Reconstructing late Neolithic plant economies at the Eastern Adriatic site of Veli [ tak ( 5 th millennium cal BC )

Archaeological research into the plant economies of the late Neolithic Hvar culture in the Eastern Adriatic is rare. So the discovery of the late Neolithic settlement of Veli∏tak (formerly known as ∞ista Mala – Veli∏tak) in 2007 presented a unique opportunity to examine the development of farming in the Eastern Adriatic, as well as providing the first archaeobotanical results from an open-air Hvar culture village in Croatia (Fig. 1). Hitherto, only two cave sites, Grap≠eva (Borojevi≤ et al. 2008) and Turska Pe≤ (Reed 2015) in Croatia, and a few grain impressions from daub collected at Lisi≠i≤i, Herzegovina (Benac 1958.84; Hopf 1958), had provided archaeobotanical evidence of the plant economies of the Hvar culture. However, the social role of caves and the range of activities conducted in them are probably very different from those of open-air settlements (e.g., Bonsall, Tolan Smith 1997; Sampson 2008; Trantalidou et al. 2010). Therefore, this paper presents the archaeobotanical results from the 2007–2013 field seasons at Veli∏tak in order to explore plant exploitation at the settlement more fully, which is important for understanding the development of farming communities in the Eastern Adriatic.


Introduction
Archaeological research into the plant economies of the late Neolithic Hvar culture in the Eastern Adriatic is rare.So the discovery of the late Neolithic settlement of Veli∏tak (formerly known as ∞ista Mala -Veli∏tak) in 2007 presented a unique opportunity to examine the development of farming in the Eastern Adriatic, as well as providing the first archaeobotanical results from an open-air Hvar culture village in Croatia (Fig. 1).Hitherto, only two cave sites, Grap≠eva (Borojevi≤ et al. 2008) and Turska Pe≤ (Reed 2015) in Croatia, and a few grain impressions from daub collected at Lisi≠i≤i, Herzegovina (Benac 1958.84;Hopf 1958), had provided archaeobotanical evidence of the plant economies of the Hvar culture.However, the social role of caves and the range of activities conducted in them are probably very different from those of open-air settlements (e.g., Bonsall, Tolan Smith 1997;Sampson 2008;Trantalidou et al. 2010).Therefore, this paper presents the archaeobotanical results from the 2007-2013 field seasons at Veli∏tak in order to explore plant exploitation at the settlement more fully, which is important for understanding the development of farming communities in the Eastern Adriatic.

The Eastern Adriatic during the Late Neolithic
The division of the Neolithic is based on three major pottery stylistic traditions or cultures; Early (Impresso culture, c. 6000-5400 cal BC), Middle (Danilo culture, c. 5400-4900 cal BC) and Late Neolithic (Hvar culture, c. 4900-4000 cal BC).Hvar-style pot-ABSTRACT -The archaeobotanical remains from Veli∏tak are the first evidence of plant economies from an open-air settlement dating to the late Neolithic Hvar culture in .The results presented here are from the 2007-2013 field seasons.Based on an examination of carbonised macro-remains, it is suggested that emmer, einkorn, and barley were the main crops at Veli∏tak, along with lentils, bitter vetch, and possibly peas and flax.Wild plants were also exploited, with evidence of wild fruits, such as cornelian cherry.Similarities with archaeobotanical finds from the early/middle Neolithic (c.6000-4900 cal BC) also suggest that plant economies remained relatively unchanged during the Neolithic. .Predstavljeni rezultati so iz podatkov, pridobljenih med izkopavanji med letoma 2007 in 2013.Na podlagi preiskave karboniziranih makroostankov ugotavljamo, da so dvozrnica, enozrnica in je≠men predstavljali glavna ∫ita na Veli∏taku, navzo≠i pa so tudi ostanki le≠e, le≠nata gra∏ica ter morda grah in lan.Nabirali so tudi divje rastline, saj so ohranjeni ostanki divjega sadja kot je rumeni dren.Podobnosti z arheobotani≠nimi ostanki iz ≠asa zgodnjega in srednjega neolitika (ok.6000-4900 cal BC) ka∫ejo, da je rastlinska ekonomija na tem obmo≠ju ostala relativno nespremenjena v celotnem obdobju neolitika.

