Harvesting cereals and other plants in Neolithic Iberia: the assemblage from the lake settlement at La Draga.

Palomo, Antoni ; Gibaja, Juan F. ; Pique, Raquel 等

[ILLUSTRATION OMITTED]

Introduction

The first evidence of the Neolithic in north-eastern Iberia appears about the middle of the sixth millennium cal BC. Until that time, the area had been occupied by hunter-gatherer communities, living both on the plains and in mountain areas of the pre-Pyrenees and Pyrenees and the pre-littoral sierras. These sites were occupied later by groups whose subsistence was based on the joint exploitation of wild and domestic resources, obtained by hunting, gathering, agriculture and animal husbandry. However, the first Neolithic communities also sought out places with greater farming potential, located on the plains and in the valleys. This is the case of the site presented here. the Neolithic settlement of La Draga at Banyoles, Girona, Spain (Tarrus et al. 1994; Bosch et al. 1999, 2000, 2006b). The exceptional preservation encountered here has permitted the recovery of implements of wood as well as stone, some of them certainly used for harvesting.

The site

The settlement at La Draga is located in the north-east of Iberia, on the eastern shore of L'Estany de Banyoles, a small lake 50km from the Mediterranean coast and 40km south of the Pyrenees. It was discovered in 1990, and since then it has been excavated under the direction of the Museu Arqueologic Comarcal de Banyoles (MACB) and the Centre d'Arqueologia Subaquatica de Catalunya (CASC). Recently, the research team has been enlarged with the participation of the Consejo Superior de Investigaciones Cientificas (CSIC), the Universitat Autonoma de Barcelona (UAB) and the Museu Arqueologic de Catalunya (MAC).

The lake is a karst landform and is fed by underground waters. Originally it was drained on its eastern side by a small river, the modern Terri, a tributary of the Ter. This river, as it exited the lake, must have created an area of marshes on the northern shore, which is the location of the La Draga archaeological site. During the Neolithic occupation this shore would have taken the form of a peninsula, which stretched out into the lake with a gentle continuous slope from east to west and from north to south. A survey suggests that the settlement occupied an area of about 8000[m.sup.2].

The location of the site confortos to common practice in the western Mediterranean region: Neolithic settlements are found in wetland locations, on the shores of lakes, lagoons or marshes yet close to agricultural land. This pattern has been documented in Italy at La Marmota (Fugazzola et al. 1993), southern France at Leucate (Guilaine et al. 1984), by lakes in the Alps and the Jura, and inland in the Iberian Peninsula (Rojo et al. 2008).

The fact that the site is now partially covered by the waters of the lake has favoured the extraordinary state of conservation of organic remains made from plant matter (Figure 1). These range from the wooden posts in the buildings to the smallest objects made or gathered by the occupants. This makes La Draga a privileged place to carry out subsistence or technological studies in order to understand these first Neolithic populations in the western Mediterranean in greater depth.

The archaeological excavations carried out to date have documented a habitation level, situated immediately above the lacustrine chalk and largely underwater, which corresponds to an occupation by a Neolithic population linked to the Cardial culture (Figure 2). Numerous radiocarbon determinations have been obtained from samples of wood, bone and charcoal whose results place the oldest occupation between 5300 and 5150 cal BC. Equally, based on the dendrochronological analysis of wooden posts, we may consider that the occupation perhaps went through different phases, during a period of approximately 80 or 100 years.

[FIGURE 1 OMITTED]

The archaeological excavation has recovered hundreds of posts that supported large huts, sunk in the ground to depths of over 2m. Next to these, small posts or stakes correspond to complementary structures. The archaeological layer has accumulated all the wood from the collapse of the huts, including several boards, fragments of topes made from plant matter and rolls of lianas used to hold together the different building elements such as posts, beams and boards. Additionally, fragments of oak bark have been found and these may have been part of the roofs or the floors of the huts.

