Kadowaki, S., ..., F. Guliyev, and Y. Nishiaki (2015) Geoarchaeological and palaeobotanical evidence for prehistoric cereal storage at the Neolithic settlement of Göytepe (mid 8th

lable at ScienceDirect

Journal of Archaeological Science 53 (2015) 408e425

Contents lists avai

Journal of Archaeological Science

journal homepage: http: / /www.elsevier .com/locate/ jas

Geoarchaeological and palaeobotanical evidence for prehistoric cerealstorage in the southern Caucasus: the Neolithic settlement of G€oytepe(mid 8th millennium BP)

Seiji Kadowaki a, *, Lisa Maher b, Marta Portillo c, Rosa M. Albert d, Chie Akashi e,Farhad Guliyev f, Yoshihiro Nishiaki g

a Nagoya University Museum, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japanb Department of Anthropology, University of California, Berkeley, 232 Kroeber Hall, Berkeley, CA 94720-3710, USAc GEPEG, Department of Prehistory, Ancient History and Archaeology, University of Barcelona, c/ Montalegre 6-8, 08001 Barcelona, Spaind ICREA/GEPEG, Department of Prehistory, Ancient History and Archaeology, University of Barcelona, c/ Montalegre 6-8, 08001 Barcelona, Spaine JSPS Research Fellow, The University Museum, The University of Tokyo, Hongo 7-3-1, Bunkyo, Tokyo 113-0033, Japanf Museum of Archaeology and Ethnology, The Institute of Archaeology and Ethnography of National Academy of Sciences, Azerbaijang The University Museum, The University of Tokyo, Hongo 7-3-1, Bunkyo, Tokyo 113-0033, Japan

a r t i c l e i n f o

Article history:Received 3 December 2013Received in revised form19 September 2014Accepted 27 October 2014Available online 5 November 2014

Keywords:StorageNeolithicPhytolithsDung spherulitesMicromorphologyPalaeobotanySouthern Caucasus

* Corresponding author.E-mail addresses: [email protected]

berkeley.edu (L. Maher), [email protected] (M.(R.M. Albert), [email protected] (C. Akashi), [email protected] (Y. Nishiaki).

http://dx.doi.org/10.1016/j.jas.2014.10.0210305-4403/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

This paper presents direct evidence for cereal storage by Neolithic farmers in west Asia. Storage featuresanalyzed this study are circular clay bins that frequently occur at Neolithic settlements (8th millenniumcal. BP) in the southern Caucasus. We examined contexts and uses of clay bin features at the Neolithicsettlement of G€oytepe (Azerbaijan). We analyzed biogenic microfossil evidence (primarily from phyto-liths and dung spherulites) and the sediments of the clay bins through micromorphology, in combinationwith their associated charred macrobotanical remains. While phytoliths and charred botanical remainsindicate direct remnants of stored plants, mainly chaffs, micromorphology and the analyses of faecalspherulites allow us to examine depositional and diagenetic processes of the archaeological sedimentsinside and outside these features. As a result, one of the clay bins was found to retain deposits at its baseexhibiting high concentrations of grass phytoliths with relatively high proportions of inflorescences andlow percentages of anatomically connected phytoliths in comparison with its upper fill deposits andareas outside of the bin. These finds, combined with the association of the two complete grinding stonesinside the bottom of the bin, suggest that the remains of cereal processing activities, specificallydehusking, may have been placed in the bin. This interpretation is corroborated by the recovery ofcharred rachises and chaffs of wheat and barley as well as micromorphological observations that thebottommost fill of the bin consists almost entirely of grass phytoliths with few very small charcoalfragments and fine amorphous organic matter.

© 2014 Elsevier Ltd. All rights reserved.

1. Early farming settlements in the southern Caucasus andstorage features

Storage is one of the main foci of archaeological research onprehistoric communities during the transition from foragers toearly farmers over the late Pleistocene and early Holocene. This

p (S. Kadowaki), maher@Portillo), [email protected][email protected] (F. Guliyev),

fundamental socio-economic practice is essential to settled com-munities as it provides a wide range of insights into past life-ways,such as subsistence practices, settlement patterns, and social re-lations (Flannery, 2002; Kuijt, 2008, 2011). However, direct evi-dence for storage practices by prehistoric hunter-gatherers or earlyfarmers has rarely been identified due to poor preservation oforganic material and, more probably, because stored food wasusually consumed subsequently and left only accidentally asarchaeological remains. In fact, there are many archaeological in-stances of the re-use of abandoned storage facilities for otherpurposes, such as receptacles for domestic refuse or human burials(e.g., Banning et al., 1992) that complicate (or even obliterate) tracesof the original usage of apparent storage features.

Delta:1_given name

Delta:1_surname

Delta:1_given name

Delta:1_surname

Delta:1_given name

Delta:1_surname

Delta:1_given name

mailto:[email protected]








http://crossmark.crossref.org/dialog/?doi=10.1016/j.jas.2014.10.021&domain=pdf

www.sciencedirect.com/science/journal/03054403

http://www.elsevier.com/locate/jas

http://dx.doi.org/10.1016/j.jas.2014.10.021



S. Kadowaki et al. / Journal of Archaeological Science 53 (2015) 408e425 409

Therefore, identification of prehistoric storage facilities hasdrawn predominantly on indirect evidence from architectural re-mains. Such examples in west Asia, where one of the oldest agro-pastoral practices emerged, include plastered or stone-lined pits(Bar-Yosef, 1998), small enclosed spaces in buildings (Verhoehen,1999; Banning, 2003; Goring-Morris and Belfer-Cohen, 2008), ar-chitecture with raised/paved floors (Garfinkel, 2002; Kuijt andFinlayson, 2009), and various bin features (Rollefson et al., 1992;Molist, 1998; Moore, 2000). However, discoveries of organic re-mains in direct association with these features are extremely rare(e.g., horsebean seeds at Yiftahel: Garfinkel, 1987).

Here we present direct evidence for prehistoric cereal storage.Storage features analyzed and discussed in this study occur atNeolithic settlements (8th millennium cal. BP) in the southernCaucasus in an area south of the Greater Caucasus Mountains andbetween the Black and Caspian Seas (Fig. 1). These Neolithic set-tlements, including G€oytepe (Fig. 2), represent fully-fledgedfarming communities since botanical and faunal remains fromthe sites are rich in domesticated species, including naked andhulled wheat/barley, legumes, sheep, goats, and cattle (Chataigner,1995; Badalyan et al., 2007; Lyonnet et al., 2012). Agrarian life-waysof the Neolithic inhabitants are also directly and indirectly illus-trated by their material cultures, including sickle elements, some ofwhich are still hafted (Narimanov, 1987), large sizes and quantitiesof food processing tools (Hamon, 2008), pottery assemblages(Palumbi, 2007; Lyonnet et al., 2012), a wide range of bone im-plements made of domesticated species (Badalyan et al., 2007), anddensely constructed architecture, including mud-walled circularhouses and presumed storage bins (Fig. 3). Architectural featuresare particularly well-developed in the Neolithic sites located in themiddle Kura and Araxes Rivers. As these sites are more or lesssimilar to each other in the characteristics of material culture(Narimanov, 1987; Kiguradze and Menabde, 2004), as well as thereported ranges of radiocarbon dates (ca. 6900e6400 14C yr BP)

Fig. 1. Map of the southern Caucasus, showing the location of G€oytepe and other Neolithic s(3) Shomutepe, Gargalartepesi, Toiretepe, (4) Imiris Gora, Shulaveri Gora, Khramis Didi Gor

(Badalyan et al., 2010; Lyonnet et al., 2012), they are often groupedin a chrono-cultural unit, called the “Shomutepe-Shulaveri” or“Aratashen-Shulaveri-Shomutepe” culture.

Circular clay bin features, less than 1 m in diameter and depth,are often interpreted as storage facilities (Narimanov, 1992;Badalyan et al., 2007) and commonly occur in the Shomutepe-Shulaveri sites. They are semi-subterranean structures, partiallydug into surrounding surfaces and fill. These types of clay binsoccur in association with circular mud-walled structures andcurved appendicular walls that constitute domestic areas of thesettlement. Round clay features are frequently shown in architec-tural plans of the Shomutepe-Shulaveri sites, such as Shomutepe(Narimanov, 1992) and Shulaveri Gora (Chataigner, 1995). Morerecently, the excavators of Aratashen in the Ararat plain found smallcircular features distributed in clusters inside and outside buildingsthey call houses and interpreted them as silos that contained grainor tools (Badalyan et al., 2007). However, there have been fewdetailed analyses of the clay bins despite their potential signifi-cance as evidence for many socio-economic aspects of these earlyagro-pastoral communities.

In an effort to provide direct evidence for cereal storage, wefocus here on the detailed, multi-scalar analysis of several clay binsfrom domestic contexts at the Neolithic site of G€oytepe. Weanalyzed phytoliths (plant silica cells; Piperno, 2006), dungspherulites (microscopic calcitic particles produced in the guts ofanimals, especially ruminants, and later expelled in excrement;Canti, 1999), in situ processes of construction, deposition, use andpreservation through micromorphology of the clay bins, in com-bination with a more conventional study of associated charredbotanical remains. While phytoliths and charred botanical remainscan represent direct remnants of stored plants, the analyses of dungspherulites and use of micromorphology allowed us to examinemicro-depositional processes of the bins themselves, their fill, andsediments outside the features.

ites affiliated with the Shomutepe-Shulaveri culture. (1) G€oytepe, (2) Guseingulutepesi,a, (5) Aruchlo I, (6) Aratashen, (7) Aknashen.

Fig. 2. Topographic map of G€oytepe, showing the excavation squares and a disturbed area to the northeast.

