ARCHAEOLOGY INTERNATIONAL

), four were pulses (, Early in Southwest Asia , and bitter vetch), and one was a fibre (flax). It is thought that and : re-examining the the were domesticated first and archaeobotanical evidence that the other species evolved at about the same time or possibly somewhat later.8 Sue Colledge & J ames Conolly Our knowledge of the evolution and Agriculture is widely recognized as a defining characteristic of spread of the depends on the examination and accurate identifica­ th e Ne olithic period in Southwest Asia and Europe, but, despite tion of charred grains found in samples many years of research, and th e discoveryof much new arch a eo­ recovered fromsecurely dated occupation botanical evidence, there have been fe w attempts to investigate levels at early Neolithic sites. The reported its origins and spread in th e region as a wh ole. No w, in a new presence, or absence, at different localities project at th e In stitute of Archaeology, th e scattered evidence for and times of domestic species (when cor­ rectly identified as such) has provided the th e emergence and dispersal of crops is being systematically evidence for our understanding of the dis­ assessed and documented both sp atially and chronologically. tribution of the earliest crops and of their subsequent dispersal throughout South­ west Asia. But this procedure is not with­ ome 40 years ago the pioneer and humidity. Although the effects on the out its pitfalls, as the example of archaeobotanist Hans Helbaek landscape of these changes were most shows. emphasized the importance of noticeable in northern latitudes, they were S correctly dated plant remains as a also apparent to a lesser degree world­ Jericho revisited means of investigating the rela­ wide. The impact of the cold spells, in Very few PPNA sites in Southwest Asia tionship between "man and nature",1 and particular, had profound consequences have yielded evidence of domestic cereals; he later developed a technique for separat­ for how people adapted to, and interacted to varying degrees, the finds fromthem are ing charred plant materials from the sedi­ with, their environments. controversial. The domestic wheat and ments in which they had been preserved.2 Semi-sedentary Natufian groups of found in the PPNA levels at Jericho Although many archaeobotanical investi­ hunter-gatherers inhabited much of the are no exception to this. gations have been carried out since then at Levant6 when the climate was at its warm­ The large mound at Jericho - a about Neolithic sites in Southwest Asia and est and wettest during the so-called cli­ 20m high, 320m long and 140m wide ­ Europe, we still lack a database that is matic optimum some 12,500 years before represents several thousand years of occu­ sufficiently large, systematic and well present (c. 12,500 bp). They benefited pation. Kathleen Kenyan excavated at Jeri­ founded to enable us to gain a better from increased local availability of the cho in the early 1950s when she was a understanding of the emergence and plant foods they had come to rely on, such member of staff of the Institute, and in the spread of agriculture in the region. as the grains of wild grasses, as is largest of her trenches she dug down to the In July 2001 a new three-year project3 indicated by mortars and pestles for grind­ earliest occupation layers of the settle­ was started at the Institute with the aims of ing them and other possible processing ment and exposed the full extent of a mas­ collecting, assessing and systematically found at Natufian sites. About 1500 sive stone-built tower and walls that were recording in a chronological framework all years later, the Natufians experienced a firstconstructed during the PPNA (Fig. 1).9 the available data on the plant remains sudden change of climate to much colder Together with fragments of burnt plaster (mainly charred ) that have been and drier conditions (a period referred to recovered from early Neolithic and Epi­ as the YoungerDryas, c. 11,000 to c. 10,000 palaeolithic or sites in the bp), which caused their supplies of wild region.4 In this article, we describe the plant fo ods to diminish. It is probable that, background to the project and outline as a direct result of this change, they some of our preliminary findings. After attempted some form of management of only a few months work, and having not the wild that had become staples in yet begun to consider any early Neolithic their diet. Thus, it is thought, the first sites in Europe, we have surprising new attempts to cultivate wild cereals took evidence indicating that at least one site place. By sowing the wild grains in tilled (Jericho), hitherto accepted as providing fields, the Natufians ensured that they had evidence of some the earliest domesti­ annual harvests sufficient for their needs cated crops, has been inaccurately inter­ and they gained some control over the sup­ preted. We have therefore to reconsider plies of plant foods regardless of climatic the questions of where and when crops instability. Domestic species of cereals first appear in the archaeological record of evolved as a result of the cultivation of the Southwest Asia. wild forms. This led to the selection of plants with tough rachises7 that prevented The origins of domestic crops in the the separate spikelets, which enclose the 5 Southwest Asian grains, fromshedding naturally once ripe. As the European icesheets began to retreat We encounter the earliest evidence of at the end of the most recent ice age soon the domestic crops on which Neolithic after 18,000 years ago, the climate slowly agriculture was founded in the Fertile became wetter and warmer. This gradual Crescent during the period referred to as amelioration was punctuated by much the Pre- Neolithic A or PPNA (c. colder and drier periods, lasting for at 10,200 to c. 9500 bp), soon after the first Figure 1 Part ofthe circular tower atferi­ most a thousand years or so, and the dis­ experiments at cultivation had apparently cho (front left), with a deep excavated sec­ tribution and composition of the natural taken place a few hundred years earlier. tion beyond showing the deposits that vegetation cover altered significantly in There were eight founder crops, three of accumulated around the tower during the response to the changes in temperature which were cereals (barley, einkorn and Pre-Pottery Neolithic A period.