One of the huts must have been destroyed by fire, leaving inside it a good sample of the implements used by the occupants, including some that are partially carbonised. There were many wooden utensils, of types that are repeated in different modules and sizes (Figure 2). of these, the most important are agricultural tools (digging sticks and sickles), building tools (adzes, wedges), hunting instruments (bows, arrows, spears) and domestic utensils (wooden bowls, baskets made from aquatic plants, mixers, combs, spindles, spoons and spatulas). In addition, a group of bull's bucrania has been found which can be associated with a series of wooden hooks in the form of a yoke, and which could be the hangers at the ends of the beams. However, we do not rule out the possibility that these bucrania had a ceremonial use and meaning.

We believe that the first large huts stood on the lake shore elevated a little above the original ground level of lacustrine chalk. This building method would have avoided flooding if the level of the lake rose. It is not surprising, therefore, that no hearths have been found in situ. However, the recovery of accumulations of charcoal and cobblestones affected by heat probably derive from hearths cleared out from inside the huts.

[FIGURE 2 OMITTED]

By contrast, the part of the site situated further inland from the lake shore was slightly later in date--at the beginning of the fifth millennium cal BC--and had a lower degree of preservation. In this drier area, only the bottom part of the wooden posts was preserved and hearths were found in hollows filled with charcoal, quartz and sandstone cobbles. Some structures were built directly on the ground, with floors made from travertine flagstones, aimed at insulating the hut floor from the damp.

Harvesting tools

Although many different wooden implements have been discovered at La Draga--tips of spears, tips and shafts of arrows, containers and hafts--in this paper we focus on the harvesting tools found in the earliest phase of the site, which are represented by one wooden blade and seven sickles (Figure 4). The wooden blade (Figure 4.3) is made from oak (Quercus subg. quercus) and consists of a cylindrical haft, finished with a spherical knob at the proximal end and an active part at the opposite end. The active part is a rectangular appendage with a concave cutting edge. This carried several marks that would suggest it was used to pull up non-woody plant fibre, such as cereals or aquatic plants. At La Draga, not only have numerous seeds from different species of domestic and wild plants been found (see below), but also some basketwork containers made of vegetable fibre from aquatic plants.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Five of the sickles were fractured, but two of them were intact, one still with an embedded flint blade (Figure 4). One was made from Sambucus sp., one of Juniperus sp. and the other five of Buxus sempervirens (Bosch et al. 2006a). All these types of wood are hard and resilient. The tools have the same general form: a cylindrical shaft (the haft) attached to a thinner branch (the hook). The slots for the flint blades were in the shaft--not the hook--and inserted along an axis orthogonal to the hook, being set either parallel to the haft or diagonal to it. In six cases there was only one slot for a blade, whereas in one case there were two slots on the same shaft (Figure 4.4). The dimensions are variable (Table 1), the one made from Juniperus being the largest but it is fragmented and its full length cannot be appreciated. In all cases advantage was taken of the morphology of a natural branch to form the hook.

The form and function of the sickles can be best appreciated from that made of elder (Sambucus sp.) which retained its flint blade in place (Figure 4.5). It is a tool with a cylindrical haft terminating in a cylindral knob at the proximal end, with a branch forming a right-angle at the distal end. The flint blade was fixed in a groove on the axis of the haft and lies obliquely in relation to the haft. The flint blade was affixed with pine resin: Pinus silvestris, according to the phytolith study carried out by researcher Dr Jordi Juan. Use-wear analysis of the blade has confirmed that this is a sickle. We observed that there was very shiny micro-polish spreading substantially towards the inside area with a compact weft and smooth appearance. In the inner part of the micro-polish area we noted some deep narrow striations both parallel and diagonal to the edge. All along the edge, small nicks appeared intermittently along the active area.