S. Kadowaki et al. / Journal of Archaeological Science 53 (2015) 408e425410

2. Storage features at G€oytepe in the middle Kura (westernAzerbaijan)

2.1. Chrono-cultural and natural context of G€oytepe

G€oytepe (40�58011.8400N 45�42017.8100E, ca. 430 m a.s.l.) is aNeolithic site in the middle part of the Kura River in the southernCaucasus (western Azerbaijan). The site is located south of the KuraRiver and sits on a large alluvial plain at the northern foot of theLesser Caucasus Mountains, where volcanic and pyroclastic rocks(mainly rhyolitic-andesitic) associated with limestone bedrockform geological substrates (Azerbaijan SSR Academy of Sciences,Institute of Geography (1963):26e27). The site is on alluvial de-posits with silts and clays that were deposited during the Pleisto-cene by rivers flowing from the Lesser Caucasus to the Kura River.Such alluvial plains are covered mainly with chestnut soil(Azerbaijan SSR Academy of Sciences, Institute of Geography,1963:82e83), stretching in a southeast-northwest direction fromthe Ganja-Qazax areas (western Azerbaijan) to the Kwemo-Kartliregion (eastern Georgia), where a number of Neolithic tepe sites,including G€oytepe, are located. Ground and river water is readilyavailable on these alluvial plains, which today support agriculturalfields of wheat and grape orchards, as well as meadows for grazing

livestock including cattle, sheep, and goats. The mean annual pre-cipitation around G€oytepe is ca. 300 mm/yr (World MeteorologicalOrganization, 2014). Precipitation occurs throughout the year, butis concentrated in the spring to early summer months (March-eJune). The mean monthly temperature in January (the coldestmonth) is �2.3e6.5 �C and in July (the warmest month) rangesbetween 19.5 and 31.7 �C. The local vegetation around the site issemi-desert or steppe including Bromus, Eremopyrum, Atriplex,Chenopodium, Malva, Heliotropium, Alhagi among others. Artemisiais dominant at the northern bank of the Kura River.

G€oytepe (meaning green hill) is an anthropogenic mound thatwas created by repeated occupation and the construction of manymud-walled buildings over a period of less than two hundred years.The mound measures ca. 145 m in diameter and is raised 8 m inheight above the surrounding alluvial plain. The site was originallyinvestigated by Narimanov (1987). In 2008, a joint Azerbaija-nieJapanese research group renewed excavations at the site(Guliyev and Nishiaki, 2012) following a small-scale sounding by anAzerbaijanieFrench mission. The recent excavations (2008e2011)took place over an area of 1000 m2 and revealed ca. 11 m-thickNeolithic deposits divided into fourteen building levels (Levels1e14 from the surface). Despite the thickness of the Neolithiclayers, radiocarbon dates from these levels cluster in a rather

Fig. 3. Architectural plan revealed during the 2008e2011 seasons at G€oytepe. Four clusters of clay bins in Levels 3, 4, 5, and 10 are circled by dashed lines.


narrow time span of 6700e6500 14C yr BP, falling in the middle ofthe Shomutepe-Shulaveri cultural period (Guliyev and Nishiaki,2012).

Excavations at the site have recovered dense distributions ofcircular mud-brick houses and a rich repertoire of artifacts (i.e.,chipped and ground stones, pottery sherds, and bone tools) thattypify the Shomutepe-Shulaveri culture. Well-preserved botanicaland faunal remains include high proportions of domestic species,including wheat, barley, legumes, goat, sheep, and cattle, indicatingamajor role of agro-pastoral practices in the subsistence of G€oytepeinhabitants. Thus, the Neolithic site at G€oytepe has been clearlycontextualized in terms of chronology and cultural history, and canbe considered a suitable site for detailed examination of the prac-tices of early agricultural communities in the southern Caucasus.

2.2. Excavation and sampling of clay bins at G€oytepe

Recent excavations of the site uncovered a number of clay binsthat are often clustered adjacent to buildings and outdoor wallsconnecting the buildings. At least four such clusters have beenrecovered in Levels 3, 4, 5, and 10 (Fig. 3), indicating that thesefeatures continued to be common architectural components duringmost of the Neolithic occupation at G€oytepe. Among the fourclusters of bins, those in Levels 3, 4, and 5 were excavated by theAzerbaijani investigation, while those in Level 10were excavated bythe Japanese mission during the 2008e2011 seasons. A cluster ofbins in Level 3 is located at Square 1Awithin an apparent courtyardsurrounded by four round buildings connected to each other bycourtyard walls. Another cluster is found in Level 4 of Square 3AII,also located in an outdoor space surrounded by round buildingsand courtyard walls. Bins in Level 5 of Square 4A are arranged in arow along a wall that is attached to a round building. Lastly, acluster in Level 10 is located in an open space near two roundbuildings connected with each other by a courtyard wall.

Another significant aspect of bins at G€oytepe is that they aresituated adjacent to open spaces that often contain complete or

nearly complete ground stones (particularly food processing tools),other artifacts that were apparently left as de facto refuse, as well ashigh concentrations of charcoal fragments and ash. Our assessmentof the context and finds from these areas suggests that such openspaces probably represent places where domestic activities tookplace.

We sampled sediments from various contexts, primarily insideand outside the bins as well as building floors and fill in Level 10 ofSquare 4B and in Level 4 of Square 3AII. Because of the largenumber of sediment samples, the analysis is still in progress. Here,we report on a detailed examination focusing on three bins fromthese levels (Table 1). This initial study on storage features atG€oytepe selected samples from a clay bin (4BIIX-94 in Level 10) thatis notably associated with possible primary deposits and in situartifacts, as well as those from two bins (3AII-1 and 2 in Level 4)that appear to be filled with secondary deposits like most of the binfeatures recovered at the site. The three bin features analyzed inthis study are similar to each other in their size, form, and contextin the settlement (i.e., located in the open-air activity space), andconsidered fairly representative of other similar bin features com-mon at the site.

2.2.1. Level 4 in Square 3AIIThe excavation of Square 3AII by the Azerbaijani team recovered

more than ten clay bins belonging to Level 4 (Fig. 4). They areclustered in an apparent outdoor area (ca. 4 � 3 m), surrounded byround buildings and courtyard walls. Adjacent to this cluster is anopen space (ca. 3 � 3 m), where we interpret domestic activities tohave taken place based on the architectural layout, although theexcavation did not record any features, artifacts, or the formationprocesses (and taphonomic) nature of the deposits in this area.

Phytolith and faecal spherulites analyses were conducted onsediments sampled from the fills (at the middle and bottom por-tions of the bins) and the bottom walls of two bins (Table 1and Fig. 4: 3AII-1 and 3AII-2), as well as outside adjacent spaces(3AII-1/2). For micromorphological analysis, a block of sediment

Table 1List of sediment samples examined for phytoliths, faecal spherulite, and micromorphology.

Sample locations Contextnumber

Sampled deposits(thickness)

Samplenumber

SoilpHa

Descriptions of the samples

Level 4 inSquare 3AII

Inside the bins 3AII-1 Middle fill (20e25 cm) G€oy10-S29 8.4 Clayey sediments with abundant mineral and rock clasts in additionto fragments of charcoal, bones, and mudbricks (up to 5 cm).

Bottom fill (10e15 cm) G€oy10-S30 8.3 Speckled groundmass fabric with dense clay groundmass andabundant large mineral and rock clasts, charcoal (up to 1 cm),amorphous organic matter, bone, shell, and other anthropogenicdebris.

Bottom wall G€oy10-S31 NA The bottom wall of the bin. Dense clay.Bottom fill and wall G€oy10-S36 NA A block of sediment, including bottom fill and a bottom part of the

installation, was sampled for micromorphology.3AII-2 Middle fill (20e25 cm) G€oy10-S32 8.1 Clayey sediments with abundant mineral and rock clasts, large

fragments of charcoal (up to 1 cm) and mudbricks (up to 10 cm).Bottom fill (10e15 cm) G€oy10-S33 8.1 Similar to the middle fill with less inclusion of mudbricks.Bottom wall G€oy10-S34 NA The bottom wall of the installation. Dense clay.

Outside the bins 3AII-1/2 Occupational surfacenear bins

G€oy10-S35 8.3 Sediments were sampled from an area between the two bins.

Level 10 inSquare 4BII

Inside the bin 4BIIX-94 Middle fill (20e25 cm) G€oy10-S21 8.2 Loose, silty clay sediments including a few charcoal fragmentsBottom fill (4 cm) G€oy10-S22 8.3 White fibrous, loose deposits, on top of which a handstone and a

grinder were found. Highly porous microstructure consisting ofphytoliths with 5e10% density of very small charcoal fragments,very fine amorphous organic matter, and even higher densities (10e20%) of fine carbonate and gypsum crystals and intergrowths.

Bottom wall G€oy10-S26 NA The bottom clay wall of the installation. White fibrous materials ofthe bottom fill are attached on the interior surface.

Bottom fill and wall G€oy10-S25 NA A block of sediment, including bottom fill and a bottom part of theinstallation, was sampled for micromorphology.

Outside the bin 4BIIX-92 Occupational surfacenear bins

G€oy10-S23 7.6 Sediments were sampled from an area with a charcoalconcentration associated with burnt stones, burnt sediments, andash.

4BIIX-92 Occupational surfacenear bins

G€oy10-S24 8.3 Sediments were sampled from a small area (ca. 30e40 cm) withwhite deposits.

Ground stonesin the bin

4BIIX-94a Bifacial handstone G€oy10-S38 NA Sediments were brushed off a working surface.G€oy10-S40 NA Sediments were washed off a working surface after sampling

G€oy10-S38.G€oy10-S41 NA Sediments were washed off the other working surface.

4BIIX-94b Unifacial grinder G€oy10-S42 NA Sediments were brushed off a working surface.G€oy10-S43 NA Sediments were washed off a working surface after sampling

G€oy10-S42.

a Soil pH (water ratio of 1:2.5) was measured using the glass electrode method by Palynosurvey Co.


(15�12� 10 cm) containingmaterial from the bottommost fill, thebin wall/fill interface, and the bin wall was collected from a bin in3AII-1.

2.2.2. Level 10 in Square 4BIIThe excavation of Level 10 in Square 4BII by the Japanese

mission found eleven bin features, two of which are located inside around building (4BIIX-16), while the rest are distributed in outdoorareas (Fig. 5). Among the latter, eight bins are clustered in a space

Fig. 4. (A) A cluster of bin features in an outdoor area (Level 4 in Square 3AII) indicating th(3AII-1 and 2), indicating sampling locations (B for phytoliths and faecal spherulites, and

that could have been a courtyard given the layout of two roundbuildings (4BIIX-12 and 16), their annex walls, and the location of adoorway of a building at 4BIIX-16. The deposits in the probablecourtyard are generally ashy and associated with a high concen-tration of charcoal fragments and burnt cobbles on a reddish,hardened (burnt?) surface (4BIIX-92). This area is also character-ized by the recovery of complete and nearly complete, large arti-facts (such as a grinder, a grinding slab, an abrader, and boneartifacts), located near building walls or bins, indicating that they

e contexts sampled for sediment analyses. (B and C) Cross-sections of fills in two bins, for micromorphology). See Table 1 for sample numbers.


were left as de facto or provisional refuse of activities performed inthis area.

