by Chris Fowler
A similar diversity of cereals (barleys, emmer, einkorn, free-threshing wheat), pulses (pea, lentil, bitter vetch, grass pea), and fruits/nuts (hazelnut, apple, acorn, blackberry, hawthorn, plum, grape) characterizes sites in northern Italy, such as mid-sixth to mid-fifth millennium BC Sammardenchia (Rottoli and Pessina 2007). A burned house at Lugo di Romagna provides a snapshot of household-level plant use, including emmer (in store), barley, free-threshing wheat and a little einkorn, peas, lentils, acorns, and hazelnuts (Rottoli and Pessina 2007). The role of wild plant foods is underlined by frequent concentrations of acorn shell (Rottoli and Castiglioni 2008).
In southern France, the scarcity of archaeobotanical data from EN (Cardial) caves and rock shelters has been linked with slow uptake of farming relative to pottery and domestic animals but probably reflects site type and location (Mills 1984). Though sparse, data suggest use of a range of cereals, including einkorn, emmer, free-threshing wheat, and naked barley (Hopf 1991). Limited archaeobotanical investigation of MN (Chasséen) open-air sites suggests use of bitter vetch and broad bean alongside cereals (Hopf 1991; Marinval 1991). Intensive sampling and flotation at a high-altitude MN (mid-fifth to mid-fourth millennium BC) open-air site, Le Chenet des Pierres in the French Alps, has revealed abundant evidence for cereals (especially naked wheat and barley), pea, opium poppy, arable weeds, and collected fruits and nuts (Martin et al. 2008). Together with regional ethnohistorical and palynological evidence, the assemblage suggests high-altitude farming as a possible alternative to widely assumed seasonal transhumance between plains and mountains (see ‘Animal exploitation in south-west Europe’).
Open-air sites in Iberia have yielded crop processing residues (chaff as well as grain) lacking at cave sites (Zapata et al. 2004; Stika 2005). The partially submerged later sixth millennium BC open site of La Draga in Catalonia exemplifies the contrasting preservational biases of charring (evidence for cereals and pulses) versus waterlogging (evidence for gathered wild plants) (Buxó et al. 2000; Buxó, pers. comm.). Iberian cave sites co-existed with open settlements and probably served specialized purposes such as animal shelters. Archaeobotanical datasets from EN caves in Catalonia range from a lack of crop remains at Bauma Serrat del Pont to multiple cereal and pulse crops at Cova 120 (Buxó 2007; Zapata et al. 2004). Further down the Mediterranean coast, caves provide most of the available evidence, encompassing cereals, pulses, and wild plants, especially acorns (Buxó 2007; Zapata et al. 2004). In Andalusia, Cueva del Toro yielded a wide spectrum of cereals (emmer, free-threshing wheat, naked barley) and pulses (pea, lentil, broad bean, bitter vetch, grass pea), and Cueva de los Murciélagos mid-sixth millennium BC opium poppy alongside free-threshing wheat, emmer, and naked barley (Peña-Chocarro 2007). In north-central Iberia, La Vaquera cave yielded a range of cereals and two pulses (lentil, vetch) as well as acorns and grapes; a lack of chaff contrasts with finds of chaff and grain at open-air La Lámpara and La Revilla del Campo in the northern Meseta (Peña-Chocarro 2007; Stika 2005). In mostly fifth millennium BC assemblages from caves in north-west Spain, Pico Ramos seems specialized in wild resources, whilst others (e.g. El Mirón) yielded sparse remains of cereals alongside domestic fauna (Zapata et al. 2004; Zapata 2007).
Wood charcoal from caves and rock shelters in north-west Italy and southern France (Thiébault 2001, 2005; Vernet 2005) generally suggests weak EN human impact on woodland in karstic hill and mid-altitude mountain zones. Increased evergreen oak and garigue scrub plants from the MN onwards may reflect use of deciduous oak, ash, and other species for leaf fodder. The scale of EN–MN upland herding probably varied across the western Mediterranean but was apparently modest compared with recent practice (see ‘Animal exploitation in south-west Europe’) (Thiébault 2001; Delhon et al. 2009).
Caves used as animal pens (grottes bergéries) have been identified from dung deposits and shed deciduous teeth (see ‘Animal exploitation in south-west Europe’). Archaeobotanical analysis of burnt dung layers at early fifth to mid-third millennium BC La Grande Rivoire suggests leaf and twig foddering using oak, ash, lime, and hazel, echoing evidence from Alpine Foreland lake villages (Delhon et al. 2008) and supporting previous charcoal- and pollen-based inferences at cave sites across the French Alps (e.g. Thiébault 2005). Given the opportunistic nature of gathering twig fodder during winter/early spring, and the high labour costs of gathering and storing leafy fodder, evidence of both practices at upland and lowland/lakeshore sites suggests herding on a small scale. The same may be argued for south-east Spanish caves (Badal 2002, 143).