Lithic tools: raw materials, technology and use-wear analysis

The raw material used for stone tools at La Draga is mainly quartz, hyaline quartz and different varieties of flint, with flint predominating. Other kinds of stone, such as quartzite, lydite and jasper were also used, bur these were tare. The majority of the retouched pieces are made of flint (93 per cent). Most numerous are the retouched blades, denticulates and notches. Geometrics and borers appear in smaller numbers. Quartz and hyaline quartz from areas surrounding the site were worked at the settlement itself. But for the flint, the abundance of blades and the absence of certain products such as cores and uncut blocks, show that the shaping of cores took place outside the settlement. Nor can we discount the possibility that flint blades came to La Draga as finished products.

The use-wear analysis of the lithic assemblage from La Draga has detected a large number of pieces used for working with non-woody vegetable matter: 42 objects representing 24.6 per cent of used material. Most of these pieces are blades (72 per cent) of which 18 (60 per cent) have been used on both their edges. The rest of the tools are flakes (28 per cent) where, in general, only one edge had been used. The characteristics of the traces on several of these pieces, the micro-polish in particular, are probably connected with cereal reaping (Figure 5). But in others, the presence of micro-polish that is not very extensive, with a very compact convex weft, showing little striations and few micro-holes, leads us to the hypothesis that they were used for cutting unripe cereals for limited rime periods or for cutting other kinds of wild plants, such as the reeds used to make containers found at the site. Analyses of the lake sediment show that this kind of plant grew next to the settlement (Burjachs 2000; Buxo et al. 2000).

On various blades hafted parallel to the handle we noted the presence of a very extensive micro-polish, its weft half-closed and compact, with little shine and many striations, produced by a totally rounded edge (Figure 6). Our experiments have revealed that these kinds of traces might have been produced as a consequence of continuous contact with the ground (Clemente & Gibaja 1998).

Flora and fauna

Due to the good preservation of organic material, it has been possible to analyse thousands of samples of plant remains that together with faunal remains provide exceptional information about the subsistence of the first farming communities. The population of La Draga primarily practised agriculture and animal-husbandry, whilst hunting and gathering of wild plants was a secondary activity. Cereals like wheat and barley (Hordeum vulgare L., Hordeum vulgare L. var. nudum, Triticum aestivum/durum, Triticum dicoccum Sch.) would have made up the main agricultural base, always accompanied by pulses such as peas (Pisum sp.) and broad beans (Vicia faba) (Buxo et al. 2000). There is a far smaller percentage of fruit remains such as wild grapes (Vitis vinifera), blackberries (Punus spinosa), pine kernels (Pinus pinea) and hazelnuts (Corylus avellana). These plant remains are scattered throughout the settlement, but are concentrated principally around hearths and granaries. In this context, we must mention the discovery of a cylindrical storage vessel in association with the granaries, found filled to the brim with wheat seeds (Tritticum durum/aestivum). The domestic animals raised would have been, mainly, herds of bovidae (Bos taurus), flocks of ovidae-capridae (Ovis aries and Capra hircus) and, to a lesser extent, suidae (Sus domesticus). It is surprising to note the large numbers of remains of domestic cattle in comparison to those of sheep and goats, which generally form the bulk of livestock found at cave-based Mediterranean sites from the Neolithic. If we examine the overall nutritional base, cattle stand out as the most significant species among the livestock raised at La Draga (Sana 2000). They also make their appearance as a wild animal (aurochs), along with deer, wild goats, boar and roe deer, amongst others, all of which were hunted (Palomo et al. 2005).

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Discussion

As can be seen, the flint blades were not inserted into the hook of the sickle, but straight into the cylindrical part of the instrument which also formed the handle. The cylindrical part of the tool thus acted as both the haft for the blade (upper end) and the grip for a hand, resting against the knob (lower end). The bent branch or hook at the distal end lies at right-angles to the axis of the haft and to the axis of the flint blade. In our interpretation, this wooden hook would have been used to gather the stalks, which were then cut with the blade. The wrist movement (45o) required to execute this double action (collecting and cutting) meant that, when the stalks were cut, the bent branch had to be pointing at the ground. This would have made it awkward for the cutting to take place on the ground itself, but easier at a minimum distance of 120mm above it. Experiments with replicas of the La Draga sickles show their efficiency (Figure 7) but confirmed the difficulty of cutting stalks at ground level, although some use-wear results suggest that frequent contact was made with the ground in some cases.