IZVLE∞EK -Arheobotani≠ni ostanki z najdi∏≠a Veli∏tak predstavljajo prve dokaze o rastlinski ekonomiji iz naselbine na prostem, ki datira v pozni neolitik hvarske kulture na
KEY cereal cultivation;archaeobotany;Croatia DOI> 10.4312\dp.43.19 tery is decorated with a rich variety of incised and painted designs and is best known from Grap≠eva Cave on the island of Hvar (Novak 1955;Forenbaher, Kaiser 2008;Forenbaher et al. 2010).Traces of paint, which would have been applied after firing, indicate that several pigments were used, including red ochre and cinnabar (Forenbaher et al. 2010).So far some 30 Hvar culture sites have been identified, showing that this cultural tradition extended throughout Dalmatia (Croatia) and the hinterlands (Hercegovina).Most of these sites are caves, probably because they are easier to locate than flat open-air settlements.
Research on the Neolithic in the Eastern Adriatic has largely focused around these stylistic and technological changes (e.g., Chapman 1988;Spataro 2002;McClure et al. 2014), while in comparison very little is known about possible underlying socio-economic and cultural shifts.
For the late Neolithic (c.4900-4000 cal BC), plant remains are equally rare, having only been identified from two cave sites in coastal Croatia: Grap≠eva (Borojevi≤ et al. 2008) and Turska Pe≤ (Reed 2015).At Grap≠eva, the evidence consists of a few grains of emmer, einkorn, naked wheat (Triticum cf.aestivum), and lentil (Lens culinaris), as well as shell fragments of almond (Amygdalus communis), acorns (Quercus sp.), and juniper (Juniperus sp.).The naked wheat grain from Grap≠eva was also securely radiocarbon dated to 4838-4712 cal BC, confirming its presence during the late Neolithic (Borojevi≤ et al. 2008).From Turska Pe≤, emmer, einkorn, barley, naked wheat (Triticum turgidum ssp.durum/Triticum aestivum ssp.aestivum), possible spelt (Triticum cf.aestivum ssp.spelta), and broomcorn millet (Panicum miliaceum) were identified (Reed 2015).In addition, a large number of weed species were also recovered and are thought to have been associated with episodes of dung burning, possibly to clear the cave of excess waste during seasonal habitation of the cave by herders and livestock (ibid.).The only exclusively Hvar culture settlement excavated prior to Veli∏tak was at Lisi≠i≤i (Hercegovina) in the 1950s, where the only evidence of plant remains came from grain impressions found in daub fragments.These were only briefly listed as wheat, barley, and rye (Benac 1958.84;Hopf 1958), although it is likely that the rye was a weed, as evidence suggests that rye was not grown as a crop until at least the Iron Age (Behre 1992).
Zooarchaeological studies of Neolithic assemblages in the Eastern Adriatic have also demonstrated the predominance of domesticated species such as cattle, ovicaprids, and pig (Miracle, Forenbaher 2005;Moore et al. 2007a;2007b;Legge, Moore 2011).Sites such as Pokrovnik and Danilo-Bitinj suggest that hunting played a minor role in subsistence during the early and middle Neolithic (Moore et al. 2007a;2007b;Legge, Moore 2011), while stable isotope studies of early Neolithic human remains suggest that marine resources were also consumed by some individuals during this time (Lightfoot et al. 2011).Stable isotope studies on animal remains have also provided information about animal management strategies, suggesting that pig management changed during the Neolithic, possibly through different foddering practices, while in contrast, cattle and ovicaprid management remained unchanged (Zavodny et al. 2014).

