid climatic regions. The prehistory of broomcorn millet in Europe is of interest because it does not belong to the Near Eastern suite of crops, and its presence in the European agricultural package therefore requires separate explanation.

In their review of records of the millet genera Panicurn and Setaria across Eurasia dating to before 5000 BC, Hunt et al (2008) highlighted the interesting and unusual pattern of the earliest records of broomcorn millet. It appeared unique among crops in being found at this early period, at both ends of the continental landmass, in Europe and in northern China. This distribution has prompted speculation and debate about the possibility of multiple domestications or unusually early cross-continental spread (de Wet & Harlan 1975; Harlan 1975; Verloove 2002; Jones 2004; Lawler 2009; Zohary et al. 2012).

The available genetic data, from microsatellite markers, lend more weight to the hypothesis of a single (Chinese) domestication rather than multiple (European and Chinese) domestications, but the evidence is still equivocal (Hunt et al 2011). Furthermore, in relation to the archaeobotanical evidence, Kreuz et al (2005) cautioned against drawing an inference of intentional cropping from individual grain finds. Also in relation to the archaeobotany, the validity of the early broomcorn millet chronology in Europe has been recently called into question by Boivin et al. (2012) as none of the Neolithic millets found in Europe has hitherto been directly dated.

In northern China, a few charred broomcorn millet grains have been directly dated from one Early Neolithic site, Xinglonggou in Inner Mongolia, producing a single date of 7670-7610 cal BP (Zhao 2011). Both botanical plant remains found in quantity and stable isotope analysis of human bones have shown that millet was a major food crop in Neolithic northern China (e.g. Cohen 1998, 2002; Zhao 2005; Crawford 2006; Barton et al 2009; Lu et al 2009; Liu et al 2012). Neolithic broomcorn millet records west of the Caucasus typically occur in small quantities, frequently only one or two grains, while the evidence for broomcorn millet becomes more prominent only in the Late Bronze Age in Europe.

The main aim of the work presented in this paper is to establish whether direct dating of broomcorn millet supports the presence of this crop in Europe before 5000 BC. If this is the case then there is a clear discrepancy between the spread of this crop and the earliest evidence from material culture of east-west contact across Eurasia, which falls at a much later period, in the second millennium BC (e.g. Sherratt 1996; Mei 2003; Kohl 2007; Frachetti 2012).

The small size of broomcorn millet grains has, until recently, precluded the possibility of direct radiocarbon ([sup.14]C) dating of individual grains. The latter constraint has now been overcome, as continued methodological refinements allow ever-smaller samples to be reliably [sup.14]C dated (Santos et al 2007). We employed these refinements to scrutinise the dates of the early western broomcorn millet grains, and thus to re-examine the chronology of its unusual prehistoric geography.

[FIGURE 1 OMITTED]

Broomcorn millet records from west Eurasian Neolithic contexts

Hunt et al. (2008) list finds from 31 sites between Germany in the west and the Caucasus in the east, the northernmost in Poland and the southernmost in Greece, for which broomcorn millet records prior to 5000 BC had been published (Table 1). Since this 2008 publication there have been further archaeobotanical finds, but none, so far as the authors are aware, significantly modify this overall time range. These records split evenly between grain impressions in ceramic items, and charred grains from archaeological sediments. The number of records per site is typically low, ranging from 1 to 97 grain impressions or 1 to 13 charred grains per site record. Whilst around a third of the site records have [sup.14]C dates from associated archaeological materials, none of the charred broomcorn millet grains have been directly dated, for the methodological reasons described above.

The records reviewed by Hunt et al. (2008) relate to work conducted up to 50 years ago, and not all of the primary material is easily accessible. Through widespread communication with active European archaeobotanists, we received broomcorn millet grains from seven archaeological sites in Europe (Figure 1) dating to the Neolithic period, relating to Linearbandkeramik (LBK) sites in Germany, the Sopot culture in Hungary, the Dudesti culture in Romania, the Butmir culture in Bosnia-Herzegovina and an Early Neolithic site in Bulgaria (Table 2). With the exception of the Fajsz 18 site in Hungary, all of these site records have been published, and some cited by Hunt etal. (2008) reportedly date to before 5000 BC. The details of each record are summarised in Table 1.

Three broomcorn millet samples from the German sites of Weterau, Fechenheim and Goddelau are associated with LBK pottery attributed to the second half of the sixth millennium BC (Kreuz et al. 2005; Kreuz & Schafer 2011). The grains were recovered from sealed pits. Together with the solitary broomcorn millet grains, a large quantity of wheat and barley grains, including a range of weeds, has been found. Consequently, the broomcorn millet grains were themselves interpreted as weeds (Kreuz et al. 2005).

Further broomcorn millet grains were available from the Fajsz 18 site in Hungary, in association with artefacts of the Late Neolithic Sopot culture (5000-4700/4600 BC). These grains too were found in a secure archaeological context, together with the Neolithic pottery and tools. The Fajsz 18 site also contains Bronze Age levels above the Neolithic levels.