[FIGURE 7 OMITTED]

These sickles have much in common with those discovered in the middle stages, the older Cortaillod and Pfyn, of the Neolithic in the Alpine region. These were classified as sickles with oblique direct hafting, such as those found at Egolzwil VI in Lucerne, Switzerland (Mueller-Beck 1965; Voruz 1991; Schlichtherle 1992). Through use-wear analysis, both kinds of sickles have been recorded in other parts of the Iberian Peninsula. The straight sickles are common in many Neolithic contexts dated to the sixth and fifth millennia cal BC in north east Iberia and south-west France. We have fewer references to sickles consisting of a single blade hafted diagonally but they have been recorded at sites dated to the mid sixth millennium cal BC in central Spain, such as Casa Montero, Madrid, and La Revilla and La Lampara, both in the province of Soria (Gibaja 2002, 2008; Ibanez et al. 2008).

Conclusions

The inhabitants of La Draga made use of reaping implements of different forms: a hefty wooden blade and two types of sickle with the blade parallel or oblique to the haft. We do not have a specific explanation for these different ways of hafting, but suggest it is connected with the kind of plant to be cut or with the process for which the tools are intended, such as weeding or harvesting. Several plants have been documented at La Draga that would require different reaping techniques to be harvested, e.g. the different kinds of cereals collected for use as grain or straw and aquatic plants such as Carex sp. or reeds that were cut to make baskets.

Although we have documented a possible implement made entirely out of wood, used to pull up and cut plants, it should be stressed that most of the sickles consist of a wooden haft and one or several pieces of flint. Just as hard woods were always chosen for the hafts (Buxus, Juniperus, Sambucus), flint, in the form of both blades and flakes, was always used for the lithic elements.

At their distal end, these sickles had a branch that acted as a hook or gatherer. Experiment has verified the function of this appendage and its effectiveness in reaping cereals. It enables several stalks to be gathered together so that they can be easily cut with a small wrist movement by the flint pieces that are fitted into the side of the sickle. The observation of the distribution of microscopic cereal traces on the lithic elements shows that they were fitted obliquely or in parallel with the haft. Although in the only sickle that has been preserved with a lithic element still in place this is positioned obliquely, other sickles exhibit a groove for flint blades that is aligned with the haft. In these cases we cannot be sure how many lithic pieces would have been used nor their orientation.

The study of use-wear traces has also enabled us to determine that whereas some of the flint artefacts were used to cut cereals near or over the ground--as a way of separating the stalks from the roots and ears or the stalks themselves with certain lengths--others were used to cut the stalk at a greater height in order to use the ear and a part of the stalk in particular.

Acknowledgements

Funding for excavation, conservation and study was provided by the Departament de Cultura de la Generalitat de Catalunya, Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya and Ajuntament de Banyoles. The research was also supported by the Ministerio de Ciencia e Innovacion (HAR2009-13494-C02-01 and HAR2009-13494-C02-02. Las ocupaciones lacustres y la gestion de los recursos entre las primeras sociedades agricolas y ganaderas del NE peninsular. Subproyecto Tecnologia de las producciones materiales y usos instrumentales. Subproyecto Estrategias agroforestales y ganaderas), the European Research Council (ERC-AdG 230561. Origins and spread of agriculture in the south-western Mediterranean region), Fundacao para a Ciencia e a Tecnologia (PTDC/HAH/64548/2006. The last hunter-gatherers and the first farming communities in the south of the Iberian Peninsula and north of Morocco: a socio-economic approach through the management of production instruments and exploitation of the domestic resources) and Agencia de Gesti6 d'Ajuts Universitaris i Recerca (2009 SGR 734. AGREST: Arqueologia de la Gestio deis Recursos Sociais i el Territori). Our sincere appreciation goes to all those who supported and participated in the excavations at La Draga. Text corrections and adaptation are by David Passingham and Peter Smith.