The case study
The archaeological site of Veli∏tak is situated in the Velim Valley to the north of Vodice in northern Dalmatia, Croatia (Fig. 1).The site was discovered in 2007 during agricultural work, and excavations subsequently were begun in the same year by Emil Podrug of πibenik Municipal Museum (Podrug 2010;2014).The excavations are continuing, but after nine excavation campaigns (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015), an area of 335m 2 has been opened.A large quantity of classic Hvar culture pottery has been recovered (Fig. 2), along with tools made of animal bone, knapped and polished stone, as well as jewellery and polished shells (Spondylus gaederopus) (Podrug 2010;2013).The Veli∏tak pottery corresponds to the initial 'Outlined' decorative stage (e.g., sub-phase 1.1 of Grap≠eva -'bordered' classic Hvar) (Forenbaher, Kaiser 2008.51-52).This is supported by six radiocarbon dates that show the settlement was founded At the subsoil level, numerous pits and other dugout features have been revealed, often intersecting each other, suggesting that they were not all dug and used at the same time (Fig. 3).Most of the pits are round or oval in plan and cylindrical or spherical in shape, and it is suggested that some were probably used for grain storage (Podrug 2013).The pits were usually filled with pottery and animal bones, which, along with numerous layers of soot and ash,

Fig. 3. Plan of the excavated area at the subsoil level and the eastern profile at Veli∏tak.
suggest that the pits were eventually used as disposal places for household refuse and other communal waste (ibid.).Several fragments of house floors, fireplaces and hearths are also present in undisturbed general cultural layers (in the northern part of the excavation area, these layers are preserved up to 60cm in thickness).Overall, the stratigraphy at Veli∏tak shows continuous successive occupation at the settlement over several generations.

Material and methods
During 2007-2013, 52 samples were collected from 24 contexts relating to house floors, pits and fireplaces from trenches A to F (Tab. 1).A total of 52 samples were collected, totalling 571.5 litres of sediment, and processed through bucket flotation using 1mm and 250μm mesh sizes.
The plant taxa were established with a low resolution (7-40x) binocular microscope and comparisons made from the modern reference collections at the School of Archaeology and Ancient History at the University of Leicester.The nomenclature of scientific plant names for cultivars follows Daniel Zohary et alii (2012) and for wild plants Tom G. Tutin et alii (1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980).All floats were 100% sorted and the charcoal volumes recorded.A standardised counting method was used, whereby each grain counts as one and the whole grain equivalent (WGE) was estimated for fragments of grains.Glume base fragments were counted as one unless clearly representing part of another fragment, while whole spikelet forks were counted as two glume bases.The fruit and weed seeds were counted as one, even when only a fragment was found, except when large seeds were broken and the parts were clearly from the same seed (e.g., Cornus mas).

Results
Carbonised plant remains were recovered from 45 of the 52 samples, as well as two mineralised seeds of Buglossoides from samples 34 and 62.A total of 3491 seeds were identified, with a further 973 indeterminate plant items.Seed density was particularly low, with 37 samples having a density of less than 1 seed per litre of sediment (Tab. 1, Fig. 4).The two samples with the highest density were 184 and 185, both from the same lower half of a pit fill, with a seed density of 16.7 and 93.5, respectively.Charcoal density was also low, with all the samples having less than 2.0cm 3 per litre.

Crops
Cereals were the most commonly recovered plant remains at the site, accounting for 94% of the identified assemblage, not including cereal fragments and indeterminate remains (Fig. 5).Of the cereal grains, 93% of the remains were of barley (Hordeum vulgare ssp.vulgare), totalling 879 grains, although 838 of these were recovered from samples 184 and 185 (Fig. 6).
Cereal chaff (cereal husk separated by winnowing and/or threshing), which accounts for 16% of the assemblage, is dominated by glume wheat glume bases.Preservation was particularly poor, but where possible, einkorn, emmer and the 'newtype' glume wheat were identified, although in low numbers (Tab.2).The identification of the 'new type' of glume wheat glume base was based on observations made by Glynis Jones et alii (2000) and Marianne Kohler-Schneider ( 2003) and stood out, as they were more 'robust' than the emmer and einkorn glume bases.
A small number of other crops were also identified from pit samples, including four lentils (Lens culinaris), two possible peas (Pisum sp.), one bitter vetch (Vicia ervilia) and two possible flax seeds (Linum sp.) (Tab.3).