Two broomcorn millet grains were chosen for dating from the Magura-Buduiasca site in southern Romania. The occupation levels bearing the broomcorn millet samples belong to the Dudesti culture of the sixth millennium BC. Previous AMS-dating of three barley grains and one einkorn grain from pit 13A (Starcevo-Cris) yielded dates in the early sixth millennium cal BC, while a fourth barley grain dated to the later sixth millennium cal BC. No dates, however, were obtained from the context containing the broomcorn millet remains. In total, seven grains of broomcorn millet were found in Early and Late Dudesti culture pits from the site (Bogaard & Walker 2011).

Twelve broomcorn millet grains, among thousands of wheat grains, were recovered from a tell site at Okoliste in the hilly region of central Bosnia-Herzegovina. They were found in association with material of the Late Neolithic Butmir culture, which corresponds to the period between 5200 and 4500 ca] BC (Muller-Scheessbel et al. 2010).

The final series of dated broomcorn millet grains came from the Yabalkovo Early Neolithic site in Bulgaria (beginning of the sixth millennium BC), which also contains some scattered occupation from the Bronze Age. The age of the site was previously determined by [sup.14]C dating of charred cereal grains from the same archaeological context as the broomcorn millet grains which are dated and presented in Table 2 (T. Popova pers. comm.). The millet grains (over 20 in total) were found in a pit together with other crops, principally wheat and barley (Leshtakov et al. 2007; Popova 2010).

Material and methods

Owing to the low starting weights of the individual broomcorn millet grains submitted to the Oxford Radiocarbon Accelerator Unit (ORAU) for radiocarbon dating, a less rigorous chemical pre-treatment methodology was applied than usual. This involved ultrasonication of samples in 1 M HC1 and ultrapure water only, rather than the more robust acid-base-acid (ABA) method more routinely applied to plant macrofossil samples (Brock et al. 2010). The fact that the millet grains presented in this paper appeared to be in a good state of preservation, and physically 'clean', gives us confidence that, despite the less rigorous pretreatment, the [sup.14]C measurements obtained remain reliable. After subsequent freeze-drying and combustion, graphitisation took place with the addition of the desiccant magnesium perchlorate in the water trap of the reactor rigs, to optimise the conversion of C[O.sub.2] to graphite. The resulting graphite was pressed into aluminium targets for accelerator mass spectrometry (AMS) radiocarbon dating (see online supplementary information for further details).

Results of direct dating

Panicum miliaceum, like other millets, photosynthesises using the Hatch-Slack ('C4') pathway. During photosynthesis, C4 plants discriminate against [sup.13]C less than 'C3' plants (plants utilising the most common photosynthetic pathway), and hence have higher [delta][sup.13]C values than C3 plants (approximately-12.5[per thousand] in C4 plants, compared with approximately -26.5[per thousand] in C3 plants) (van der Merwe 1982; Tieszen 1991). From the [delta][sup.13.C] values obtained here (between -11.16[per thousand] and-9.57[per thousand]), we can be confident that all of the dated samples utilise the C4 pathway (Table 2). Vegetation in temperate Eurasia is predominantly C3, as are the majority of cultivated plants. Before the post-Columbian introduction of maize from the Americas, the millets were the C4 species most heavily used in Eurasia. Although not conclusive of species identification, the [delta][sup.13]C values obtained therefore corroborate the identification of the samples as millet.

All 10 seeds submitted for AMS [sup.14]C dating provided radiocarbon measurements, even where the weight prior to pre-treatment was as small as 0.92mg (Table 2; Figure 2). None of the dates obtained is older than 5000 BC; most of the dates are from the middle of the second millennium BC, ranging from 1606 to 417 cal BC and later, and some, such as from the Magura-Buduiasca site in Romania, are just a few centuries old. The group of radiocarbon dates from the LBK sites in Germany correspond to the Middle Bronze Age and the Early Iron Age of Central Europe: sample OxA-26700 (from Bruchenbrucken/Friedberg) dates to 1505-1386 cal BC (all calibrated [sup.14]C ages are given at 95.4%, i.e. 2[sigma] probability). OXA-26701 (from Fechenheim/Frankfurt) dates to 1055-851 cal BC, and OXA-26702 (from Godddau/Riedstadt) dates to 772-417 cal BC. The dates from Hungary (Fajsz 18) are the next oldest, and correspond to the Late Bronze Age (OXA-26703:1428-1262 cal BC, and OXA-26704:1606-1414 cal BC). One date from Romanian (Magura-Buduiasca) broomcorn millet is similar to the Hungarian data and is attributed to the Late Bronze Age (OXA-26706:1434-1268 cal BC). The remainder of the samples, those from Bosnia-Herzegovina, Bulgaria and one from Romania, are attributed to the centuries AD of medieval and post-medieval Europe.