References

BOSCH, A., J. CHINCHILLA & J. TARRUS. 1999. La Draga: un poblado del neolitico antiguo en el lago de Banyoles, Girona, Catalunya, in J. Bernabeu Auban & T. Orozco Kohler (ed.) Actes de 2 Congres del Neolitic a la Peninsula Iberica, Universitat de Valencia, 7-9 d'abril, 1999 (Saguntum, Extra 2): 315-21. Valencia: Departament de Prehistoria i d'Aqueologia, Universitat de Valencia.

--2000. Elpoblat lacustre neolitic de la Draga: excavacions de 1990 a 1998 (Monografies del CASC 2). Girona: Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya.

--2006a. Els objectes de fusta del poblat neolitic de la Draga: excavacions 1995-2005 (Monografies del CASC 6). Girona: Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya.

BOSCH, A., J. CHINCHILLA, J. TARRUS & R. PIQUE. 2006b. Els objectes de fusta i fibres vegetals, in A. Bosch, J. Chinchilla & J. Tarrus (ed.) Els objectes de fusta del poblat neolitic de la Draga: excavacions 1995-2005 (Monografies del CASC 6): 27-126. Girona: Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya.

BURJACHS, F. 2000. El paisaje del neolitic antic: les dades palinologiques, in A. Bosch, J. Chinchilla & J. Tarrus (ed.) El poblat lacustre neolitic de la Draga: excavacions de 1990 a 1998 (Monografies del CASC 2): 46-50. Girona: Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya.

BUXO, R., N. ROVIRA & C. SAUCH. 2000. Les restes vegetals de llavors i fruits, in A. Bosch, J. Chinchilla & J. Tarrus (ed.) Elpoblat lacustre neolitic de la Draga: excavacions de 1990 a 1998 (Monografies del CASC 2): 129-39. Girona: Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya.

CLEMENTE, I. & J.E GIBAJA. 1998. Working processes on cereals: an approach through microwear analysis. Journal of Archaeological Science 25(5): 457-64.

FUGAZZOLA, M.A., G. D'EUGENIO & A. PESSINA. 1993. La Marmotta. Scavi 1989. Un abitato perilacustre di eta neolitica. Bulletino di Paletnologia Italiana 84: 181-342.

GIBAJA, J.F. 2002. Las hoces neoliticas del noreste de la Peninsula Iberica. Prehistoire Anthropologie Mediterraneennes 10: 83-96.

--2008. La funcion del utillaje litico documentado en los yacimientos neoliticos de Revilla del Campo y La Lampara, Ambrona, Soria, in M.A. Rojo, M. Kunst, R. Garrido, L Garcia & G. Moran (ed.) Paisaje de la memoria: asentamientos del neolitico antiguo en el Valle de Ambrona, Soria, Espana (Arte y Arqueologia 23): 451-93. Valladolid: Universidad de Valladolid.

GUILAINE, J., A. FREISES & R. MONTJARDIN. 1984. Leucate-Correge: habitat noye du Neolithique cardial. Toulouse: Centre d'Anthropologie des Societes Rurales.

IBANEZ, J.J., J.E. GONZALEZ, J.F. GIBAJA, A. RODRIGUEZ, B. MARQUEZ, B. GASSIN & I. CLEMENTE. 2008. Harvesting in the Neolithic: characteristics and spread of early agriculture in the Iberian peninsula, in L. Longo and N. Skakun (ed.) Prehistoric technology 40 years later: functional studies and the Russian legacy (British Archaeological Reports international series 1783): 183-95. Oxford: Archaeopress.

MUELLER-BECK, H. 1965. Seeberg Burgaschisee-Sud: 5, Holzgerate und Holzbearbeitung (Acta Bernensia 21). Berne: Stampfli.