South-west European weed assemblages are generally too sparse for in-depth ecological analysis, but the taxa attested suggest established plots rather than shifting cultivation (e.g. Rottoli and Pessina 2007; Stika 2005; cf. Bogaard 2002). Relatively abundant evidence from LN Clairvaux Station III, Chalain in the French Jura suggests cereal husbandry akin to intensive ‘gardening’ (Lundström-Baudais 1986).
Animal exploitation in south-west Europe
Although faunal analysis has focused on caves and rock shelters, fairly consistent evidence is emerging from lowland open settlements. In Dalmatia, lowland EN–MN (sixth millennium BC) villages at Pokrovnik, Danilo, Nin, and Tinj-Podlivade resemble those in Greece and the eastern Balkans: very sparse evidence for hunting (mainly small game at Pokrovnik and Danilo) and heavy dominance of sheep, which exhibit ‘meat’ mortality (Legge and Moore 2011; Miracle 2006; Mlekuz 2005). In southern Italy, early Neolithic (sixth to fifth millennium BC) ditched enclosures as at Passo di Corvo, Rendina, and Santa Tecchia (see Skeates, this volume ch. 41) likewise have little evidence for hunting and sheep/goats account for 50–65% of domesticates, with sheep/goats and cattle at Torre Sabea again matching ‘meat’ mortality (Vigne 2003). The nature of animal exploitation at many smaller open settlements is largely unknown. In lowland central Italy, domesticates predominate in small samples from EN (sixth millennium BC) open settlements at Villaggio Leopardi and La Marmotta (Cassoli and Tagliacozzo 1995). On the coast of southern France, the EN open settlement at Portiragnes displays little hunting and specialization in sheep that exhibit ‘meat’ mortality (Tresset and Vigne 2007; Vigne and Helmer 2007, 25, fig. 6). In Spain, there is growing evidence for EN open settlements, which include substantial ditched enclosures as at Mas d’Is (Bernabeu et al. 2003), but faunal evidence is still sparse. Domesticates make up 93% of the sample from EN La Draga, with sheep/goats most abundant and mortality among sheep/goats and cattle conforming to a ‘meat’ strategy (Saña 2000).
With EN open settlements displaying scarcity of game, predominance of sheep (/goats) over cattle and pigs, and ‘meat’ mortality, faunal as well as archaeobotanical evidence broadly resembles that from south-east Europe and a similar regime of small-scale, integrated mixed farming has been suggested (Bernabeu et al. 1995, 269–281; Robb and Van Hove 2003). Early predominance of sheep is often less marked than in south-east Europe (though most Italian assemblages are small or cover broad temporal spans), but coastal sites in southern France and eastern Spain, specializing in sheep, have been interpreted as ‘beachheads’ of colonist farmers (Vigne 2000).
Later Neolithic open sites to varying degrees display the more even balance of domesticates seen in south-east Europe. In Dalmatia, sheep(/goats) were heavily dominant in the EN, but drop to c. 60% of domesticates at FN Bukovic (Miracle 2006). In Italy, at MN–LN Masseria di Gioia and FN Neto-Via Verga, small samples are again inconsistent. Southern Iberia is more informative, though LN–Copper Age (fourth to third millennium BC) open sites must be compared with EN caves such as Cendres, Nerja, Or, and Sarsa (Pérez Ripoll 1999). At the latter, sheep are the commonest domesticate, followed by pigs, with goats and cattle scarce, whilst on later open sites these four species are fairly evenly represented. Although the EN caves were apparently not specialized herding sites (see below, this section), the trend from sheep at EN caves to goats at later open sites, which does not match local grazing conditions, suggests an overall regional expansion of animal husband
ry (Pérez Ripoll 1999, 98), as has tentatively been proposed for LN south-east Europe. Charcoal data, however, suggest that the combined impact of crop and livestock husbandry on local vegetation was modest until the third millennium BC (Badal et al. 1994). Mortality data for the LN–Copper Age show high proportions of juvenile and sub-adult deaths for combined sheep and goats, consistent with management primarily for meat. An exception is third millennium BC Arenal de la Costa, where goats outnumber sheep and many animals reached old age. Together with much higher adult male survivorship for sheep than goats at third millennium BC Cerro de la Virgen, Valencina de la Concepción, and Zambujal, this implies management for different goals: sheep for meat (and conceivably wool); goats for milk (Pérez Ripoll 1999). The survival of most cattle to adulthood (with 20–40% achieving old age at fourth millennium BC Jovades and third millennium BC Ereta del Pedregal, Arenal de la Costa, and Cerro de la Virgen) and high proportions of adult males (including probable castrates) favour traction, as do ‘traction pathologies’ (Pérez Ripoll 1999). Unfortunately, faunal evidence is insufficient to explore the relationship between animal husbandry strategies and different types and sizes of sites.