We tookmore than 100 sediment samples fromvarious contextsin this level, including building floors, outdoor occupational sur-faces, and inside the bins. Here we report the analysis of the sam-ples from the inside and outside a bin at 4BIIX-94 (See Table 1 andFigs. 5 and 6 for the list and locations of the samples). Although thisfeature is structurally similar to other bins in this level, it issomewhat distinct from the others for two reasons. First, twocomplete upper grinding stones (4BIIX-94a and 94b) were placednear the base of the bin. Second, the sediments at the very base ofthe bin, approximately 4 cm in thickness, between the binwalls andthe overlying grinding stones, exhibited a white, fibrous appear-ance, rich in white “light chaffs” and phytoliths (see below). Ac-cording to these observations, the fill deposits were subdivided intofour layers (Upper, Middle, Lower, and Bottom: Fig. 6). The bottomfill corresponds to the white fibrous deposits at the bin base, whilethe lower fill is immediately above the bottom fill.

Sediment samples for macrobotanical remains were taken fromall four layers of the fills. For the phytolith and faecal spheruliteanalyses, we sampled middle and bottom fills, the bottom wall ofthe bin, working surfaces of the grinding stones left in the bin, aswell as an adjacent activity area. The feature was also sampled formicromorphological analysis; a column of sediment approximately14 � 11 � 12 cm in size was left unexcavated at the base of the binand then taken en bloc to be prepared into thin section slides.

3. Phytolith and spherulite analyses

3.1. Methods

3.1.1. Phytolith analysesPhytolith analyses followed the methods of Albert et al. (1999).

Samples of approximately 1 g of dried sediment were treated with3 N HCl, 3 N HNO3 and H2O2. Phytoliths were concentrated using2.4 g/ml sodium polytungstate [Na6(H2W12O40)�H2O]. Slides wereprepared by weighing about 1 mg of sample using Entellan New(Merck). A minimum of 200 phytoliths with diagnostic morphol-ogies were counted at 400� magnification. Morphological identi-ficationwas based on standard literature (Twiss et al., 1969; Brown,

Fig. 5. A clay bin (4BIIX-94) and its adjacent area

1984; Mulholland and Rapp, 1992; Rosen, 1992; Twiss, 1992;Madella et al., 2005; Piperno, 2006), as well as on modern plantreference collections (Albert and Weiner, 2001; Tsartsidou et al.,2007; Albert et al., 2014; Portillo et al., 2014).

3.1.2. Spherulite analysesSamples were prepared following Canti's (1999) methodology.

Approximately 1 mg of dried sediment was mounted on a micro-scope slide, as described above for phytoliths. Spherulite countingwas performed using a polarized light microscope at 400�magnification. Samples were compared to modern dung referencecollections (Albert et al., 2008; Portillo et al., 2012, 2014).

3.2. Results of phytolith and spherulite analyses

Table 2 shows the quantitative phytolith and spherulite results,percentage of acid insoluble fraction (AIF), estimated amounts ofphytoliths per gram of AIF and per gram of sediment, percentages ofgrass phytoliths, phytoliths from grass inflorescences, percentagesof weathered morphotypes (WM) and multicelled phytoliths (MC),and numbers of spherulites per gram of sediment. The insolublefraction data (% AIF), which is the fraction that remains after acidand peroxide treatment, indicates the presence of siliceousmaterialby percentage that include heavy minerals, quartz, clay and phy-toliths. The AIF percentage in the samples ranged from 53 to 78%(Table 2). This means that siliceous minerals were major compo-nents of these sediments, although variations exist among thesamples. Sediments obtained from bin fills (S21, S30, and S32)showed the lowest AIF percentages (53e59%), whereas in mostsamples the AIF was 65%, or more.

Phytoliths and spherulites were noted in different amounts inthe samples. Indications of partial dissolution of phytoliths wereobserved in all samples by the presence of surface pitting andetching at different degrees. Those phytoliths that were unidenti-fiable because of some degree of dissolution were counted andpercentages of the total phytolith count were listed as weatheredmorphotypes (Table 2). The dissolution index ranged from 1.9 to8.9% despite moderately alkaline soil conditions (Table 1), and thus,did not interfere in the overall morphological identification of thephytoliths.

(4BIIX-92) distributed with domestic refuse.

Fig. 6. (A) Plan and section of a bin (4BIIX-94), showing sampling locations (B for phytoliths and dung spherulites, and , for micromorphology). See Table 1 for sample numbers.(B) Two in situ grinding stone found at the bottom of the bin overlying a whitish, fibrous sediment at the base of the bin. (C) A column of whitish sediments sampled formicromorphology analysis at the base of the bin.


Other siliceous biogenic microremains, primarily diatoms, werealso observed in most of the bin fill sediments and their bottomwalls. In contrast, samples collected from outside the bins did notyield these silica microfossils. Diatoms can grow in almost anyenvironmental condition where moisture is present (i.e., soils, de-posits, mud-bricks and plasters) (Coil et al., 2003). The results ob-tained from this study have been analyzed separately according tothe different excavation areas (Squares 3AII and 4BII).

3.2.1. Level 4 in Square 3AIISeven samples were analyzed from Level 4 in Square 3AII

(Table 2). Most of the samples correspond to the fill of two bins(3AII-1 and 2). Two of these samples were collected in the bottomwall of each feature. An additional sample belongs to sedimentsfrom outside the bins (S35).

Phytoliths were abundantly identified in the examined assem-blages with concentrations ranging from 500,000 to 1.6 millionphytoliths per gram of sediment (Table 2). Both bins showed thelargest abundances in their middle fills. Grasses dominated thephytolith record, comprising around 85% or more of all the countedmorphotypes. Inflorescences constitute between 31 and 43% of allthemorphotypes. According to the short cell morphologies, that arecommonly produced both in leaves and inflorescences, most ofthese grasses belonged to the C3 Pooid subfamily (Fig. 7a). Inaddition, other short cells from C4 Panicoids were also identifiedbut in lesser amounts (Fig. 7b). Inflorescences were characterized

mainly by diagnostic epidermal elongate echinate long cells(Fig. 7c). Other distinctive morphotypes, such as dendritic andpapillae cells were also identified, although in lower proportions(Fig. 7dee). Although sampled sediments are moderately alkaline(Table 1), such decorated morphologies are considered good in-dicators of preservation in phytolith assemblages (Cabanes et al.,2011). Multi-cellular structures (multi-celled or interconnectedphytoliths) from both floral parts of cereals, and the leaves and thestems of grasses (Fig. 7f) were also noted in most of the samples indifferent proportions (Table 2). The identified morphologies cor-responded to the floral parts or husks of grass seeds, primarily fromwheat (Triticum sp.) and barley (Hordeum sp., Fig. 7g).

Dung spherulites were noted in most of the bin samples indifferent amounts (Table 2 and Fig. 7h) with the exception ofsamples from the bottomwalls of the bins. The middle fills yieldedthe largest numbers (over 100,000 spherulites/g sediment) thatcorrelate with large phytolith concentrations in these samplesindicating that dung material was dumped in that area.

3.2.2. Level 10 in Square 4BIITen samples were selected from level 10 in Square 4BIIX

(Table 2). Three of these samples correspond to fill deposits from abin (4BIIX-94), including the white colored sediments observed atthe base of the bin. Additionally, five samples were obtained fromthe working surfaces of two grinding stones that were placed nearthe base of the bin. Two sediment samples were obtained from

Table

2Mainphy

tolithan

dsp

heruliteresu

ltsob

tained

from

Squares

3AIIan

d4B

II.

Sample

location

sCon

text

numbe

rSa

mpleddep

osits

Sample

numbe

r%AIF

N.P

hytoliths

%Grass

phytoliths

%Inflorescence

phytoliths

%W

Mphytoliths

%MC

phytoliths

N.S

pherulites

1gof

AIF

1gof

sedim

ent

1gof

sedim

ent

Leve

l4in

Square

3AII

Insidethebins

3AII-1

Middle

fill

G€ oy

10-S29

67.5

2,20

0,00

01,20

0,00

086

.731

6.6

18.4

159,00

0Bottom

fill

G€ oy

10-S30

56.1

1,20

0,00

062

0,00

086

.331

.75

8.9

38,000

Bottom

wall

G€ oy

10-S31

64.5

820,00

049

0,00

088

.136

.64.8

20

3AII-2

Middle

fill

G€ oy

10-S32

58.9

3,00

0,00

01,60

0,00

088

.137

.56.9

8.5

101,00

0Bottom

fill

G€ oy

10-S33

68.6

2,50

0,00

01,50

0,00

086

37.8

8.2

13.5

61,000

Bottom

wall

G€ oy

10-S34

722,10

0,00

01,30

0,00

085

.843

.38.9

2.1

0Outsidethebins

3AII-1/2

Occupational

surfacenea

rbins

G€ oy

10-S35

67.2

800,00

051

0,00

085

.842

.47.4

6.9

51,000

Leve

l10in

Square

4BII

Insidethebin

4BIIX-94

Middle

fill

G€ oy

10-S21

52.8

1,10

0,00

053

0,00

087

.854

.75

515

,000

Bottom

fill

G€ oy

10-S22

68.6

55,000

,000

33,900

,000

96.3

74.4

2.4

5.8

0Bottom

wall

G€ oy

10-S26

72.8

19,000

,000

13,100

,000

94.2

78.3

1.9

1.3

0Outsidethebin

4BIIX-92

Occupational

surfacenea

rbins

G€ oy

10-S23

70.1

980,00

047

0,00

085

.424

.22.5

9.8

0

Occupational

surfacenea

rbins

G€ oy

10-S24

61.2

3,50

0,00

01,70

0,00

083

.538

.17.7

3.2

7,40

0,00

0

Groundston

esin

thebin

4BIIX-94a

Bifacialhan

dston

eG€ oy

10-S38

76.8

4,60

0,00

03,10

0,00

093

59.8

3.6

10.6

0G€ oy

10-S40

68.6

2,50

0,00

081

0,00

091

.650

.73.7

3.7

0G€ oy

10-S41

67.2

2,40

0,00

01,30

0,00

087

.253

.98.6

0.5

04B

IIX-94b

Unifacialgrinder

G€ oy

10-S42

70.4

2,90

0,00

01,60

0,00

093

.360

.73.6

5.8

0G€ oy

10-S43

66.1

5,50

0,00

03,50

0,00

092

.958

.54

2.4

0


outside the bins. Phytoliths were especially abundant in all thesamples (from 470,000 to 34 million phytoliths/g of sediment,Table 2). Contrary to level 4 in Square 3AII, the sediments richest inphytoliths were the white deposits from the bottom fill of the bin(S22 and S26, over 13 million phytoliths/g of sediment). The whitesediments (S24) from the area adjacent to the bin yielded also ahigh amount of these micro-remains (around 1.7 million phyto-liths/g of sediment).