Wild plants
Wild plants make up 5% of the assemblage, and consist of fruit remains, possible weed species and other wild plants (Fig. 4; Tab. 3).Of the fruits, only one cornelian cherry (Cornus mas), one rose hip (Rosa canina), and one bramble fruit seed (Rubus sp.), as well as two acorn fragments (Quercus sp.) were identified within four samples taken from pits (Tabs.1, 3).Due to poor preservation, the vast majority of the wild plant remains are identified only to family or genus, with the highest number of remains being identified as either large-seeded legumes (Fabacaeae) or large-and small-seeded grasses (Poaceae) (Tab.3).

Formation processes
In order to understand the archaeobotanical results, it is important to explore the formation processes at the site in order to identify any possible bias in the samples that may influence interpretations.The most common form of preservation at Veli∏tak was carbonisation or charring, which results from organic material being exposed to heat either accidentally or deliberately, such as cooking, burning rubbish or fuel (Hillman 1984;Miller, Smart 1984;Charles 1998;Valamoti, Charles 2005; Van der Veen 2007).Thus, carbonised plant remains will be heavily biased towards items that come into contact with fire more frequently and survive the charring process (Dennell 1972;Hillman 1981;Jones 1981;Boardman, Jones 1990;Van der Veen 2007).
The deposition of these remains within the archaeological record also needs to be considered.For example, Richard N. L. B. Hubbard and Alan J. Clapham (1992) suggested that charred assemblages can be divided into three groups: Class A, where remains have been burnt within the context within which they are recovered; Class B, where an assemblage derives from one burning event that was subsequently moved to the context (secondary deposition); and Class C, where the assemblage derives from different charring events that were subsequently deposited within the same context.Seed density has also been used to reflect the rate of deposition, whereby a low density of plant remains indicates slow accumulation, while high densities suggest rapid deposition (Jones 1991).
At Veli∏tak, seed density was particularly low, with 37 samples having a density of less than 1 seed per litre of sediment (Fig. 4).In addition, only ten samples were collected from a fireplace or area where burning occurred and only a few seeds were recovered.The remaining contexts and the overall low seed and charcoal densities of the samples suggests that the remains probably reflect Class C assemblages, which derive from different charring events that were subsequently deposited within a secondary or even tertiary context.The only exceptions to this are the two samples with particularly high densities, 184 and 185, which were both from the same lower half of a pit fill.Both are dominated by barley grains, although poor preservation led to a high number of grains being identified as 'cerealia' fragments.The samples also contain a small number of wheat grains and chaff, a few pulses and a handful of weeds.In contrast, samples 179 and 180 from the upper half of the same pit contain only a small number of barley grains (<13).These differences may suggest that the charred plant remains at the bottom of the pit were deposited more intensively, as part of one activity, or more rapidly, from a number of activities, than at the top of the pit.
But what activities might these plant remains represent at Veli∏tak?Since the 1970s researchers have determined that carbonised remains are more likely to result from food production processing, e.g., crop processing, rather than food consumption (Knörzer 1971;Dennell 1972;1974;1976;Hillman 1984;Jones 1984).Predictive models have since been created to identify which stage of the crop processing sequence an assemblage represents, based on the assumption that each stage produces a characteristically different ratio of cereal, chaff and weeds within the sample (Hillman 1984;Jones 1984;Van der Veen 1992;Van der Veen, Jones 2006).
The general uniform composition of the assemblage at Veli∏tak of cereal grains, chaff and weeds would suggest that the remains represent harvested cereals.Overall, the low seed density makes it difficult to apply any of the predictive models, as over 50 seeds per sample are required for a meaningful result (e.g., Bogaard 2004.63).Only four samples (nos. 179, 184, 185, 258) have over 50 plant items, although sample no.179 is dominated by cereal fragments.The dominance of barley grain within sample nos.184 and 185 and their high density within the context may suggest that these samples represent a deposit of a barley crop.In contrast, sample no.258, which is from a pit a few metres away, has a high proportion of glume wheat glume bases and  no weed remains.This sample may therefore represent the dehusking of cleaned glume wheats (Reed 2015).Some suggest that the daily processing of stored glume wheats occurred within the household, where the waste (cereal chaff) was then swept into fires and carbonised (cf.Hillman 1984;Gregg 1989;Meurers-Balke, Lüning 1992;Bogaard 2004.68;Kreuz 2012).The waste from these fires may have then been deposited outside the house, and so sample no.258 may indicate the secondary or tertiary deposition of discarded wheat chaff.Archaeological finds of querns and flint sickle blades also attest to crop processing activities at Veli∏tak.However, the recovery of only one barley rachis at the site may suggest that either barley was mainly processed away from the settlement or that the more fragile barley rachis did not survive the carbonisation process (cf.Dennell 1976;Hillman 1981;Boardman, Jones 1990).
There are advantages and disadvantages to growing both naked and hulled barley.In terms of crop processing, naked barley is easier to process, as the grain is looser in the spikelets and can be easily threshed, while hulled barley requires an extra crop processing stage to release the grain from the tighter chaff.On the other hand, it has been suggested that hulled barley has a greater tolerance to adverse environmental conditions (e.g., soil salinity, parasites) (Ceccarelli et al. 2001;Lister, Jones 2013), while it has been suggested that naked barley grains are