PALOMO, A., R. PIQUE, M. SANA, A. BOSCH, J. TARRUS, J. CHINCHILLA & J.F. GIBAJA. 2005. La caza en el yacimiento lacustre de La Draga, Banyoles-Girona, in P.A. Cabal, R. Ontanon & C. Garcia (ed.) Actas del 3 Congreso del Neolitico en la Peninsula Iberica, Santander, 5 a 8 de octubre de 2003 (Monografias del Instituto Internacional de Investigaciones Prehistoricas de Cantabria 1): 135-44. Santander: Servicio de Publicaciones, Universidad de Cautabria.

ROJO, M.A., M. KUNST, R. GARRIDO, I. GARCIA & G. MORAN. 2008. Paisaje de la memoria: asentamientos del neolitico antiguo en el Valle de Ambrona, Soria, Espaca (Arre y Arqueologia 23). Valladolid: Universidad de Valladolid.

SANA, M. 2000. La gestio i esplotacio dels recursos animals, in A. Bosch, J, Chinchilla & J. Tarrus (ed.) El poblat lacustre neolitic de La Draga: excavacions de 1990 a 1998 (Monografies del CASC 2): 150-64. Girona: Centre d'Arqueologia Subaquatica de Catalunya, Museu d'Arqueologia de Catalunya.

SCHLICHTHERLE, H. 1992. Jungsteinzeitliche Erntegerate am Bodensee. Plattform 1: 24-44.

TARROS, J., J. CHINCHILLA & A. BOSCH. 1994. La Draga, Banyoles: un site lacustre du Neolithique ancien cardial en Catalogne. Bulletin de la Societe Prehistorique Francaise 91(6): 449-56.

VORUZ, J.L. 1991. Le Neolithique suisse: bilan documentaire (Document du Departement d'Anthropologie et d'Ecologie de l'Universite de Geneve 16). Geneve: Universite de Geneve.

Antoni Palomo (1), Juan F. Gibaja (2), Raquel Pique (3) *, Angel Bosch (4), Julia Chinchilla (5) & Josep Tarrus (4)

(1) Departament de Prehistoria, Universitat Autonoma de Barcelona, 08193 Bellaterra, Spain (Email: tpalomo@arqueolitic.com)

(2) Ramon y Cajal Research Program, Departament d'Arqueologia i Antropologia CSIC-IMF, Carrer Egipciaques 15, 08001 Barcelona, Spain (Email: jfgibaja@imf.csic.es)

(3) Departament de Prehistoria, Universitat Autonoma de Barcelona, 08193 Bellaterra, Spain (Email: raquel.pique@uab.cat)

(4) Museu Arqueologic Comarcal de Banyoles, Placeta de la Font, 11-17820 Banyoles, Catalonia, Spain (Email: abosch15@xtec.cat; macb@ajbanyoles.org)

(5) Escola de Restauracio i Conservacio de Bens Culturals de Catalunya, c. Aiguablava, 109-113 Barcelona, Spain (Email: juliachinchi@hotmail.com)

* Author for correspondence

Table 1. Dimensions of sickles (in mm).

 Haft

Ref. Taxa Length Thickness Width

FG 91 Sambucus 180 22.5
1

JE 83 Juniperus 140.7 12 14.6
31

JI 87 Buxus 194 18 24
13

KB 89 Buxus 210 15 45
6

KA 88 Buxus 204 11 40
12

JG 90 Buxus 200 20 35
23

KID 92 Buxus 200 15 42
5

 Appendix

Ref. Length Thickness Width Slot position Handle

FG 91 83.4 13.3 91.6 No
1

JE 83 138.4 8.4 15
31

JI 87 116.8 13 21.5 95 93
13

KB 89 601135 No
6 (2 slots)

KA 88 125 95
12

JG 90 83 78 (broken)
23

KID 92 92 13 28 100 90
5