Grasses constitute more than 84% of all the morphotypes in allthe examined samples. Again, C3 Pooid grasses (Fig. 7a) were themost common group identified. Interestingly, grass inflorescencesdominated in all the bin and grinding stones samples, whereasleaves and stems of these plants were common in sediments fromthe outside area. Similar to Square 3AII samples, epidermal elon-gate echinate cells were abundantly identified in all the assem-blages. The phytolith-rich white sediments from the bin fills (S22and S26) yielded the highest concentrations of inflorescent phy-toliths with around 75e78%, and a low proportion of anatomicallyconnected phytoliths.

Significantly, overwhelmingly high densities of spherulites wereobserved in the white deposits in the adjacent area (S24 from4BIIX-92) (over 7 million spherulites/g of sediment, Table 2)indicative of dung accumulation. This concentration overlaps withhigh proportions of grass phytoliths. Inflorescences were alsoobserved in this sample, although in much smaller proportions(38%). These findings indicate that these sediments were composedof animal dung derived from a grass-rich diet. This area wasdescribed in the field as white-colored sediment associated withcobbles, bone tools, faunal remains, charcoal and ash (Fig. 5). Thestudied sample (S24) may thus represent the ashy residue of dung.

4. Macro-botanical remains

4.1. Methods

The examination of macro-botanical remains from the clay binswas based on sediment samples for flotation from both inside andoutside the bins. The fill deposits of the bins were divided into two(Upper and Bottom) to four layers (Upper, Middle, Lower, andBottom); from each layer we sampled (i.e., conducted water-flotation of) ca. 2e6 L of sediment. We also sampled depositsfrom apparent outdoor surfaces near the bins. Flotation was con-ducted with 0.3 mm mesh sieve, and retrieved light fractions weredried and brought to Japan for identification under a binocularmicroscope. The identification is based on modern botanicalreference collections made by our field surveys around the site.Here, wemainly present the results of the sorted botanical samplesfrom inside the bins (3AII-2 and 4BIIX-94) and their adjacentoutside areas.

4.2. Results

We have analyzed 19macro-botanical samples (60.7 L of soil), ofwhich seven samples (Table 3) are directly relevant to this study. Ingeneral, barley grains and rachises (Hordeum vulgare) occur mostfrequently (18% of the charred remains), followed by free-threshingwheat (Triticum durum/aestivum) (6%) among the charred botanicalremains. The former show morphological features of the nakedtype (i.e., plump and round in cross-section). The presence of nakedcereals is also indicated by the recovery of a large volume of fragileparts (“light chaffs”) like glume tips and awn in addition to rachisesor spikelet bases. Notably, rachises and chaffs outnumber grainsand straw fragments in all the samples. Other crop plants includehulled wheat and lentils although these are minor components incomparison with the above-mentioned cereals.

Fig. 7. Photomicrographs of phytoliths and other microremains identified in the samples. The photographs have been taken at 400� (aeg: PPL, h: XPL). a) short cell rondel, b) shortcell bilobate, c) long cell with echinate margin, d) dendriform long cell, e) papillae cells, f) multicellular structure from grass leaves/stems, g) multicellular structure from Hordeumsp. husk, h) dung spherulites.


Common wild species in our samples include Asteraceae, Bro-mus, Silene and Chenopodium-type. Seeds of the Asteraceae are verysmall (ca. 1 mm) and shaped in an asymmetrical oval, showingclose similarity to those of Artemisia. This identification is alsosupported by the observation of vertical striations, a distinctfeature of Artemisia seeds, on some of the well-preserved Aster-aceae seeds (van Zeist and Bakker-Heeres, 1985).

There are no distinct differences in the composition of charredbotanical remains among the samples from different parts of thebin fills or outside the bins (Table 3). However, the lower andbottom fills of the bin at 4BIIX-94 are particularly similar to eachother in the range of taxa, including grains and rachises of nakedbarley and wheat, wild or weedy species like Artemisia-type, Poa-ceae (Bromus, Aegilops, Lolium), Silene and Brassicaceae seeds, andnumerous “white light chaffs”. The light chaffs include fragments ofglume, lemma/palea and awn, and some of them were found inwhite, mineralized appearance instead of being carbonized (Fig. 8).These “white light chaffs” occur most frequently in whitish, fibrousdeposits at the bin base. As described below in the micromor-phology section, the white appearance of fibrous remains is createdby gypsum (and rarely calcite) crystals and crystal intergrowthsformed around phytoliths.

5. Micromorphological examination

5.1. Methods

The primary aim of micromorphological examination of sedi-ment samples from two clay bins (S36 from 3AII-1 and S25 from4BIIX-94) was to understand the nature of the fills within thesebins, and to compare formation processes between the two binfeatures that differed macroscopically during excavation. Morespecifically, we aim to explore the depositional processes that led tothe formation of a whitish, fibrous deposit at the base of the 4BIIX-94 bin.

During the excavation of the two clay bins, an intact block ofsediment (approximately 15 � 10 � 10 cm) from the inside of eachbin was left undisturbed. Each block was described, oriented,photographed and collected en bloc for micromorphological pro-cessing and analysis. Theywere transported to Nichika Geo-ScienceMaterials Inc. in Kyoto, Japan. In the thin section laboratory, eachintact sediment block was thoroughly dried and then impregnatedwith a clear polyester resin under a vacuum. Once consolidated,each block was prepared into five thin section slides, each 3 � 5 cmin size, representing the bottom fill sediments contained within the

bin, a portion of the clay bin wall and, importantly, the boundarybetween these two deposits. The microscopic observations of thesethin section slides are systematically described in SupplementaryTable.

5.2. Results

5.2.1. Level 10 in Square 4BIIIn general, the samples of bin walls (both S25 and S36) are

dominated by a massive, dense clay groundmass and those of bin-fills by phytoliths and gypsum crystal intergrowths. There is a cleardifference in composition and character between samples takenfrom the bottom fill of bin 4BIIX-94 (Slides 1e3 of S25) versus thosethat represent the wall/base of the bin (Slides 4e5 of S25) (Fig. 9).The bin fill samples exhibit a spongy, porous, sometimes laminatedmicrostructure and are a notably gray color in Plane Polarized Light(PPL). They consist almost entirely of phytoliths, particularly elon-gate phytoliths oriented parallel to each other and to the sampledbin wall/base creating, in places, a layered, laminated appearance,with 5e10% density of very small charcoal fragments, very fineamorphous organic matter, and even higher densities (10e20%) offine carbonate and gypsum crystals and intergrowths (Fig. 10:AeD).

The bottom fill is almost entirely dominated by grass phytoliths(predominantly inflorescences, and lesser amounts of stems andleaves; see phytolith section above for details). The fill is quiteporous, a distinct gray/whitish color, 10e15% density of very finecharcoal fragments (<0.05 mm), and laminated following thecontours of the underlying bin shape. In particular, Slide 3 shows avery distinct and well-marked boundary between the fill and binwall/base, with both a color and composition change, and a thin(~3 mm-thick) layer of dark gray, very porous, crumb clay aggre-gates and individual gypsum crystals. The whitish/grey color seenmacroscopically during excavation is also documented in the thinsection. It results fromvery fine crystals and crystal intergrowths ofgypsum (and, rarely, calcite) that form a fine-textured and sparsegroundmass around the phytoliths and charcoal fragments (Fig. 10:EeF).

Well-formed gypsum crystals and crystal intergrowths form insitu as the mineral precipitates out from solution in water presentin (or moving through) a deposit (Brewer, 1976; Fitzpatrick, 1993;Stoops et al., 2010). They can form from infiltration of moisturethrough a deposit or as it pools in particular portions of a deposit,such as the undersides of stones (in this case two handstones) orwhen reaching a comparably impermeable layer such as the dense


clay bin wall/base. In this case, the laminated phytolith layersconsisting of grass parts (seeds, leaves and stems) in the bottom fill,some of which suggest a possible matted texture (Fig. 11: AeD),could represent the remains of a very thin lining on the bottom of

Table 3Counts of macrobotanical remains from different parts of the bin fills and from outside t

Context number 4BIIX-94 4BIIX-94 4B

Deposits Bottom fill Lower fill MSoil amount (liter) 6 6 6Charcoal amount (ml) 42 27 7.“White light chaffs”“White light chaffs” (count) 11,020a 1071 6Ratio of “white light chaffs” to charred remains 8.55 0.38 0.Charred remainsCereal grainsBarley 44 211 11Barley (hulled type) 0 0 0Wheat (naked type) 20 22 9Wheat (hulled type) 3 1 0Fragments 70 218 14Sub-total 137 452 26% of total 10.6% 16.2% 20Rachises & chaffsBarley rachis 130 281 16Naked wheat spikelet bases 60 78 97Hulled wheat spikelet bases 37 24 23Rachis (barley/wheat) 95 125 92Light chaff 162 824 48Aegilops spikelet bases 2 7 3Straw fragments 2 5 2Grass rachis 66 45 6Sub-total 554 1389 87% of total 43.0% 49.6% 66Other food plantsLens 5 0 0Legumes 2 5 0Vitis 0 1 0Crataegus 0 1 0Sub-total 7 7 0% of total 0.5% 0.3% 0.Wild plantsAegilops 2 4 1Bromus 28 31 24Lolium 8 7 8Poa type 12 5 1Stipa 3 18 4Panicaceae 2 0 1Poaceae 32 137 22Artemisia type 139 520 3Heliotropium 2 10 2Boraginaceae 24 23 6Alyssum/Lepidium 6 0 0Brassicaceae 32 48 3Silene 27 25 15Gypsophilla type 0 0 0Caryophillaceae 2 0 0Hippocrepis type 3 0 0Trifoliae 6 0 4Medicago (pod) 0 0 0Fabaceae 0 0 0Suaeda 3 11 5Chenopodiaceae 56 10 7Cistaceae 7 0 0Erodium 3 0 0Lamiaceae 4 0 1Malva 7 1 0Adonis 1 3 0Verbena 1 9 1Other wild species 4 7 0Indetermine 177 81 62Sub-total 591 950 17% of total 45.8% 34.0% 13Total counts of charred remains in each sample 1289 2798 13

a This number was estimated from the actual count of white light chaffs in the one-fi

the bin (Goldberg, 2000; Goldberg et al., 2009; Wadley et al., 2012).The gypsum likely formed as moisture introduced into the plantmatter contained in the bins percolated down to the lowest layers,and into the lining, where gypsum precipitated here to form small

he bins.