At Molino Casarotto (c. 5000 BC) in the Alpine foothills of north-east Italy, lake-side huts with ceramics are associated with sparse remains of domestic animals and crops, but abundant red deer, boar, and gathered water chestnuts (Jarman 1971), whilst EN (sixth millennium BC) open sites further east, at Piancada and Nogaredo al Torre, are overwhelmingly dominated by domesticates. Only regional-scale analysis of seasonality and human mobility will clarify whether Molino Casarotto represents limited adoption of domesticates by foragers or seasonal foraging by farmers. In north-east Spain, EN open sites with domesticates and caves with mainly wild fauna (e.g. Chaves, Fosca) raise similar questions. Likewise, on the Atlantic coast of north-west Spain, domesticates are increasingly documented from the fifth millennium BC, but wild plants and animals predominate at some sites (González Urquijo et al. 1999). In the extremely heterogeneous western Alps, however, where domesticates predominate at EN open sites in valleys (e.g. Sion Planta), it seems implausible that caves with mainly wild animals (e.g. La Balme de Thuy) represent independent groups of foragers (Chaix and Sidi Maamar 1993).
Open-air sites devoted largely to hunting are also known from the LN, as at fourth millennium BC Roucadour or Mulino Sant’ Antonio (Albore Livadie et al. 1987–88) in the hills of southern France and southern Italy respectively. In southern Italy, ceremonial deposition of deer crania in the Ipogeo Manfredi and of wild animal bones in Grotta Scaloria, and use of deer canines as personal ornaments, all underline the symbolic importance of game (Robb 2007, 128). In north-east Spain bones of wild animals are selected in fourth millennium BC graves at Camí de Can Grau (Gibaja 2004). Painted human and animal representations in southern Italian caves, such as Grotta del Genovese, suggest a link between hunting and rites of social reproduction (Pluciennik 2002; Skeates 1994). The same may be argued for the potentially LN rock art in eastern Spain (McClure et al. 2008). The performance of these rites at a distance from agricultural settlements presumably played a part in regional-scale social integration and landscape enculturation (e.g. Bernabeu et al. 2003).
The use of caves to shelter livestock is widely attested in southern France, eastern Spain, northern and southern Italy, and along the eastern side of the Adriatic (e.g. Boschian and Montagnari-Kokelj 2000; Brochier et al. 1992). A distinction must be drawn, however, between the presence of dung with rich cultural material (e.g. Font Juvénal rock shelter, south-west France—Brochier 1991, 306), implying management of livestock among other activities, and thick layers of dung with little cultural material, reflecting specialized use for penning livestock. In southern France, specialized grottes bergéries seem particularly characteristic of the MN (fifth to fourth millennium BC) Chasséen period, with more mixed use attested before and afterwards (Brochier 1991, 2006). Corralling of livestock in caves likewise began, or intensified, a few centuries into the local Neolithic at Edera, Zingari, and perhaps Pupicina in the Istrian karst (Boschian and Montagnari-Kokelj 2000; Forenbaher and Miracle 2005), at Arene Candide on the coast of north-west Italy (Courty et al. 1991), and at Cendres in eastern Spain (although loss of cultivable land to rising sea level may have been a factor in the last case). From a few centuries after the inception of farming, therefore, in several regions of south-west Europe, livestock were sheltered in caves in greater numbers or for longer periods and, at Cendres, charcoal registers their impact on local vegetation (Badal 2002). MN livestock numbers remained well below the level needed to create the ‘degraded’ landscapes of the recent past, however, and at Bélesta cave in the Pyrenean foothills did not significantly transform vegetation until the Bronze Age (Brochier et al. 1998).