Discussion

The new data obtained by using the methodological refinements for small mass samples resulted in direct [sup.14]C dating of 10 broomcorn millet grains. The dates indicate that the chronology previously proposed for the substantial number of Central and Eastern European broomcorn millet macrofossils was too early by at least 3500 years.

A possible explanation of the dates recorded in Table 2 is that the small broomcorn millet grains have repeatedly moved downwards through stratigraphic sequences, giving the spurious impression of an early date. The site of Fajsz 18 provides a plausible example, with Bronze Age levels deposited directly above Neolithic levels. Most sites also contain later occupations not directly above the Neolithic one but in a neighbouring area, and thus the land surface above the Neolithic levels could have been part of the field system in these later periods.

[FIGURE 2 OMITTED]

That explanation is more difficult to apply, however, where later stratigraphic levels are absent. The Yabalkovo site, where a pit contained cereal grains of a range of species, is particularly enigmatic; our first millennium AD dates on millet grains place them some seven millennia later than the early sixth millennium BC dates for the wheat and barley.

This substantial revision of the macrofossil chronology raises the question of the reliability of the other source of pre-5000 BC broomcorn millet evidence: that from grain impressions in pottery. In these cases, the dating of the impressions is as secure as the dating of the ceramic typologies, and it seems unlikely that the age of such well-studied ceramic groups as the LBK will be shifted by millennia. However, the question in the case of impressions is whether the identification of the small voids as casts of broomcorn millet is secure. We have not had the opportunity of re-examining the impression-bearing ceramic fragments, so we offer the following observations in relation to future scrutiny of small grain impressions.

Yanushevich, who identified a sizeable proportion of the broomcorn millet impressions from Moldova and Ukraine, notes that the impressions in pottery and daub at the Neolithic and Early Chalcolithic sites in Moldova are not very clear, and that "possibly some of them belong to Setaria glauca, Setaria viridis or Echinochloa crus-galli species" (Yanushevich 1976: 153). In Moldova, Yanushevich seems to have taken the shape and size of broomcorn millet spikelet imprints with glumes and lemmas as the main criteria for identification. It has been suggested, however, that the scutellum, grain dimensions and shape offer more reliable criteria for identification (Fuller 2006; Motuzaite-Matuzeviciute et al. 2012). It would therefore be useful to re-examine the early impressions identified as broomcorn millet to establish whether any of these key criteria are discernible, which will depend amongst other things on the presence/absence of glumes potentially obscuring these features.

The increase in quantity and frequency of broomcorn millet in the west Eurasian archaeological record

By the Late Bronze Age in Europe, stable isotope analysis on human bone collagen at some sites in southern Lithuania and northern Italy has indicated significant C4 plant input into the human diet (Antanaitis & Ogrinc 2000; Tafuri et al. 2009). During the same period, the quantity and frequency of broomcorn millet increase across Europe (e.g. Marinval 1992; Pashkevich 2003; Jacomet 2004; Tafuri et al. 2009).

Records of broomcorn millet become more frequent and the grains are found in higher quantities at sites in Eastern and Central Europe by around 3000 BC. Lumps of charred broomcorn millet were found in a Polish Funnel Beaker site (Wasylikowa et al. 1991). Eighty-three millet grains were reported from six contexts within the burned house at the Kleiner Anzingerberg site of the Jevisovice culture in Austria and have been dated to 3200-2800 BC (Kohler-Schneider & Caneppele 2009). Charred broomcorn millet macrofossils have been frequently reported from the Baden culture in Slovakia and dated to 3600-2900 BC (Hajnalova 1989). Clay temper in 70 cult statuettes from Usatovo culture sites, dated to 3600-3000 BC, as well as pottery vessels and daub, contained many impressions of broomcorn millet grains (Kuzminova & Petrenko 1989).

The Eurasian pattern

The direct dating of broomcorn millet grains from west Eurasian Neolithic contexts has produced estimates considerably younger than dates associated with the contexts in which they were found. The most parsimonious explanation is that the vertical movement of small grains through the stratigraphic profile has led to a spurious impression of greater antiquity. This certainly calls into question half of the site records reviewed by Hunt et al. (2008), though without resolving how younger grains might enter older archaeological contexts in every case. We have, in addition, suggested how the other set of early broomcorn millet records, those of impressions in ceramics, might be further scrutinised.

Until these questions are satisfactorily answered, we cannot exclude the possibility that some of the reported pre-5000 BC broomcorn millet records in Europe are correct. Nonetheless, our direct dating of a small but substantial set of charred macrofossils indirectly questions the Early Neolithic attributions of the remaining broomcorn millet records across Europe.

As mentioned above, there is considerably more evidence for the presence of broomcorn millet in western Eurasia from the second half of the fourth millennium BC onwards, especially in the Carpathian region and southern Ukraine, where records start to become more ubiquitous and grains present in higher quantities (Hajnalova 1989; Kuzminova & Petrenko 1989; Pashkevich 2003; Kohler-Schneider & Caneppele 2009). Nevertheless, this evidence too would also benefit from confirmation by direct [sup.14]C dating, and that will be the subject of future work.