IIX-94 4BIIX-94 4BIIX-92 3AII-2 3AII-1/2

iddle fill Upper fill Outside the bin Bottom fill Outside the bin5 4 2 2

6 6.1 12 21 17.5

5 50 197 300 0.00 0.05 0.21 0.00

4 77 41 82 590 3 2 06 12 10 00 1 2 0

1 116 45 83 1304 199 102 179 189.2% 14.2% 10.6% 19.0% 22.7%

6 167 91 127 6772 33 64 1223 8 7 089 74 67 48

7 610 425 305 1960 7 7 33 1 6 024 8 19 21

6 988 647 602 347.9% 70.6% 67.0% 63.8% 41.7%

2 6 7 00 0 0 10 0 0 00 0 0 02 6 7 1

0% 0.1% 0.6% 0.7% 0.1%

1 0 3 212 13 11 57 7 3 18 5 1 00 4 6 00 0 1 029 8 20 2015 10 38 1062 2 10 981 4 20 80 0 0 06 5 11 1721 7 7 612 0 0 00 0 0 00 0 0 13 9 1 90 0 0 80 0 0 00 1 1 035 58 5 10 2 0 01 0 0 05 6 0 00 1 0 01 0 4 00 1 0 41 1 2 051 67 11 10

0 211 211 155 296.0% 15.1% 21.8% 16.4% 35.5%10 1400 966 943 833

fth of the light fraction because of their large quantity.

Fig. 8. Charred or mineralized chaffs from the base of the bin (4BIIX-94 in Level 10).


crystals and intergrowths at the bin wall boundary that was morecompact and impermeable. In essence, the clay wall formed a moreimpermeable boundary that the water did not infiltrate as easily asthe fill. Instead after water pooled at the base of the bin as it dried itprecipitated gypsum and some carbonate from solution, formingthe fine dust-like layers of gypsum and carbonate that give it adistinctive whitish, fibrous macroscopic appearance. The clay wallsof the bin form a comparatively water-resistant boundary thatconcentrated any moisture percolating down through the bin andits contents by trapping it in the lining and bottom fill, wheregypsum (mobilized from the local geology and adjacent archaeo-logical deposits) precipitated out of solution here to form fibrouscrystals and crystal intergrowths in the pores between the grassmaterial and undersides of the ground stones.

The walls/base of the bin (Slides 4e5 in Fig. 9), on the otherhand, are massive in appearance, with varying crumb, granular,and vughy or vesicular microstructures (termed ‘complex’ inStoops et al., 2010). The groundmass is dense and clay-rich, withvoids (vughs, channels and vesicles) indicative of decayed plantmatter (Fig. 13: AeB). The samples are a reddish-brown color in PPLand, as a clay-enriched deposit, exhibit a mosaic b-fabric in Cross

Fig. 9. Excavated sediment column from S25 removed en bloc (right) with five thin sectionslides 4e5 consist of the bin wall material.

Polarized Light (XPL) caused by very fine gypsum and calcitecrystals. These dense clay samples also contain 10% or more smallcalcitic (micritic) limestone fragments, gypsum crystals and in-tergrowths, charcoal, burnt bone, and very rare clasts of quartz orobsidianmicroflakes (Fig. 13: AeB). These samples also contain rarespherulites from animal dung. Phytoliths are also present in thesesamples, but in much lower densities (<5%) than the overlying filldeposit. Instead, a high biological content is represented by voidspaces resulting from decay of biological content in the deposit.

These samples are characteristic of prepared clay constructionmaterial or mudbrick, with their groundmass density, high claycontent, gypsum crystal intergrowths and evidence for somedissolution features (Friesem et al., 2011, 2014). This is not sur-prising as erosion by water is the most common post-depositionalalteration affecting clay construction material (Rosen, 1986;Goodman-Elgar, 2008). Although the large planes and channelsshow a parallel orientation from molding and forming of the clay(and likely also from inclusions of grass stems and leaves astemper), the samples from the clay bin walls (slides 4e5) display adense, massive microstructure with no laminations or micro-stratigraphy and, thus, no evidence for regular re-surfacing of thebin walls. Although there are small fragments of the reddish-colored clay bin deposits contained within the fill (0.5e1 mmdiameter; Fig. 10: BeC), the fill and binwall samples are really quitedistinct from each other in their content and structure and, un-surprisingly, represent very different depositional events anddiagenesis. The samples from S36 are much less clear (see below).

5.2.2. Level 4 in Square 3AIIA sediment block (S36) sampled from bin 3AII-1 covers deposits

of the bottom fill and the bin wall (Fig. 12). However, the boundarybetween the bottomwall of the bin and the bottom fill is unclear asthere was admixture between the fill and bin lining noted duringexcavation of the feature, and so the sample may contain only a few(1e2) cm of the bottom fill (slide 1) and consist otherwise of eitherthe bottom bin wall or a secondary fill that is identical to the binwalls (slides 2e5).

The five slides from bin 3AII -1 are quite homogenous and, thus,different from the fill samples from S25 of bin 4BIIX-94. The bottomfill and bin wall samples from S25 are distinct from each other in

slides cut from the block (left). Slides 1e3 (from top) represent the bottom fill, while

Fig. 10. Photomicrograph images of features from the bottom fill (S25) of the bin (4BIIX-94 in Level 10). (A) Image of the articulated and laminated phytolith layers in the bottom fill(PPL). (B) Image of the bottom fill showing phytolith layers perpendicular to each other (PPL). (C) Bottom fill layer with randomly oriented phytoliths, fine charcoal, and smallfragments of clay aggregates in a fine groundmass (PPL). (D) Large fragment of charcoal incorporated into phytolith-rich bottom fill (PPL). (E) Fine calcite crystals and crystalintergrowths within the groundmass of the phytolith-rich bottom fill (XPL). (F) Gypsum crystals and crystal intergrowths within the fine, dispersed groundmass of bottom fill (XPL).


terms of both fine- and coarse-fraction components and structure.All of the slides from bin 3AII-1, representing the bottom fill (upperportion of slide 1) and binwall/base (slides 2e5), are comparativelyuniform in composition, consisting of both remnants of the binsfinal contents (grass phytoliths), other debris (charcoal, burnt bone,shell, limestone/calcite clasts), and fragments of clay bin debris(Fig. 13: CeF).

There are no distinct differences between any of the slides interms of lamination, other fine- and coarse-fraction content, ormineralogy. Unlike S25, S36, especially slides 1 and 5, contain verylarge fragments (up to 1 cm) of charcoal and ash. The presence ofvery small (<0.5 mm) fragments of charcoal like those from S25 isexpected within stored processed grains and associated linings,where charcoal flying out of a hearth could easily become incor-porated into foodstuffs or textiles being processed or dried nearby.Instead, these large fragments of charcoal (up to 1 cm) are morecommonly found in secondary fill deposits that contain theheterogenous debris from household activities, such as hearthcleaning or floor sweeping (Matthews, 2010; Matthews et al.,1997; Stoops et al., 2010; Shillito et al., 2011; Matthews, 2012a,

2012b). These large fragments of charcoal are randomly orientedand distributed. The accompanying ash reinforces this source forthe charcoal.

The bin wall samples (slides 2e4) also contain higher densitiesof rock fragments (limestone, obsidian or quartz flakes) and soil/clay-like aggregates (silty and sandy clays with amorphousorganic matter, fine charcoal, a coarse fraction of calcite, gypsum,and quartz) (Fig. 13: CeF) (Finlayson et al., 2003; Kuijt andFinlayson, 2009; Stoops et al., 2010). In essence, all the S36 sam-ples, including the upper portion of slide 1, more closely resemblethe bin wall samples from both S25 than the distinctive bottom fillfrom S25 (slides 1e2). In addition, with the exception of increasedlarge charcoal and mineral constituents in slides 1 and 5, the S36slides are very consistent with each other in terms of thearrangement of components and groundmass (very fine fraction).There are no distinct boundaries between fill and bin wall (as seenin S25), since the slides show no internal structure (massive, ratherthan laminated), and their composition (fine and coarse matter andmineralogy) is much more internally variable than the bottom fillfrom S25.

Fig. 11. Photomicrograph images of the bottom fill (S25), dominated by phytoliths and showing a laminated, in some cases, matted appearance. (AeB) Images of the laminated andlayered, usually perpendicularly, phytoliths within the bottom fill (PPL), along with inclusions of weathered charcoal fragments, amorphous organic matter and soil aggregates.(CeD) Close-up (100�) of laminated and perpendicularly layered phytoliths within the bottom fill of the bin (PPL).


6. Discussion

6.1. Depositional processes of the bin features (3AII-1 and 2) inlevel 4

Phytolith results indicate a vegetal component dominated by amixing of inflorescence and leaves/stems in the middle and bottomfills of the two bins, as well as in the sediments from the outsidearea. Dung micro-remains were observed in the samples from both

Fig. 12. Excavated sediment column from S36 removed en bloc (right) with five thin section2e5 come from the bin wall material.

inside and outside the bins. These similarities between the insideand outside the bins in the vegetal component and the occurrenceof dung spherulites indicate that these micro-remains inside thebins may represent material accumulation resulting from house-hold debris or the remains of dumped material derived from do-mestic activities. These findings could represent either livestockdung remains or faecal material burnt as fuel dumped at the lo-calities studied here. The latter possibility is consistent with theidentification of a large amount of ash and household debris in the

slides cut from the block (left). Slide 1 (from top) represents the bottom fill, while slides

Fig. 13. Photomicrograph images of features from the clay bin of S25 and bottom fill and clay wall of bin S36. (AeB) The clay bin wall consists of a dense clay groundmass with fineinclusions of charcoal (black), quartz silt (B-white grains), gypsum crystals (B-grey mineral in center of image), and bone fragments (A-orange inclusions in center of image). Bothimages contain elongate and circular voids and vughs (A-white voids, B-black voids) characteristic of decayed plant matter that likely represented straw added as temper during binconstruction. Secondary features in the clay bin wall deposit include localized areas of iron staining and clay translocation (darkened areas in reddish-brown groundmass) (A: PPL,B: XPL). (CeD) Large fragments of charcoal within a crumb, clay-rich aggregate in the bottom fill (PPL). (CeD) Large fragments of charcoal and amorphous organic matter (smallblack specks) within a calcite- and gypsum-rich clay groundmass of the bin wall sample (C: PPL, D: XPL). (E) Soil aggregate (top center of image) contained within the dense claygroundmass of the bin wall samples (PPL). (F) Clay-rich groundmass with fine charcoal, ash, amorphous organic matter, gypsum and quartz crystals, clay aggregates, abundantvughy and planar voids, and localized areas of calcite dissolution (darker brown areas) and precipitation (light, speckled areas) are characteristic of the bin wall samples (PPL).


micromorphological samples from the bottom fill of bin 3AII-1(S36).