The scarcity of cultural material in grottes bergeries implies primary human habitation elsewhere. Moreover, whilst shed milk teeth (fallen from live animals) are present in some of these caves (e.g. Baume Ronze), their absence in others implies removal of at least part of the herd (Brochier 2006; Helmer et al. 2005) for at least part of the year. ‘Home’ settlements were perhaps not adjacent to the caves given that accumulations of dung were left in situ rather than removed to fertilize arable plots (Brochier 2006, 141), but hints of year-round slaughter at caves across the region (Forenbaher and Miracle 2005; Helmer et al. 2005, 180; Rowley-Conwy 1991) imply a distance that could be covered in hours rather than days. Caves with Neolithic evidence for penning of livestock are mostly at low to medium altitudes (0–600m) and several have faunal evidence (newborn lambs or kids, shed deciduous teeth of young adult sheep or goats) of late winter–early spring use (Forenbaher and Miracle 2005; Helmer et al. 2005; Rowley-Conwy 2000). Possibly livestock were removed in late winter from open settlements, where very young lambs and kids are scarce (Helmer et al. 2005), to safeguard growing crops or to shelter in caves from cold weather (especially important for pregnant/lactating females and newborn offspring). Neolithic flint scatters have been found above 1900 m in the southern French Alps (Walsh et al. 2006), close to potential summer pasture, but faunal remains from mid-altitude caves provide as much evidence of hunting as herding (e.g. Chaix and Sidi Maamar 2003). There is no hint that MN grottes bergeries were integrated in long-distance transhumant pastoralism (Brochier 2006) and some subsequently saw more intensive habitation and a more diverse domestic animal fauna, suggesting use as (or proximity to) mixed farming residential bases (Brochier 2006, 147–148). The same trend is apparent east of the Adriatic, at Pupicina (Forenbaher and Miracle 2005) and perhaps Grapčeva (Miracle 2006), and in southern Greece, at Zas. Grottes bergeries may thus represent a short-term phase in the expansion of mixed farming, rather than the development of specialized pastoralism.
Milk residues have been found in EN pottery from Mala Triglavca rock shelter in Slovenia (Soberl et al. 2008), and early lamb or kid mortality consistent with intensive ‘milk’ management at several caves across the region from Ciclami, Edera, Mitreo, Pupicina, and Zingari at the head of the Adriatic (Miracle 2006; Mlekuz 2005) to EN Arene Candide (Rowley-Conwy 2000) and MN Combe Obscure (Vigne and Helmer 2007) further west. ‘Milk’ mortality could be an artefact of seasonal use of caves, around lambing/kidding time, but is encountered in caves with evidence of slaughter in other seasons too (Mlekuz 2005). Moreover, contrasting levels of infant mortality for sheep and goats at EN–MN Arene Candide (Rowley-Conwy 2000) and EN Baume d’Oulen (Helmer et al. 2005) suggest a real difference in management goals (as also in LN–Copper Age southern Iberia—see above, this section). Relatively enriched stable nitrogen isotope ratios from Neolithic humans at EN–MN (sixth to fourth millennium BC) Arene Candide and, to a lesser extent, EN (sixth millennium BC) Pendimoun are consistent with diets high in animal protein (Le Bras-Goude et al. 2006).
Dairying yields more energy per animal than consumption of meat alone, but is more labour-intensive and riskier. Neolithic sheep and goats exhibit ‘meat’ mortality at Cavallo, Madonna, an
d Uzzo caves in southern Italy (Tagliacozzo 1993, 2000) and at most such sites in southern France (Vigne and Helmer 2007). ‘Milk’ mortality profiles are restricted to the earlier Neolithic in several caves at the head of the Adriatic (Miracle 2006; Mlekuz 2005), at Arene Candide in north-west Italy (Rowley-Conwy 2000), and (for cattle as well as sheep/goats) in open settlements of the lower Alpine valleys. Dairying was perhaps characteristic of the inception of farming, when limited clearance restricted numbers of livestock, and gave way to less labour-intensive ‘meat’ management when expanding clearance allowed larger herds (Legge 1981; also Rowley-Conwy 2000).
SYNTHESIS: SUBSISTENCE AND SOCIETY IN NEOLITHIC SOUTHERN EUROPE
Stable isotope evidence of the Neolithic human diet in southern Europe points to subsistence on terrestrial plants and animals, the remains of which are overwhelmingly from domestic cereal and pulse crops and livestock. Livestock were exploited for milk as well as meat and, alongside wild resources, contributed to a more diverse and balanced diet. Prevalent ‘meat’ mortality precludes widespread specialized dairying, however, and implies primary dependence on crops—especially for ‘village’ communities. At a few caves on the northern margins of the west Mediterranean, however, sheep and/or goat approximate to a ‘milk’ pattern, especially in the EN. Apparently not just an artefact of seasonal mobility, this might reflect greater dietary reliance on livestock where cultivation was unreliable, as human stable isotope data from Arene Candide perhaps imply.
Evidence of indoor storage, principally from burnt levels on south-eastern tells and rare south-western waterlogged sites, suggests diversified crop production by small residential groups or ‘households’, whilst dehusking by-products of hulled wheats on open-air sites are consistent with piecemeal, household-level food preparation. Some Mediterranean pulses did not initially spread to the north Balkans, but farmers grew a wide range of grain crops, thus spreading labour and risk and introducing dietary diversity. Labour-intensive pulses hint at small-scale cultivation, consistent with low anthropogenic impact on local and regional vegetation in charcoal and pollen records, respectively; crop weeds also suggest stable and probably intensive cultivation. Temporal and regional variability in archaeobotanical data arguably reflect differential preservation and retrieval, rather than differences in crop husbandry.