In summary, the new dating results have cast significant doubt upon the earliest records of Panicum miliaceum in the west. The apparent association of small numbers of millet grains with archaeological contexts associated with a pre-5000 BC date has repeatedly been called into question. The direct dates of these particular grains brings them much closer into line with the earliest evidence for shared metallurgical techniques across Eurasia (e.g. Sherratt 1996; Mei 2003). There remains, however, a significant body of published evidence for a presence by the third millennium BC of Panicum miliaceum in the archaeobotanical record, and of C4 plants in some western diets. These records may prove to be more robust than the very earliest millet records re-examined in this paper. Subject to scrutiny of these later records, the re-dating reported in this paper greatly reduces the chronological gap between the earliest dates for crop contact and metallurgical contact across Eurasia, but that gap as yet remains.

Acknowledgements

We would like to acknowledge the Leverhulme Trust for funding the 'Pioneers of Pan-Asian Contact' project, the European Research Council for funding the 'Food Globalisation in Prehistory' project, and the Research Council of Lithuania for funding postdoctoral research. We would like to thank all of the archaeobotanists who kindly provided their broomcorn millet samples for dating: Angela Kreuz, Helmut Kroll, Amy Bogaard, Tzvetana Popova and Peter Pomazi; thanks also to all the staff at the Oxford Radiocarbon Accelerator Unit for their team efforts in dating the material.

References

ANTANAITIS, I. & N. OGRINC. 2000. Chemical analysis of bone: stable isotope evidence of the diet of Neolithic and Bronze Age people in Lithuania. Istorija 45:3-12.

BALTENSPERGER, D.D. 2002. Progress with proso, pearl and other millets, in J. Janick & A. Whipkey (ed.) Trends in new crops and new uses: 1-9. Alexandria (VA): American Society for Horticultural Science Press.

BARTON, L., S.D. NEWSOME, EH. CHEN, H. WANG, T.P. GUILDERSON & R.L. BETTINGER. 2009. Agricultural origins and the isotopic identity of domestication in northern China. Proceedings of the National Academy of Sciences of the USA 106: 5523-28.

BOGAARD, A. & A. WALKER. 2011. Plant use and management at Magura-Buduiasca (Teleor 003), southern Romania: preliminary report on the archaeobotanical analysis. Report prepared for the European Union, Brussels.

BOIVIN, N., D.Q. FULLER & A. CROWTHER. 2012. Old World globalization and the Columbian exchange: comparison and contrast. World Archaeology 44: 452-69.

BROCK, E, T. HIGHAM, P. DITCHFIELD & C. BRONK RAMSEY. 2010. Current pre-treatment methods for AMS radiocarbon dating at the Oxford Radiocarbon Accelerator Unit (ORAU). Radiocarbon 52: 103-12.

COHEN, D.J. 1998. The origins of domesticated cereals and the Pleistocene-Holocene transition in East Asia. The Review of Archaeology 19: 22-29.

--2002. New perspectives on the transition to agriculture in China, in Y. Yasuda (ed.) The origins of pottery and agricuhure: 217-27. New Delhi: Lustre & Roll

CRAWFD, G. 2006. East Asian plant domestication, in S.T. Stark (ed.) Archaeology of Asia: 77-95. Oxford: Blackwell.

DE WET, J.M.J. & J.R. HARLAN. 1975. Weeds and domesticates: evolution in the man-made habitat. Economic Botany 29: 99-108.

FRACHETT1, M.D. 2012. Multiregional emergence of mobile pastoralism and nonuniform institutional complexity across Eurasia. Current Anthropology 53: 2-38.

FULLER, D.Q. 2006. A millet adas: some identification guidance. Report prepared for University College London. Available at: http://www.homepages.ucl.ac.uk/~tcrndfu/Abot/Millet%20Handout06.pdf (accessed 30 August 2013).

GRAYBOSCH, R.A. & D.D. BALTENSPERGER. 2009. Evaluation of the waxy endosperm trait in proso millet (Panicum miliaeeum). Plant Breeding 128: 70-73.

HAJNALOVA, E. 1989. Katalog zvyskov semien a plodov v archeologickych nalezoch na Slovensku. Acta Interdisciplinaria Archeologica 6: 3-192,

HARLAN, J.R. 1975. Crops and man. Madison (WI): American Society of Agronomy.

HUNT, H.V., M. VANDER LINDEN, X. LIU, G. MOTUZAITE-MATUZEV1CIUTE, S. COLLEDGE & M.K. JONES. 2008. Millets across Eurasia: chronology and context of early records of the genera Panicum and Setaria from archaeological sites in the Old World. Vegetation History and Archaeobotany 17: 5-18.