44 ARCHAEOLOGY INTERNATIONAL

(a) (c) Cafer Hoyuk 244

Asiklihoyuk Cayonu .. 512 · �ayonu 833 .

M'lefaat Qermez Dere 18 Halula Tepe jerf alAhmar 80 678 Te ll Abu Hureyra 260 gTell Mureybit � li 2343 Tepe Abdul Hosein•

Ali Kosh 6.....: 476 Parekklisha-Shillourokambos ... Tell Ghoraife 77

. Te ll Aswad 449 449

jilat 7 273

250 250 km wild cereals Beidha km '------' D status indeterminate (b) (d) • domestic cereals Cafer Hoyuk c->,7 A 422 L.-.,) Cayonu Cayonu 1522 . "� • 313 Can Hasan III • 213 'I Halula Djade fli 361 Tell Mureybit Tell Ras 10 Shamra 32 G 4308 Kissonerga- • Mylouthkia 630 �Parekklisha- 180 kklisha Shillourokambos �Shillourokambos- Kissodnerga-Myloulhki.a ..Te ll Ghoraife 150 ..1371 Te ll Aswad Q Tell Ramad 157 56252

Jeric 9012 Azraq 31 8

\'Vadi Finan A 267 0 km 250 ' km 250

Figure 2 Th e distribution of early Neolithic sites in the Fertile Crescent, southeasternAnat olia and Cyprus; the changing proportions of wild and domestic cereals identified at them have been calculated for each period: (a) PPNA , 10,200-9500 bp; (b) Early PPNB, 9500- 9200 bp; (c) Middle PPNB, 9200-8500 bp; (d) Late PPNB, 8500-8000. Th e numbers by the site names refer to the total quantity of cereal remains (seeds and ch aff) identified. Such inform ation is lacking for the sites without numbers. that bore the impressions of grains, sam­ subsequent spread of the founder crops dates of their settlements become pro­ ples of charred plant remains were taken during the PPNA and PPNB may need to be gressively later as the distances from the for analysis from the deposits adjacent to redrawn. Figure 2 shows the results of our Southwest Asian heartland increase. The the walls 10 Domestic einkorn and emmer first attempts at reconstructing the chang­ founder crops (and domestic animals) that wheat and domestic barley (the three ing proportions of wild and domestic cere­ evolved in Southwest Asia formed the founder-crop cereals) were identified in als found at PPNA and PPNB sites around the basis of the so-called Neolithic farming these PPNA contexts.11 However, our closer Fertile Crescent. Wild cereals predomi­ package that disseminated throughout examination of the relationship between nate in a majority of the PPNA (phase 1) continental Europe and across the sea to these finds and the radiocarbon dates of sites, whereas at the few Early PPNB (phase Britain and Ireland. Whether it was the the layers with which they were associ­ 2) sites, the pattern is reversed, with most Neolithic farmers themselves who spread, ated has shown that the cereal remains of them having more domestic cereals. or their crops and the knowledge and tech­ were chronologically contemporary with This trend continues during the Middle nology to grow them, or some combination the succeeding Early Pre-Pottery Neolithic and Late PP B (phases 3 and 4), with of both, is hotly and at times acrimoni­ B phase (c. 9500-9200 bp).12 Therefore, we increasing evidence for higher propor­ ously debated. The timing and tempo of should probably no longer regard the tions of domestic crops at the more numer­ this Neolithic diaspora has been the sub­ domestic cereals found at Jericho as ous sites now spread throughout the ject of much research over the past 30 among the earliest known, although (no Fertile Crescent and extending west into years. less significantly) they may represent evi­ south-central . A major source of evidence that has dence for the initial stages of dispersal been used in the past to investigate the fromtheir areas of origin elsewhere in the Modelling the spread of early process of dispersal across Europe has region. fa rming in Europe been the chronology of the first appear­ This reassessment of the archaeobotan­ Farming communities became established ance of farming communities. Grahame ical evidence from Jericho suggests that across Europe some time after their initial Clark was the first to attempt this, in maps showing the areas of origin and appearance in the Fertile Crescent, and the 1965.13 He established that the initial