Tab. 3. The total number of other crops and wild plants per sample at Veli∏tak. (n) = cf.
more susceptible to increased moisture in the atmosphere and to fungal attack (Van der Veen 1992.74).
In the archaeobotanical results at Veli∏tak, barley dominates the crop remains, accounting for 94% of the cereals.As already discussed, sample no.185 (lower half of a pit fill) contained 752 barley grains, but very few other plant remains, although over 1000 cereal grain fragments were also identified.Due to poor preservation, most of the grains could not be identified as either hulled or naked varieties, but many did have the more rounded shape of naked barley, while a couple were more angular, characteristic of hulled barley.A few grains also had the characteristic lateral twist of six-rowed barley (Zohary et al. 2012.56).Thus, poor preservation prevents further discussion as to the preference for hulled or naked barley growing at Veli∏tak from the current assemblage.
Apart from this pit feature (samples nos. 179, 180, 184, 185), the quantity of barley, emmer and einkorn grains is the same, suggesting that barley may not have been the main 'preferred' crop at the site.This is further supported by the relatively high number of glume wheat glume bases seen in samples nos.49 and 258.The ratio of cereal species may therefore indicate differences in preservation rather than the predominance of barley at the site.

Plant economies of the Hvar culture
The entire area of the Eastern Adriatic, from the Trieste Karst in the north to the Gulf of Kotor in the south, is characterised by alternating dry barren karst landscapes, sunken karst fields and ridges, limestone plateaus and fertile valleys filled with springs, underground streams and gorges (Maga∏ 1998.195)Tab. 3. continued recovered from this region, has led researchers to believe that this area provided the best conditions for Neolithic crop agriculture (e.g., Brusi≤ 2008.63-64).In contrast, the cave sites are located in mountainous areas away from land suitable for agriculture, and are usually interpreted as seasonal herding camps, marked by low density pottery and animal bone finds (e.g., Miracle, Forenbaher 2005;Mleku∫ 2005).This is probably the case at Turska Pe≤, where the archaeobotanical remains identified an area resulting from human activities (where cereal remains were recovered) and an area that was probably subject to the periodic burning of animal dung (where wild, weed species were recovered) (Reed 2015).An exception to this is Grap≠eva, where the large number of animal remains, artefacts and scattered human remains suggests ritual activities, where feasts, offerings and secondary burials took place (Forenbaher, Kaiser 2008;Forenbaher et al. 2010).
Therefore, the three late Neolithic sites present archaeobotanical remains from a ritual context, a possible herding camp and a valley settlement.Although they cannot be directly compared, as bias in the range of plant remains found in such locations will be present, we can begin, if only tentatively, to reconstruct the plant economies of the late Neolithic period in the Eastern Adriatic.For example, emmer, einkorn, and barley are the most common cereals identified from the three sites.This corresponds with archaeobotanical evidence from northern Italy (Rottoli 2014.76),Slovenia (Tolar et al. 2011), and continental Croatia (Reed 2015) and would suggest that these were the main cereals grown at the time.Evidence of naked wheat, spelt, rye, and