HUNT, H.V., M.G. CAMPANA, M.C. LAWES, Y.J.I.N. PARK, M.A. BOWER, C.J. HOWE & M.K. JONES. 2011. Genetic diversity and phylogeography of broomcorn millet (Panicum miliaceum L.) across Eurasia. Molecular Ecology 20: 4756-71.

JACOMET, S. 2004. Archaeobotany. A vital tool in the investigation of lake-dwellings, in E Menotti (ed.) Living on the lake in prebistoric Europe. 150years of lake dwelling research: 162-77. London: Routledge.

JONES, M.K. 2004. Between Fertile Crescents: minor grain crops and agricultural origins, in M.K. Jones (ed.) Traces of ancestry: studies in honour of Colin RenJ%w: 127-35. Cambridge: McDonald Institute for Archaeological Research.

KOHL, EL. 2007. The making of Bronze Age Eurasia. Cambridge: Cambridge University Press.

KOHLER-SCHNEIDER, M. & A. CANEPPELE. 2009. Late Neolithic agriculture in eastern Austria: archaeobotanical results from sites of the Baden and Jevigovice cultures (3600-2800 BC). Vegetation History and Archaeobotany 18:61-74.

KREUZ, A. & E. SCHAFER. 2011. Weed finds as indicators for the cultivation regime of the early Neolithic Bandkeramik culture? Vegetation History and Archaeobotany 20: 333-48.

KREUZ, A., E. MARINOVA, E. SCI4KFER & J. WIETHOLD. 2005. A comparison of early Neolithic crop and weed assemblages from the Linearbandkeramik and the Bulgarian Neolithic cultures: differences and similarities. Vegetation History and Archaeobotany 14: 237-58.

KUZMINOVA, N.N. & V.G. PETRENKO. 1989. Kulturnye rasteniya na zapade Stepnogo Prichernomorya v seredine 3-2 tis. do n. e. (po dannym paleobotaniki), in RE Tolochko (ed.) Problemu Drevnei Istorii i Arkheologii Ukrainskoi SSR: 119-20. Kiev: Naukova Dumka.

LAWLER, A. 2009. Millet on the move. Science 325: 942-43.

LESHTAKOV, K., N. TODOROVA, V. PETROVA, R. ZLATEVA-UZUNOVA, O. OZBEK, T. POPOVA, N. SPASSOV & N. ILIEV. 2007. Preliminary report on the salvage archaeological excavations at the Early Neolithic site Yabalkovo in the Maritsa Valley. Anatolica 33:185-234.

LIE, X., M.K. JONES, Z. ZHAO, G. LIU & T.C. O'CONNELL. 2012. The earliest evidence of millet as a staple crop: new light on Neolithic foodways in north China. American Journal of Physical Anthropology 149: 283-90.

Lu, H., J. ZHANG, K. Llu, N. Wu, Y. LI, K. ZHOU, M. YE, T. ZHANG, H. ZHANG & X. YANG. 2009. Earliest domestication of common millet (Panicum miliaceum) in East Asia extended to 10,000 years ago. Proceedings of the National Academy of Sciences of the USA 106: 1-6.

MARINVAL, P. 1992. Archaeobotanical data on millet (Panicum miliaceum and Setaria italica) in France. Review of Palaeobotany and Palynology 73: 259-70.

MEI, J. 2003. Qijia and Seima-Turbino: the question of early contacts between northwest China and the Eurasian steppe. Bulletin of the Museum of Far Eastern Antiquities 75: 31-54.

MOTUZMTE-MATUZEVICIUTE, G., H.V. HUNT & M.K. JONES. 2012. Experimental approaches to understanding variation in grain size in Panicum miliaceum (broomcorn millet) and its relevance for interpreting archaeobotanical assemblages. Vegetation History and Archaeobotany 21: 69-77.

MULLER-SCHEESSEL, N., R. HOFMANN, J. MOLLER & K. RASSMANN. 2010. The socio-political development of the Late Neolithic settlement of Okoliste/Bosnia-Hercegowina: devolution by transhumance?, in Kiel Graduate School (ed.) Landscapes and human development: the contribution of European archaeology: 181-91. Bonn: Rudolf Habelt.

NESBITT, M. & G.D. SUMMERS. 1988. Some recent discoveries of millet (Panicum miliaceum L. and Setaria italica (L.) E Beauv.) at excavations in Turkey and Iran. Anatolian Studies 38: 85-97.

PASHKEVICH, G.A. 2003. Palaeoethnobotanical evidence of agriculture in steppe and forest-steppe of East Europe in the Late Neolithic and Bronze Age, in M. Levine, C. Renfrew & K. Boyle (ed.) Prehistoric steppe adaptation and the horse: 287-97. Cambridge: McDonald Institute for Archaeological Research.

POPOVA, T. 2010. Plant environment of man between 6000 and 2000 BC in Bulgaria (British Archaeological Reports international series 2064). Oxford: Archaeopress.

RACHIE, K.O. 1975. Millets. Importance, utilization and outlook. Hyderabad: International Crops Research Institute for the Semi-Arid Tropics (ICRISAT).