45 ARCHAEOLOGY INTERNATIONAL

spread of farming took place along the recent radiocarbon database of such sites spread of agriculture and pastoralism in Danube, thus confirming the earlier work in Europe.20 This, in turn, will be linked Eurasia, D. R. Harris (ed.), 142-58 (Lon­ by Gordon Childe that was undertaken to GIS (geographical information system) don: UCL Press, 1996), and D. Zohary, without the benefit of radiocarbon dating.14 map data of topographic, soil, temperature "Monophyletic versus polyphyletic ori­ gin of the crops on which agriculture was Then, in 1971, Ammerman & Cavalli­ and rainfall patterns across Europe. The founded in the Near East", Genetic Sforza proposed a statistical model for the overall objective is to model the spatial, Resources and Crop Evolution46, 133-42 , expansion of farming.15 They found that, temporal and ecological contexts of the 1999. although a general rate of progression of first appearance and dispersal of plant 9. Kathleen Kenyan undertook seven cam­ about 1 km per year (from anassumed ori­ domesticates, which will contribute greatly paigns at Jericho between 1952 and 1958; gin at Jericho) explained the spread of the to our understanding of the complex proc­ see Excavations at Jericho, vol. m: the Neolithic across Europe reasonably well, esses by which crops, early farming prac­ architecture and stratigraphy of the tell. there were areas that deviated substan­ tices and Neolithic peoples spread into (text), K. Kenyan & T. A. Holland (eds) tially fromthis average, notably the Linear Europe from their ecologicallyvery differ­ (London: British School ofArchaeology in Jerusalem, 1981). Pottery (Bandkeramic) culture of central ent Southwest Asian homeland. 10. See plates 236 and 238 in Excavations at Europe, which evidently spread at rates in Jericho, vol. Ill: the architecture and 5 excess of km a year. Based on this evi­ Notes stratigraphyofthe Tell (plates), K. Kenyan dence, and despite identifying territories 1. He did so as early as 1960; see p. 9 in & T. A. Holland (eds) (London: British into which the Neolithic moved at a rate H. Helbaek, "The palaeoethnobotany of School of Archaeology in Jerusalem, faster or slower than the average, they the Near East and Europe", in Prehistoric 1981). concluded that the best explanation for investigations in Iraqi Kurdistan, R. J. ll.M. Hopf, "Appendix B: Jericho plant Braidwood & B. Howe (eds), 99-1 18 the overall pattern of movement was a remains", in Excavations atJerich o, vol. 5: (Chicago: University of Chicago Press, demic diffusion caused by population the pottery phases of the tell and other Studies in Ancient Oriental Civilization fin ds, K. M. Kenyan & T. A Holland (eds), expansion, which resulted eventually in no. 31, Oriental Institute, University of 576-621 (London: British School of the replacement, by incoming Neolithic Chicago, 1960). Archaeology in Jerusalem, 1983). farmers, of indigenous Mesolithic hunter­ 2. See p. 385 in "Plant collecting, dry farm­ 12. 0. Bar-Yosef & A. Goren suggest that the gatherers. ing, and agriculture in prehis­ radiocarbon dates from Jericho indicate However, later research has muddied toric Deh Luran", in and that the PPNA lasted there from c. 10,300 the waters. The spatial and temporal pat­ ecology of the Deh Luran plain, bp to 9300-9200 bp; see p. 25 7 in An early terns of Neolithic expansion identified by F. Hole, K. V. Flannery, J. A Neely (eds), Neolithic village in the Jordan valley, part 383-426 (Memoir 1, Museum of Anthro­ Ammerman & Cavalli-Sforza were shown 1: the archaeology of Netiv Hagdud, pology, University of Michigan, 1966). to be as easily accounted for by population 0. Bar-Yosef & A. Goren (eds) (Bulletin 3. The project, entitled "The origin and 43, American School of Prehistoric diffusion (e.g. through intermarriages spread ofNeolithic plant economies in the Research, Peabody Museum of Archaeol­ between hunter-gatherer and early farm­ Near East and Europe", is supported by a ogy and Ethnology, Harvard University, ing groups) or by cultural diffusion in­ grant from the UK Arts and Humanities 1997). dependent of population movements.16 Research Board. Stephen Shennan is the 