millet is sporadic, so it is unlikely that they were grown as part of the crop package during the late Neolithic in this region.Unfortunately, the general low density of plant remains at Veli∏tak also means that the analysis of samples in relation to cultivation methods (i.e.whether crops were manured, weeded, etc.) is not possible at present.Other crops were also probably grown, including pulses, such as lentil, pea and flax.The recovery of cornelian cherry seeds (Cornus mas), brambles (Rubus sp.), oak (Quercus sp.), dog rose (Rosa canina), and dwarf elder (Sambucus ebulus) suggests that local woodland and woodland fringes were exploited by the inhabitants of Veli∏tak to supplement the diet.
Compared to the earlier Neolithic, the plant spectra in coastal Croatia does not seem to change, with the main founder crops -emmer, einkorn, barley, lentil, pea, and flax -being found throughout the Neolithic (Reed 2015).Although naked wheat (Triticum durum/turgidum/aestivum) has also been discovered at the early Neolithic site of Tinj-Podlivade (Huntley 1996), the middle Neolithic site of Danilo-Bitinj (Reed 2006), and late Neolithic levels at Grap≠eva cave (Borojevi≤ et al. 2008), it is unlikely that the few grains recovered, as well as the single grain from Veli∏tak, means it was cultivated during the Neolithic in the Eastern Adriatic.

Conclusion
The analysis of carbonised plant remains from Ve-li∏tak presents the first archaeobotanical evidence from an open-air settlement dating to the late Neolithic Hvar culture in the Eastern Adriatic.The formation processes inferred from the plant remains suggest that they probably resulted from a number of charring events that were then discarded as secondary or tertiary deposits.The only possible exception to this was the recovery of a relatively large deposit of barley grains in the bottom of a pit feature.The remains are probably from a barley crop that were carbonised and, as there was very little sign of burning in the pit, dumped as waste.The remains of a relatively high number of glume wheat (Triticum monococcum/turgidum ssp.dicoccum/'new-type' glume wheat) glume bases within a pit a few metres away also suggests that crop processing of glume wheats occurred at the site, probably within the home where the by-products could be easily thrown into the hearth and then re-deposited within the pit as waste.Thus the results from the 2007 to 2013 field seasons suggest that emmer, einkorn and barley were grown as the main crops at Veli∏tak, along with lentil, bitter vetch and possibly pea and flax.However, further work is clearly needed in recovering archaeobotanical remains from late Neolithic sites in order to understand more fully the development of subsistence economies in the Eastern Adriatic.
sometime after 5000 cal BC and lasted until 4700 cal BC (McClure et al. 2014.1027,T.1), making Veli∏tak the earliest known Hvar culture site.

177 178 179 180 184 185 193 194 216 218 221 249 250 258 259 TOTAL OTHER CROPS
. Within this landscape, early Neolithic occupation seems to have clustered in three main areas: one cluster of cave and open-air settlements in coastal Istria and the Kvarner islands, a second cluster of open-air settlements in the Zadar and πibenik area, and a third cluster of cave sites to the south in the Pelje∏ac peninsula and within the hinterlands of Herzegovina (Marijanovi≤ 2007).The fertile valleys of the Zadar/πibenik area, and the fact that most of the archaeobotanical evidence has been