Micromorphological observations of the sediment block (S36)from the bin-bottom (3AII-1) suggest that although it is high inphytolith content (in particular grass leaves and stems, versusinflorescences in S25), the fill (slide 1) and wall samples (slides2e5) of S36 contain notably less phytoliths than the S25 fill(4BIIX-94) and are more similar to the bin wall samples from S25.The S36 sample also includes other debris (charcoal, burnt bone,shell, limestone/calcite clasts) and fragments of clay bin debris.Both phytoliths and void spaces present in the bin wall samplesresult from the use of chaff temper in the construction of mud-brick that leave characteristic phytolith and void traces afterdecay (Goldberg, 1980; Courty et al., 1989; Goodman-Elgar, 2008;Love, 2012). Other dissolution features are typical fabric changesfrom mudbrick construction and decay (Berna et al., 2007;

Goodman-Elgar, 2008). Thus, the sediment block S36 likely rep-resents a secondary fill composed of an admixture of storedcontents and debris from domestic activities and clay bin con-struction or repair.

A secondary fill interpretation is also consistent with theobservation that although only a portion of slide 1 was thought torepresent a sample of the bin fill (and was identified macroscopi-cally as a greyish sediment), it and slide 5 (reported as bin wallmaterial) resemble each other, while the other slides (2e4) arehomogenous, dense clay (Fig. 13). Both macroscopically andmicroscopically (Figs. 12 and 13), it appears that slides 1 and 5might represent secondary dumping events that filled the bin 3AII-1 with large fragments of charcoal, and more ash, lending a darker,more heterogenous appearance. On the other hand, slides 2e4 arevery homogenous, with a compact groundmass and brown color.The latter slides probably represent fill material consisting of debris


of the same material as the bin wall (mudbrick) dumped into thebin to fill it.

If one can assume a similar original function to that of bin 4BIIX-94 (S25) based on contextual similarities, perhaps after use as astorage bin, this clay bin 3AII-1 was likely re-used for refusedisposal. It may also have been filled with similar material to itsconstruction used to level it off with adjacent features. While thesesamples are quite homogenous compared to S25, there is noobvious sampling reason why the topmost and bottommost slides(1 and 5) would resemble each other while the middle ones (2e4)would be homogenous; contamination during sampling may haveobscured the interpretive value of these slides. The most parsi-monious explanation is that similarities in the macroscopicappearance of the sediment column (from the bin with the thinsection slides cut from it) suggest a secondary nature for the fillwithin this bin, where slides 1 and 5 represent slightly differenttypes of fill (household refuse dumping or hearth cleaning) thanslides 2e4 (construction material). Although these depositionalprocesses are not indicative of primary bin usage, they are consis-tent with the fact that the clay bins are located near open spaceswhere household activities may have taken place.

6.2. Depositional processes of the bin feature (4BIIX-94) in level 10

Our primary question related to this feature is what depositionalprocesses are relevant to the formation of the whitish, fibrous fill atthe bottom of the bin. The phytolith analyses indicate that thewhitedeposits consist of concentrations of grass phytoliths with highproportions of inflorescences in comparison with its upper fill de-posits and areas outside of the bin. The white deposits are distinctfrom upper fills or sediments outside bins also by the absence ofdung spherulite. These observations are corroborated by themicromorphological analyses that note the high densities of inflo-rescence phytoliths with few very small charcoal fragments and fineorganic matter. In the soil thin-section, this bottom fill shows anabrupt transition to the massive clay wall of the bin. The micro-morphological study also suggests that the whitish, fibrousappearance was formed by the growth of fine gypsum and calcitecrystals and crystal intergrowths precipitated around phytolithsfromwater/moisture as it settled at the comparatively impermeabledense clay the bin wall/base. Although the moist conditions thatfacilitated the precipitation of gypsum crystals might be considereda factor contributing to the dissolution of dung spherulites, gypsumcrystals do occur in associationwith dung spherulites inside anotherbin (3AII-1). In addition, the micromorphological observations pointto particularly elongate phytoliths oriented parallel to each otherand to the sampled bin wall/base creating a layered, laminatedappearance, suggesting that the bin may have been lined with plantmaterial that absorbed excess moisture of the stored material.

In sum, these observations of the contents and structure ofmicro-remains collectively suggest that the white deposits at thebin base represent primary remains of mostly grass inflorescencescontained in this feature. The identified morphologies of phytolithsinside the bin corresponded to the floral parts or husks of grassseeds, primarily fromwheat (Triticum sp.) and barley (Hordeum sp.,Fig. 7g). These identifications are consistent with macrobotanicalevidence. The charred remains inside the bins are dominated byrachises and chaffs of barley and wheat, which outnumber cerealgrains or straw fragments. Particularly, their light chaffs, some ofwhich are mineralized rather than charred, occur most frequentlyin the bottom fill. These results indicate that the bin (4BIIX-94)contained at least chaffs and rachises of wheat and/or barley thathad been separated from straws. Although cereal grains may havebeen also included, their volume must have been limited in com-parison with chaffs.

In contrast to the primary nature suggested for the bottom de-posits, on top of which two grinding stones were placed, the fillsabove them are considered secondary influx from outside the bin.This scenario is in accord with the lower density of phytoliths andlower proportion of inflorescence phytoliths in the middle fill incomparison with the bottom fill. The occurrence of dung spheru-lites in the middle fill, in contrast its absence in the bottom fill, alsoindicate different depositional processes between them.

We are uncertain about how this secondary influx into the bininfluenced the primary contents of the bin. This is partly becausewe do not have micromorphological observations of the contactbetween the bottom fill (i.e., white deposits) and the fills above. Forexample, the bottom fill (slides 1e3 of S25) includes few clay ag-gregates averaging 1 mm in diameter (Fig. 10: BeC) that resembleinmicrostructure and content of the binwall (slides 4e5). Althoughthese aggregates in the bottom fill might suggest re-use of the binswhere, for example, re-opening and re-filling of its contents causedtiny fragments of the clay bin wall to become dislodged andamalgamated into the stored contents, they may also derive fromsecondary intrusions into the porous phytolith-rich layer.

The same problem constrains the interpretation of wild plantsassociated with wheat and barley in the bin. For example, wedetected a notable concentration of Artemisia-type seeds in thelower (and partly bottom) fill (Table 3). Distributed from Europe toCentral Asia, Artemisia is a kind of herb traditionally used forvarious purposes (e.g., as a vermifuge, antispasmodic, antiseptic,and insecticide). The smoke of Artemisia santonicum is knownethnographically to have been used to drive away snakes(Mohammad Al Sayed, 2010). Given these traditional uses, wepropose the possibility that the plant may have been intentionallyburned and placed on the stored cereals as a preservative and/orinsecticide. However, their primary nature cannot be establishedwith available records.

6.3. Implications of the chaff concentration in the bin (4BIIX-94)

The phytolith analyses noted that the phytolith-rich deposits atthe bottom of the bin (4BIIX-94) are characterized by very highproportions (74e78%) of inflorescences and low proportions(1.3e5.8%) of anatomically connected phytoliths. Given the asso-ciation of the two complete grinding stones (4BIIX-94a and 94b)inside the bin and another (4BIIX-92a) just beside the bin, thescarcity of multicellular structures can be interpreted as a result ofcereal processing (probably dehusking) that involved the use ofgrinding stones, as indicated by studies carried out with grindingtools from diverse geographical areas using a similar methodo-logical approach (Albert and Portillo, 2005; Portillo, 2006; Portilloand Albert, 2014; Portillo et al., 2009, 2013). Recent experimentalstudies conducted with einkorn wheat have demonstrated quan-titatively that multicelled phytolith size decreases as a result of thebreakdown produced by mechanical degradation suffered throughboth dehusking and grinding processes (Portillo et al., 2013). Inaddition, the decrease of anatomically connected phytoliths canalso result from varied depositional and post-depositional pro-cesses (Jenkins, 2009; Shillito, 2011; Cabanes et al., 2011). Despitethe moderately alkaline condition (Table 1), which may acceleratephytolith dissolution, this factor has not significantly affected thecomposition and state of phytoliths at the base of the bin. First,these samples showed a low dissolution index (1.9e2.4%). Inaddition, there was high concentration of diagnostic inflorescencemorphotypes (up to 78%), mostly decorated long cells, which ac-cording to Cabanes et al. (2011) are considered as fragile formsespecially susceptible to dissolution.

Therefore, despite the possibility of phytolith dissolution, thephytolith-dominated deposits in the bin (4BIIX-94) indicate a


practice that separated chaffs and rachises from straws and grains.This must have been achieved by first separating unhusked grainsfrom straws (e.g., threshing or harvesting ears selectively). Then,unhusked grains were dehusked possibly with grinding stones,followed by the separation between grains and chaffs (e.g., win-nowing or sieving). It is notable that unhusked grains were onceremoved from straws even if macrobocanical evidence suggeststhat naked barley and free-threshing wheat were mainly used atG€oytepe. While dehusked grains must have been consumed byhumans, chaffs and rachises were kept for future use in the bin.

One of the uses of chaffs may have been fodder as we find as-sociations of concentrated phytoliths (including inflorescences)with abundant dung spherulites in some of the sediment samples(S-24, 29, and 32). The storage of fodder, in turn, suggests ananticipation of occupying the site with livestock when stubbleswould be scarce in the field, i.e., late falleearly summer. Other likelyuses of chaffs include temper for mudbricks and pottery. Theformer is a major construction material at G€oytepe, and the latteruse in level 10, where the bin (4BIIX-94) is located, corresponds tothe timing of an increase in plant-tempered pottery replacing thatwith mineral-temper.

However, why were the chaffs stored in the bin not eventuallyconsumed? At G€oytepe, we found a number of contexts, whereusable tools (e.g., complete ground stones, bone artifacts, and slingstones) were placed, sometimes concentrated, besides buildingwalls or in bin features, suggesting tool caches that were noteventually used (Guliyev and Nishiaki, 2012). The chaffs in the bincan be considered one of such unconsumed caches that are prev-alent at G€oytepe. The caching behaviors in historical and archaeo-logical records are often interpreted as representing inhabitant'santicipation of reoccupying the settlement after a relatively shortperiod of planned abandonment (e.g., Stevenson, 1982; Henry et al.,2014). Thus, we raise a possibility that Neolithic inhabitants (atleast some of them) at G€oytepe left the settlement seasonally. Thismay have created opportunities for obtaining/trading long-distanceresources, such as obsidian, consumed at the site. This settlementsystem encouraged caching behaviors, and some of the caches wereinadvertently left unconsumed.