REIMER, P.J., M.G. BAILLIE, E. BARD, A. BAYLISS, J.W. BECK, P.G. BLACKWELL, C. BRONK RAMSEY, C.E. BUCK, G.S. BURR, R.L. EDWARDS, M. FRIEDRICH, P.M. GROOTES, T.P. GUILDERSON, I. HAJDAS, T.J. HEATON, A.G. HOGG, K.A. HUGHEN, K.F. KAISER, B. KROMER, EG. MCCORMAC, S.W. MANNING, R.W. REIMER, D.A. RICHARDS, J.R. SOUTHON, S. TALAMO, C.S.M. TURNEY, J. VAN DER PLICHT & C. WEYHENMEYER. 2009. IntCal09 and Marine09 radiocarbon age calibration curves, 0-50,000 years ca] BE Radiocarbon 51:1111-50.

SANTOS, G.M., J.R. SOUTHON, S. GRIFFIN, S.R. BEAUPRE & E.R.M. DRUFFEL. 2007. Ultra small-mass AMS [sup.14]C sample preparation and analyses at KCCAMS/UCI Facility. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 259: 293-302.

SHERRATT, A. 1996. Plate tectonics and imaginary prehistories: structure and contingency in agricultural origins, in D.R. Harris (ed.) The origins and spread of agriculture and pastoralism in Eurasia: 13041. Washington, D.C.: Smithsonian Institution.

STUIVER, M. & H.A. POLACH. 1977. Discussion: reporting of [sup.14]C data. Radiocarbon 19: 355-63.

TAFURI, M.A., O.E. CRAIG & A. CANCI. 2009. Stable isotope evidence for the consumption of millet and other plants in Bronze Age Italy. American Journal of Physical Anthropology 139: 146-53.

TIESZEN, L.L. 1991. Natural variations in the carbon isotope values of plants: implications for archaeology, ecology and paleoecology. Journal of Archaeological Science 18: 227-18.

VAN DER MERWE, N.J. 1982. Carbon isotopes, photosynthesis, and archaeology: different pathways of photosynthesis cause characteristic changes in carbon isotope ratios that make possible the study of prehistoric human diets. American Scientist 70: 596-606.

VERLOOVE, E 2002. A revision of the genus Panicum (Poaceae, Paniceae) in Belgium. Systematics and Geography of Plants 71: 53-72.

WASYLIKOWA, K., M. CARCIUMARU, E. HAJNALOVA, B.P. HARTYANYI, G.A. PASHKEVICH & Z.V. YANUSHEVICH. 1991. East-central Europe, in W. Van Zeist, K. Wasylikowa & B. Karl-Ernst (ed.) Progress in Old World palaeoethnobotany A retrospective view on the occasion of 20 years of the International Work Group for Palaeoethnobotany: 207-39. Rotterdam: A.A. Balkema.

WEBER, S.A. & D.Q. FULLER. 2007. Millets and their role in early agriculture. Paper presented at the First Farmers in Global Perspectives conference, Lucknow, India, 18-20 January 2006.

YANUSHEVICH, Z.V. 1976. Kulturnye rasteniya yugo-zapada SSSR po paleobotanicheskim issledovaniyam. Kishinev: Shtiintsa.

ZHAO, Z. 2005. Zhiwu kaoguxue jiqi xin jinzhan. Kaogu 7: 42-49.

--2011. New archaeobotanic data for the study of the origins of agriculture in China. Current Anthropology 52: 295-306.

ZOHARY, D., M. HOPF & E. WEISS. 2012. Domestication of plants in the Old World: the origin and spread of domesticated plants in southwest Asia, Europe, and the Mediterranean basin. Oxford: Oxford University Press.

Received: 5 February 2013; Accepted: 1 May 2013; Revised: 3 June 2013

Supplementary material is provided online at http://antiquity.ac.uk/ProjGall/motuzaitematuzeviciute33 8/

Giedre Motuzaite-Matuzeviciute (1,2), Richard A. Staff (3), Harriet V. Hunt (1), Xinyi Liu (1) & Martin K. Jones (4)

(1) McDonald Institute for Archaeological Research, University of Cambridge, Downing Street, Cambridge CB2 3ER, UK (Emaik gm327@cam.ac,uk; x1241 @cam.ac.uk; hvh22@cam,ac.uk)

(2) History Faculty/Department of Archaeology, Vilnius University, Universiteto 7, 01513 Vilnius, Lithuania (Emaik Giedre.keen@if.vu.lt)

(3) Oxford Radiocarbon Accelerator Unit (ORAU), Research Laboratory for Archaeology and the History of Art (RLAHA), University of Oxford, Dyson Perrins Building, South Parks Road, Oxford OX1 3QY, UK (Email: richard.staff@rlaha.ox.ac.uk)

(4) Department of Archaeology, University of Cambridge, Downing Street, Cambridge CB2 3DZ, UK (Email: mkjl2@cam.ac.uk)

Table 1. Sites in Europe (including the Caucasus) with
archaeobotanical evidence for Panicum miliaceum prior to 5000 BC
(modified after Hunt et al. 2008). Dating rationale: a) grains
with material culture association; b) grains with associated
[sup.14]C date; c) impressions in ceramics.