13.J. G. D. Clark, "Radiocarbon dating andthe Genetic analysis of modern Europeans has principal investigator, and James Conolly expansion of farming culture from the shown the importance of pre-Neolithic and James Steele (of Southampton Uni­ Near East over Europe", Proceedings of populations in the contemporary gene versity) are eo-investigators. The project is the Prehistoric Society 31, 57-73, 1965. one of the research initiatives at UCL pool,17 and more sophisticated analysis of 14. V. G. Childe, The Danube in prehistory taken by the Centre for the Evolutionary the radiocarbon record has shown that the (Oxford: Oxford University Press, 1929). Analysis of Cultural Behaviour, directed 15.A. J. Ammerman & L. L. Cavalli-Sforza, relationship between the decline of indig­ by Stephen Shennan. "Measuring the rate of spread of early enous Mesolithic populations and the first 4. The term "" refers to the farming in Europe", Man (new series) 6, appearance of farming communities is var­ final part of the Palaeolithic period in 674-88, 1971. ied and at times ambiguous.18 Now the Southwest Asia that immediately pre­ 16.K. Aoki, M. Shida, N. Shigesada, "Travel­ consensus of opinion is that some mixture ceded the Neolithic. In Europe the term ling wave solutions for the spread offarm­ of demic expansion, diffusion, and the "Mesolithic" is used for the equivalent ers into a region occupied by hunter­ adoption of cultural traits was responsible pre-Neolithic period. gatherers", Theoretical Population Biol­ 5. The Fertile Crescent refers to the upland 7 for the spread of the Neolithic into and ogy 50, 1-1 , 1996. arc that stretches from the southern A. across Europe, but debate still continues 17. See, for example, M. B. Richards, V. north and east around the Meso­ Macaulay, H. J. Bandelt, B. C. Sykes, as to which of these processes were effec­ potamian lowland and southeast along "Phylogeography ofmitochondrial DNA in tive in different regions.19 the foothills of the Zagros mountains in western Europe", Annals of Human Evidence used in debates on the neoli­ western . Genetics 67, 1251-76, 1998. thization of Europe is still mainly based on 6. The people known as Natufians occupied 18. M. Gkiasta, T. Russell, S. Shennan, the analysis ofradiocarbon dates fromlate sites throughout the Levant, from the mid­ J. Steele, Origins of European agriculture: Mesolithic and early Neolithic sites. It is dle valley in the north to the the radiocarbon record revisited, unpub­ surprising that the most direct evidence highlands of the Negev in the south, and lished manuscript, UCL Institute of east across the Jordanian plateau. for the spread of farming communities - Archaeology, 2000. 7. The rachis is the "spine" ofthe cereal ear. the remains of their crops - has not been 19. See for example P. Bellwood, 2001. "Early In wild cereals the rachis is brittle and the agriculturalist population diasporas? incorporated systematically into any of separate spikelets containing the seeds Farming, language, and genes", Annual the existing models. Yet archaeobotanical shed naturally once the plants are ripe. In Review of Anthropology 30, 191-207, data, when classified spatially and tempo­ the process of selection 2001, and M. Nowak, "The second phase rally, can give us the clearest evidence not favoured plants with tough rachises, to ofNeolithization in east-central Europe", only of when crops first appeared in any which the spikelets remained firmly Antiquity 75, 582-92, 2001. given region but also about the ecological attached even when the ears were ripe. 20. See Gkiasta et al., cited in n. 19 above. characteristics of early farming practices. Thus, the domesticated cereals could not survive without the intervention of the It is for these reasons that our first priority early Neolithic farmers, who had to reap is to establish an archaeobotanical data­ and sow the grains to guarantee annual base containing details of the wild and harvests. domestic plants found on pre- and early 8. See D. Zohary, "The mode of domestica­ Neolithic sites in both Southwest Asia and tion of the founder crops of Southwest Europe, which will be linked to the most Asian agriculture" in Th e origins and

46