6.4. Storage space in the Neolithic community at G€oytepe

The observations on primary and secondary contents and spatialcontexts of the bins analyzed in this study collectively indicate thatstorage space at G€oytepe was not spatially segregated from, butrather closely connected to, domestic areas. These areas consistedof round residential buildings and a courtyard where daily activ-ities, such as food processing and burning took place. A spatialrelationship of this nature between domestic and storage areas, inturn, suggests that cereal storagewasmainly performed at the levelof the group, who resided in round buildings adjacent to a court-yard. Furthermore, the appearance of this settlement organizationthrough different levels (at least Levels 3, 4, 5, and 10) suggests thatstorage management continued to be a significant component ofhousehold activities beginning with the early occupations atG€oytepe. Thus, the excavations of lower levels or sites dated earlierthan G€oytepe (e.g., Hacı Elamxanlı Tepe: Nishiaki et al., 2013)should help clarify the question of when and how this storagepractice developed at the transition from a foraging to an agricul-tural economy in the southern Caucasus.

7. Conclusion

This paper examined the context and use of storage at G€oytepethrough geoarchaeological and palaeobotanical analyses of thedeposits inside and outside clay bin features. Although this study

focused on a small number of selected clay bins, the results of themulti-scalar analyses allowed us to detect direct and indirectremnants of cereal storage by Neolithic farmers. We then discussedthe implications of the storage in terms of other aspects of Neolithicfarmers, such as cereal processing, seasonal movements, andhousehold activities. In particular, possible seasonal movementsassociatedwith the procurement of distant resources, e.g., obsidian,by G€oytepe inhabitants are reminiscent of the dispersed andsometimes transient nature of contemporary (i.e., late Neolithic)agro-pastoral societies in upper Mesopotamia and the Levant,where dispersed settlements were probably inter-linked throughregional networks (Banning et al., 1994; Kadowaki et al., 2008;Nieuwenhuyse et al., 2013). In addition, this study has shownthat storage was a significant component of household activities atG€oytepe like other early agricultural communities inwest Asia, thuspromising further studies of storage practices in order to revealearly developments of Neolithic households (Flannery, 2002;Byrd, 2005; Garfinkel, 2006; Kadowaki, 2012; Henry et al., 2014).We plan to increase the size and contexts of samples in the future inorder to present a more comprehensive picture of the reconstruc-tion of early Neolithic village and household storage dynamics andtheir socioeconomic complexity. This will allow for more system-atic analyses of synchronic and diachronic variability of storageactivities and more generally the use of space in the early agro-pastoral settlement at G€oytepe and other early farming commu-nities in the southern Caucasus.

Acknowledgments

We are grateful to Dr. Maisa N. Ragimova of the Institute ofArchaeology and Ethnography, the National Academy of Science,Azerbaijan, for permission and support for the research at G€oytepe.We also thank the crew members of the renewed expedition atG€oytepe since 2008. Kazuya Shimogama kindly translated anAzerbaijanese reference cited in this study. The financial supportfor this study was provided by the National Academy of Science ofAzerbaijan, the JSPS KAKENHI (26770265), and the TakanashiFoundation for Arts and Archeology (2011-16). The research per-formed by M.P. has been funded by the Juan de la Cierva Sub-programme (JCI2009-04217, Spanish Ministry of Competitivenessand Economy, MINECO) and the Catalan Agency for Universitiesand Research Grants, AGAUR (SGR2009 1418). Laia Maci�a (GEPEG)helped in the phytolith laboratory. We also appreciate commentsfrom two anonymous reviewers, who helped us clarify the paper.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jas.2014.10.021.

References

Albert, R., Esteve, X., Portillo, M., Rodríguez-Cintas, A., Cabanes, D., Esteban, I.,Hern�andez, F., February 20, 2014. Phytolith CoRe, Phytolith Reference Collec-tion. Retrieved February 20, 2014, from. http://www.gepeg.org/enter_PCORE.html.

Albert, R.M., Lavi, O., Estroff, L., Weiner, S., Tsatskin, A., Ronen, A., Lev-Yadun, S.,1999. Mode of occupation of Tabun Cave, Mt Carmel, Israel during the Mous-terian Period: a study of the sediments and phytoliths. J. Archaeol. Sci. 26,1249e1260.

Albert, R.M., Portillo, M., 2005. Plant uses in different Bronze and Iron Age settle-ments from the Nuoro Province (Sardinia): the results of phytolith analysesfrom several ceramic fragments and grinding stones. Anejos Complutum 10,109e119.

Albert, R.M., Shahack-Gross, R., Cabanes, D., Gilboa, A., Lev-Yadun, S., Portillo, M.,Sharon, I., Boaretto, E., Weiner, S., 2008. Phytolith-rich layers from the LateBronze and Iron Ages at Tel Dor (Israel): mode of formation and archaeologicalsignificance. J. Archaeol. Sci. 35, 57e75.



http://www.gepeg.org/enter_PCORE.html

http://www.gepeg.org/enter_PCORE.html

http://refhub.elsevier.com/S0305-4403(14)00391-4/sref2
















Albert, R.M., Weiner, S., 2001. Study of phytoliths in prehistoric ash layers using aquantitative approach. In: Meunier, J.D., Colin, F. (Eds.), Phytoliths, Applicationsin Earth Sciences and Human History. A.A. Balkema Publishers, Lisse,pp. 251e266.

Azerbaijan SSR Academy of Sciences, Institute of Geography, 1963. Atlas ofAzerbaijan Soviet Socialist Republic (In AzerbaijanIs). USSR National Committeeof Geology, Department of Geodesy and Cartography, Baku-Moscow.

Badalyan, R., Harutyunyan, A.A., Chataigner, C., Le Mort, F., Chabot, J., Brochier, J.E.,Balasescu, A., Radu, V., Hovsepyan, R., 2010. The settlement of Aknashen-Khatunarkh, a Neolithic site in the Ararat plain (Armenia): excavation results2004-2009. TÜBA-AR 13, 185e218.

Badalyan, R., Lombard, P., Avetisyan, P., Chataigner, C., Chabot, J., Vila, E.,Hovsepyan, R., Willcox, G., Pessin, H., 2007. New data on the late prehistory ofthe southern Caucasus. The excavations at Aratashen (Armenia): preliminaryreport. In: Lyonnet, B. (Ed.), Les Cultures du Caucase (VIe-IIIe mill�enaires avantnotre �ere): Leurs relations avec le Proche-Orient. CNRS �Editions, Paris,pp. 37e61.

Banning, E.B., 2003. Housing Neolithic farmers. Near East. Archaeol. 66 (1e2), 4e21.Banning, E.B., Rahimi, D., Siggers, J., 1994. The Late Neolithic of the southern Levant:

hiatus, settlement shift or observer bias? the perspective from Wadi Ziqlab.Pal�eorient 20.2, 151e164.

Banning, E.B., Dods, R.R., Field, J., Kuijt, I., McCorriston, J., Siggers, J., Taani, H.,Triggs, J., 1992. Tabaqat al-Bûma: 1990 excavations at a Kebaran and LateNeolithic site in Wadi Ziqlab. Annu. Dep. Antiq. Jordan 49, 43e69.

Bar-Yosef, O., 1998. The Natufian culture in the Levant, threshold to the origins ofagriculture. Evol. Anthropol. 6, 159e177.

Berna, F., Behar, A., Shahack-Gross, R., Berg, J., Boaretto, E., Gilboa, A., Sharon, I.,Shalev, S., Shilstein, S., Yahalom-Mack, N., Zorn, J.R., Weiner, S., 2007. Sedimentsexposed to high temperatures: reconstructing pyrotechnological processes inLate Bronze and Iron Age strata at Tel Dor (Israel). J. Archaeol. Sci. 34, 358e373.

Brewer, R., 1976. Fabric and Mineral Analysis of Soils, second ed. Krieger,Huntington.

Brown, D.A., 1984. Prospects and limits of a phytolith key for grasses in the centralUnited States. J. Archaeol. Sci. 11, 345e368.

Byrd, B.F., 2005. Early Village Life at Beidha, Jordan: Neolithic Spatial Organizationand Vernacular Architecture. The Excavations of Mrs. Diana Kirkbride-helbaek.Oxford University Press/Council for British Research in the Levant, Oxford.

Cabanes, D., Weiner, S., Shahack-Gross, R., 2011. Stability of phytoliths in thearchaeological record: a dissolution study of modern and fossil phytoliths.J. Archaeol. Sci. 38, 2480e2490.

Canti, M.G., 1999. The production and preservation of faecal spherulites: animals,environment and taphonomy. J. Archaeol. Sci. 26, 251e258.

Chataigner, C., 1995. La Transcaucasie au N�eolithique et au Chalcolithique. In: BARInternational Series, vol. 624. Hadrian Books, Oxford.

Coil, J., Korstanje, M.A., Archer, S., Hastorf, C.A., 2003. Laboratory goals and con-siderations for multiple microfossil extraction in archaeology. J. Archaeol. Sci.30, 991e1008.

Courty, M.A., Goldberg, P., Macphail, R., 1989. Soils and Micromorphology inArchaeology. Cambridge University Press, Cambridge.

Finlayson, B., Kuijt, I., Arpin, T., Chesson, M., Dennins, S., Goodale, N., Kadowaki, S.,Maher, L., Smith, S., Schurr, M., Mckay, J., 2003. Dhra' excavation project, 2002interim report. Levant 25, 1e38.

Fitzpatrick, E.A., 1993. Soil Microscopy and Micromorphology. John Wiley and Sons,Toronto.

Flannery, K., 2002. The origins of the village revisited: from nuclear to extendedhouseholds. Am. Antiq. 67 (3), 417e433.

Friesem, D., Boaretto, E., Eliyahu-Behar, A., Shahack-Gross, R., 2011. Degradation ofmud brick houses in an arid environment: a geoarchaeological model.J. Archaeol. Sci. 38, 1135e1147.

Friesem, D., Karkanas, P., Tsartsidou, G., Shahack-Gross, R., 2014. Sedimentary pro-cesses involved in mud brick degradation in temperate environments: amicromorphological approach in an ethnoarchaeological context in northernGreece. J. Archaeol. Sci. 41, 556e567.

Garfinkel, Y., 1987. Yiftahel: a Neolithic village from the seventh millennium B.C. inLower Galilee, Israel. J. Field Archaeol. 14, 199e212.

Garfinkel, Y., 2002. The effect of population size on the human organization atSha'ar Hagolan. In: Garfinkel, Y., Miller, M.A. (Eds.), Sha'ar Hagolan 1: NeolithicArt in Context. Oxbow Books, Oxford, pp. 257e262.