Country
(region/province) Period Culture

Azerbaijan Late Neolithic Shulaveri-
 Shomutepe

Bulgaria (south-west) Late Neolithic

Czech Republic Early Neolithic LBK (Rubane)
 (northern
 Bohemia)

Czech Republic Early Neolithic LBK/
 (central Bohemia) Stichbandkeramik/
 Lengyel?

Czech Republic Early Neolithic LBK/Moravian
 (northern Moravia) Painted Pottery?

Germany Early Neolithic LBK
 (south-west,
 Hessen)

Germany Early Neolithic LBK
 (north-central,
 Harz mountains)

Germany Early Neolithic LBK
 (south-west,
 Hessen)

Germany (south, Early Neolithic LBK
 Bayern)

Georgia Late Neolithic Shulaveri-
 Shomutepe

Georgia Late Neolithic Shulaveri-
 Shomutepe

Georgia Late Neolithic Shulaveri-
 Shomutepe

Greece (Thessaly) Early Neolithic Protosesklo

Greece (Thessaly) Middle Neolithic Sesklo

Moldova (Prut river) Early Neolithic LBK

Moldova (central) Neolithic LBK

Moldova (north) Early Neolithic LBK/Cris

Moldova Early Neolithic LBK
 (east-central)

Moldova (north-east) Early Neolithic Bug-Dniester/Cris

Moldova (north-east) Early Neolithic Bug-Dniester

Poland (south-east) Early Neolithic LBK

Romania Neolithic Vinca

Slovakia (south-east) Early Neolithic (eastern) LBK
 (Bukk)

Slovakia (east) Early Neolithic (eastern) LBK
 (Bukk)

Slovakia (south-west) Early Neolithic LBK

Ukraine Early Neolithic LBK
 (west-central)

Ukraine Early Neolithic Bugo-Dniestr
 (west-central)

Ukraine (central) Middle Neolithic Kievo-Cherkasskaya

Ukraine (north-west) Middle Neolithic Volynskaya

Ukraine (north-west) Early Neolithic Volynskaya

Ukraine (north-west) Early Neolithic Volynskaya

Country Dating information
(region/province) Site name cal BC

Azerbaijan Kjultepe fifth-fourth
 millennia

Bulgaria (south-west) Drenkovo-Ploshteko late sixth-early
 fifth millennia

Czech Republic Brezno u Louny second half of sixth
 (northern millennium
 Bohemia)

Czech Republic Bylany 5400-4300 (9 dates)
 (central Bohemia)

Czech Republic Mohelnice 5600-5000 (6 dates)
 (northern Moravia)

Germany Bruchenbrucken 5200-4800 (1 date)
 (south-west,
 Hessen)

Germany Eitzum-II 5500-4700 (3 dates)
 (north-central,
 Harz mountains)

Germany Goddelau 5700-5100 (5 dates)
 (south-west,
 Hessen)

Germany (south, Mintraching second half of sixth
 Bayern) millennium

Georgia Arukhlo 1 fifth-fourth
 millennia

Georgia Dikhi-Gudzuba fifth-fourth
 millennia

Georgia Imiris-gora fifth-fourth
 millennia

Greece (Thessaly) Argissa Magoula 6500-6200 (1 date)

Greece (Thessaly) Otzaki Magoula first half of sixth
 millennium

Moldova (Prut river) Denchen-I second half of sixth
 millennium

Moldova (central) Durlesht-I second half of sixth
 millennium

Moldova (north) Sakarovka-I second half of sixth
 millennium

Moldova Braneshry-I 5400-5000
 (east-central)

Moldova (north-east) Ruptura 5976-5560

Moldova (north-east) Soroki-I 6000-4800/4700

Poland (south-east) Olszanica 7000-4200 (8 dates)

Romania Liubcova second half of sixth
 millennium

Slovakia (south-east) Domica Cave 5200-4800 (1 date)

Slovakia (east) Sarisske second half of sixth
 Michal'any II millennium

Slovakia (south-west) Sturovo 5500-4800 (2 dates)

Ukraine Rovno 5629-5306 (2 dates)
 (west-central)

Ukraine Sokoltsy II 6438-6101 (2 dates)
 (west-central)

Ukraine (central) Grini 5200-4250

Ukraine (north-west) Krushniki 5100-3850

Ukraine (north-west) Mala Osnitsa 5450-5100

Ukraine (north-west) Obolon 5450-5100

Country Nature of millet find Dating
(region/province) (number of grains) rationale

Azerbaijan

Bulgaria (south-west) grain (4) a

Czech Republic grain (13) a
 (northern
 Bohemia)

