Plant remains can be recovered at archaeological sites, even in hunter-gatherer sites as long as the adequate sampling strategies are used. In contexts earlier than 5000 BP(1) we have identified different types of wild plant foods: nuts (hazelnut and acorns), pomes and fragments of parenchymatous tissues (possible roots or tubers). Hazelnut pericarp is the most abundant plant macro-remain in Holocene sites from northern Iberia, most probably because Corylus was very frequent during this period, but also because the pericarps got frequently in contact with fire and may have had a differential preservation. Acorns are less frequent but they could have been an important resource because oaks were the most important component of forest communities during this period and the ethnographic evidence shows that they have been widely used in human subsistence. We have identified Rosaceae fruits type Malus (crab apple) or Sorbus in Aizpea, Kanpanoste Goikoa and Lumentxa. The pomes could have got charred during the roasting or drying in order to improve their taste or conservation.

Even in the sites where wild plant foods are well preserved, as in the rock-shelter of Aizpea (Navarre, Western Pyrenees), we cannot forget that, when the only type of conservation is carbonisation, there must have been many other plant foods that were used but have not been preserved. The number of plants used for food, medicine, rituals, etc. must have been much higher than what we recover in our samples where only acorns, hazelnuts and Pomoideae fruits are present. Basically from the data from Aizpea, we think that among pre-farming societies there could have been an intensive exploitation of wild plant foods: (a) wild vegetal resources were abundant, easy to gather and could be stored, (b) we recover some remains, and (c) the only available chemical analysis on human bones shows that plants were the most important type of food of this individual. However, the data is still very limited to be able to adequately assess their real importance in the human diet.

At this moment, the oldest evidence of agriculture in the Basque Country comes from caves on the coast: Kobaederra (Kortezubi, Biscay) where cereal is AMS(2) dated at 5375±90 BP (4360-3990 cal. BC) and Lumentxa (Lekeitio, Biscay) where barley has been recovered in a context dated at 5180±70 (4220-3800 cal. BC) and 5095±75 (4040-3710 cal. BC). Because of lack of sampling, at present we do not have data for the Basque southern area.

The archaeozoological remains recovered in Peña Larga (Cripán, Álava) and Arenaza (Galdames, Biscay) show that domestic animals were already available in the Basque Country from at least c. 6000 BP. The dates for the coast and the inland do not differ greatly. These data and the chronology of the first domesticates in the neighbouring territories lead us to think that the Basque Country may have adopted agriculture from the period 6000-5500 BP, although this is something that future research will have to prove.

Apart from the Ebro valley, usually considered the only possible way for domesticates to arrive in northern Iberia, we cannot ignore the importance that the south of France could have had in the Neolithisation of the Western Pyrenees and the Atlantic Façade. There is artefactual evidence showing the existence of trans-Pyrenean relationships and recent data reject the idea of a late adoption of the Neolithic of western and south-western France. Domesticates could come from different foci although, in the case of plants, the information we have is still extremely scarce. It is also possible that there were no major chronological gaps in the adoption of domesticates in the different areas of the Basque Country. The data from the caves of Kobaederra and Arenaza confirm that domesticates –at least domestic animals– existed on the Basque coast from the first half of the VIth millennium BP. However, we do not know if this was the general situation or if there still existed exclusively hunter-gatherer groups. Sites like Pico Ramos (Nivel 4) in Biscay and Herriko Barra in Gipuzkoa could be hunter-gatherer sites or occasional camps of groups which were already Neolithic.

The neolithic evidence of agriculture is still extremely small in the Basque Country. For this reason we cannot value the importance of the different crops that were involved and we can only say what taxa have been identified so far during the VIth millennium BP: barley and emmer wheat in Kobaederra and hulled barley in Lumentxa. One objective for future research should be to gain more specific and concrete information on this subject and determine whether the agriculture of the Atlantic fringe of the Iberian Peninsula is in anyway different because of its humid climatic conditions.

There is different evidence –the presence of hulled cereals, humid and mountain conditions, low intensity of agricultural practices, absence of sickle lithic artefacts– that lead us to think that during the Neolithic in the Basque Country other methods rather than the use of sickles were used in the harvest of cereals: plucking the ears off by hand or using reaping sticks like the Asturian mesorias. The absence of sickles might mean that the straw was not cut, maybe because it was not necessary for animal food, building material or craftworks. Domestic animals could eat the straw on the fields –although the one from hulled wheats is not greatly appreciated– or use other types of pastures or tree leaf foddering.

From the available data, it seems likely that agriculture was not very important in subsistence systems from the VIth millennium BP. However, we cannot forget that the data is still very limited and that all of it comes from caves.

Some methodological hints

At the current stage of the research in the area –our data about prehistoric agriculture and plant foods in human diets is still extremely limited– it is necessary that the sampling of plant macro-remains be systematically integrated in the archaeological excavations from the beginning of the field work. Flotation machines can be easily incorporated into field work. However, if it is not possible, other methods like manual flotation or water sieving are also valid as long as the adequate mesh is used (0.25 mm).

In hunter-gatherer sites or in sites with Mesolithic-Neolithic transition, an effort should be made in order to process all the sediment which is excavated or at least an important part of it. This is because (1) they are periods about which very little is known about the use of wild plants; (2) it is extremely important to certify the existence or absence of agriculture, and (3) the frequency of the remains tends to be very low. With flotation, if both the flot and the residue are examined, the results improve considerably.