Czech Republic grain b
 (central Bohemia)

Czech Republic grain b
 (northern Moravia)

Germany grain (1) b
 (south-west,
 Hessen)

Germany grain (2) b
 (north-central,
 Harz mountains)

Germany grain (1) b
 (south-west,
 Hessen)

Germany (south, grain (1) a
 Bayern)

Georgia - -

Georgia - -

Georgia - -

Greece (Thessaly) grain (1) b

Greece (Thessaly) grain a

Moldova (Prut river) 60 imprints in pottery c

Moldova (central) 1 imprint in pottery c

Moldova (north) 97 imprints in pottery c

Moldova 1 imprint c
 (east-central)

Moldova (north-east) 1 imprint c

Moldova (north-east) 1 imprint c

Poland (south-east) grain (5) c

Romania grain -

Slovakia (south-east) grain b

Slovakia (east) grain (4) a

Slovakia (south-west) grain (1) b

Ukraine 2 impressions in pottery c
 (west-central)

Ukraine 1 impression in pottery c
 (west-central)

Ukraine (central) 3 impressions in pottery c

Ukraine (north-west) 2 impressions in pottery c

Ukraine (north-west) 1 impression in pottery c

Ukraine (north-west) 1 impression in pottery c

Table 2. Archaeological contexts of the 10 broomcorn millet samples,
along with their [sup.14]C (uncalibrated and calibrated) and
[delta][sup.13]C's measurements. Sample weights (initial starting
weight and combustion yield following chemical pre-treatment) are
given to emphasise radiocarbon's ability to date ever smaller
[sup.14]C samples (see online supplementary Table S1 for
additional sample data).

 Expected culture ORAU lab.
Site Country and period code

Bruchenbrucken/ Germany Bandkeramik OxA-26700
 Friedberg (Neolithic
 5500-4500 BC)

Fechenheim/ Germany Bandkeramik OxA-26701
 Frankfurt (Neolithic
 5500-4500 BC)

Goddelau/ Germany Bandkeramik OxA-26702
 Riedstadt (Neolithic
 5500-4500 BC)

Fajsz 18 Hungary Sopot (Late OxA-26703
 Neolithic
 5500-4500 BC)

Fajsz 18 Hungary Sopot (Late OxA-26704
 Neolithic
 5500-4500 BC)

Okoliste Bosnia- Butmir (5500-4800 OxA-X-2479-22
 Herzegovina BC)

Yabalkovo Bulgaria Early Neolithic OxA-26705

Yabalkovo Bulgaria Early Neolithic OxA-26477

Magura- Romania Dudesti (Neolithic OxA-26706
 Buduiasca sixth mill. BC)

Magura- Romania Dudesti (Neolithic OY-A-26707
 Buduiasca sixth mill. BC)

 Starting Combustion
 weight yield [delta][sup.13]C
Site (mg) (mg C) ([per thousand])

Bruchenbrucken/ 1.36 0.429 -10.05
 Friedberg

Fechenheim/ 2.65 0.581 -10.63
 Frankfurt

Goddelau/ 0.92 0.464 -9.59
 Riedstadt

Fajsz 18 1.03 0.361 -11.16

Fajsz 18 1.21 0.413 -9.87

Okoliste 1.62 0.528 -11.04

Yabalkovo 1.42 0.311 -9.57

Yabalkovo 2.09 1.106 -9.75

Magura- 1.57 0.441 -9.66
 Buduiasca

Magura- 1.23 0.592 -10.41
 Buduiasca

 Conventional
 [sup.14]C age Calibrated age
 BP ([+ or -] (cal BC/AD,
Site 1[sigma]) 95.4% hpd range)

Bruchenbrucken/ 3163 [+ or -] 33 1505-1386 BC
 Friedberg

Fechenheim/ 2815 [+ or -] 32 1055-851 BC
 Frankfurt

Goddelau/ 2484 [+ or -] 34 772-417 BC
 Riedstadt

Fajsz 18 3075 [+ or -] 36 1428-1262 BC

Fajsz 18 3214 [+ or -] 36 1606-1414 BC

Okoliste 1740 [+ or -] 130 * AD 4-576

Yabalkovo 1128 [+ or -] 35 AD 781-991

Yabalkovo 1176 [+ or -] 28 AD 774-953

Magura- 3093 [+ or -] 35 1434-1268 BC
 Buduiasca

Magura- 398 [+ or -] 26 AD 1438-1620
 Buduiasca

Note: [delta][sup.13]C ([per thousand]) data are relative to the
Vienna Pee Dee Belemnite standard. The conventional [sup.14]C age
BP ([+ or -]1[sigma]) data and a fractionation correction was
calculated as per Stuiver & Polach 1977. With reference to the
calculated age, 'hpd' is highest probability density.

* This sample has a higher than usual uncertainty owing to its
low AMS target current. See online supplementary information
(footnote [dagger]) for further details.