Garfinkel, Y., 2006. The social organization at Neolithic Sha'ar Hagolan: the nuclearfamily, the extended family & the community. In: Banning, E.B., Chazan, M.(Eds.), Domesticating Space: Construction, Community, and Cosmology in theLate Prehistoric Near East. ex oriente, Berlin, pp. 103e111.

Goldberg, P., 1980. Micromorphology in archaeology and prehistory. Pal�eorient 6,159e164.

Goldberg, P., 2000. Micromorphological aspects of site formation at Keatley Creek.In: Hayden, B. (Ed.), The Ancient Past of Keatley Creek, vol. 1. Taphonomy.Archaeology Press, Simon Fraser University, Burnaby, BC, pp. 81e94.

Goldberg, P., Miller, C.E., Schiegl, S., Ligouis, B., Berna, F., Conard, N.J., Wadley, L.,2009. Bedding, hearths, and site maintenance in the middle stone Age of SibuduCave, KwaZulu-Natal, South Africa. Archaeol. Anthropol. Sci. 1 (2), 95e122.

Goodman-Elgar, M., 2008. The devolution of mudbrick: ethnoarchaeology ofabandoned earthen dwellings in the Bolivian Andes. J. Archaeol. Sci. 35,3067e3071.

Goring-Morris, N., Belfer-Cohen, A., 2008. A roof over one's head: developments inNear Eastern residential architecture across the Epipalaeolithic-Neolithic

transition. In: Bocquet-Appel, J.-P., Bar-Yosef, O. (Eds.), The Neolithic De-mographic Transition and its Consequences. Springer, pp. 239e286.

Guliyev, F., Nishiaki, Y., 2012. Excavations at the Neolithic settlement of G€oytepe, themiddle Kura Valley, Azerbaijan, 2008e2009. In: Matthews, R., Curtis, J. (Eds.),Proceedings of the 7th International Congress of the Archaeology of the AncientNear East, Fieldwork and Recent Research, vol. 3. Harrassowitz Verlag, Wies-baden, pp. 71e84.

Hamon, C., 2008. From Neolithic to Chalcolithic in the southern Caucasus: economyand macrolithic implements from Shulaveri-Shomu sites of Kwemo-Kartli(Georgia). Pal�eorient 34 (2), 85e135.

Henry, D.O., Kadowaki, S., Bergin, S.M., 2014. Reconstructing early Neolithic socialand economic organization through spatial analysis at Ayn Ab�u Nukhayla,southern Jordan. Am. Antiq. 79 (3), 401e424.

Jenkins, E., 2009. Phytolith taphonomy: a comparison of dry ashing and acidextraction on the breakdown of conjoined phytoliths formed in Triticum durum.J. Archaeol. Sci. 36, 2402e2407.

Kadowaki, S., 2012. A household perspective towards the Pre-Pottery Neolithic toLate Neolithic cultural transformation in the southern Levant. Orient 47, 3e28.

Kadowaki, S., Gibbs, K., Allentuck, A., Banning, E.B., 2008. Late Neolithic settlementin Wadi Ziqlab, Jordan: al-Basatîn. Pal�eorient 34 (1), 105e129.

Kiguradze, T., Menabde, M., 2004. The Neolithic of Georgia. In: Sagona, A. (Ed.),A View from the Highlands: Archaeological Studies in Honour of Charles Bur-ney. Peeters, Leuven, pp. 345e398.

Kuijt, I., 2008. Demography and storage systems during the southern LevantineNeolithic demographic transition. In: Bocquet-Appel, J.-P., Bar-Yosef, O. (Eds.),The Neolithic Demographic Transition and its Consequences. Springer,pp. 287e313.

Kuijt, I., 2011. Home is where we keep our food: the origins of agriculture and latePre-Pottery Neolithic food storage. Pal�eorient 37 (1), 137e152.

Kuijt, I., Finlayson, B., 2009. Evidence for food storage and predomestication gra-naries 11,000 years ago in the Jordan Valley. Proc. Natl. Acad. Sci. 106,10966e10970.

Lyonnet, B., Guliyev, F., Helwing, B., Aliyev, T., Hansen, S., Mirtskhulava, G., 2012.Ancient Kura 2010-2011: the first two seasons of joint field work in theSouthern Caucasus. Arch€aologische Mittl. aus Iran Turan 44, 1e190.

Love, S., 2012. The geoarchaeology of mudbricks in architecture: a methodologicalstudy from Çatalh€oyük, Turkey. Geoarchaeology 27, 140e156.

Madella, M., Alexandre, A., Ball, T.B., Working Group, I.C.P.N., 2005. Internationalcode for phytolith nomenclature 1.0. Ann. Bot. 96, 253e260.

Matthews, W., 2010. Geoarchaeology and taphonomy of plant remains and micro-archaeological residues in early urban environments in the Ancient Near East.Quat. Int. 214, 98e113.

Matthews, W., 2012a. Defining households: micro-contextual analysis of earlyNeolithic households in the Zagros, Iran. In: Parker, B., Foster, C. (Eds.), NewPerspectives on Household Archaeology. Eisenbrauns, Winona Lake,pp. 183e218.

Matthews, W., 2012b. Household life-histories and boundaries. In: Tringham, R.,Stevanovic, M. (Eds.), Last House on the Hill: BACH Area Reports from Çata-lh€oyük, Turkey. Cotsen Institute of Archaeology Press and British Institute atAnkara, Los Angeles and Ankara, pp. 205e222.

Matthews, W., French, C.A.I., Lawrence, T., Cutler, D.F., Jones, M.K., 1997. Micro-stratigraphic traces of site formation processes and human activities. WorldArchaeol. 29, 281e308.

Mohammad Al Sayed, A., 2010. Medicinal Herbal Guide. Alfa Publishing (inArabic).

Molist, M., 1998. Espace collectif et espace domestique dans le N�eolithique desIX�eme et VIII�eme mill�enaires B.P. au nord de la Syrie: Apports du site de TellHalula (valle�e de l'Euphrate). In: Fortin, M., Aurenche, O. (Eds.), Natural Space,Inhabited Space in Northern Syria (10th-2nd Millennium B.C.). Canadian Soci-ety for Mesopotamian Studies, Qu�ebec, pp. 115e130.

Moore, A.M.T., 2000. The excavation of Abu Hureyra 2. In: Moore, A.M.T.,Hillman, G.C., Legge, A.J. (Eds.), Village on the Euphrates. Oxford UniversityPress, Oxford, pp. 189e259.

Mulholland, S.C., Rapp Jr., G., 1992. A morphological clasification of grass silica-bodies. In: Rapp Jr., G., Mulholland, S.C. (Eds.), Phytolith Systematics: EmergingIssues, Advances in Archaeological and Museum Science. Plenum Press, NewYork, pp. 65e89.

Narimanov, I.G., 1987. Культура древнейшего зеМледельческо-скотоводческогонаселения АзербайдЖана (зпоха знеолита VI-IV тыс. до н.з.). National Academy ofSciences, Baku.

Narimanov, I.G., 1992. The earliest agricultural settlements in the territory ofAzerbaidzhan. In: Kohl, P.L. (Ed.), Recent Discoveries in Transcaucasia. SovietAnthropology and Archeology 30 (4). M. E. Sharpe, N.Y., pp. 9e66 (Englishtranslation of Narimanov 1987, pp. 14e69).

Nieuwenhuyse, O.P., Bernbeck, R., Akkermans, P.M.M.G., Rogasch, J. (Eds.), 2013.Interpreting the Late Neolithic of Upper Mesopotamia. Brepols, Turnhout.

Nishiaki, Y., Guliyev, F., Kadowaki, S., Arimatsu, Y., Hayakawa, Y., Shimogama, K.,Miki, T., Akashi, C., Arai, S., Salimbeyov, S., 2013. Hacı Elamxanlı Tepe: excava-tions of the earliest pottery Neolithic occupations on the middle kura,Azerbaijan, 2012. Arch€aologische Mittl. aus Iran Turan 45. http://www.reimer-mann-verlag.de/controller.php?cmd¼detail&titelnummer¼630450&verlag¼4.

Palumbi, G., 2007. A preliminary analysis on the prehistoric pottery from Aratashen(Armenia). In: Lyonnet, B. (Ed.), Les Cultures du Caucase (VIe-IIIe mill�enairesavant notre �ere): Leurs relations avec le Proche-Orient. CNRS �Editions, Paris,pp. 63e76.

































































































































































































































































http://www.reimer-mann-verlag.de/controller.php%3Fcmd%3Ddetail%26titelnummer%3d630450%26verlag%3D4


















Piperno, D.R., 2006. Phytoliths: a Comprehensive Guide for Archaeologists andPaleoecologists. AltaMira Press, Lanham.

Portillo, M., 2006. La m�olta i triturat d'aliments vegetals durant la Protohist�oria a laCatalunya Oriental (Ph.D. thesis). University of Barcelona.

Portillo, M., Albert, R.M., 2014. Early crop cultivation and caprine herding: thephytolith and phaecal spherulite evidence. In: Henry, D.O. (Ed.), The Sands ofTime: the Desert Neolithic Settlement at Ayn Ab�u Nukhayla. ex Oriente, Berlin,pp. 121e137.

Portillo, M., Albert, R.M., Henry, D.O., 2009. Domestic activities and spatial distri-bution in Ain Abu Nukhayla (Wadi Rum, Southern Jordan): the use of phytolithsand spherulites studies. Quat. Int. 193, 174e183.

Portillo, M., Valenzuela, S., Albert, R.M., 2012. Domestic patterns in the Numidiansite of Althiburos (northern Tunisia): the results from a combined study ofanimal bones, dung and plant remains. Quat. Int. 275, 84e96.

Portillo, M., Bofill, M., Molist, M., Albert, R.M., 2013. Phytolith and use-wear func-tional evidence for grinding stones from the Near East. In: Anderson, P.C.,Cheval, C., Durand, A. (Eds.), Regards crois�es sur les outils li�es au travail desv�eg�etaux. An interdisciplinary focus on plant-working tools, �Ed. APDCA, Anti-bes, pp. 161e174.

Portillo, M., Kadowaki, S., Nishiaki, Y., Albert, R.M., 2014. Early Neolithic householdbehavior at Tell Seker al-Aheimar (Upper Khabur, Syria): a comparison toethnoarchaeological study of phytoliths and dung spherulites. J. Archaeol. Sci.42, 107e118.

Rollefson, G., Simmons, A., Kafafi, Z., 1992. Neolithic cultures at ʹAin Ghazal, Jordan.J. Field Archaeol. 19, 443e470.

Rosen, A.M., 1986. Cities of Clay: the Geoarchaeology of Tells. University of ChicagoPress, Chicago.

Rosen, A.M., 1992. Preliminary identification of silica skeletons from Near Easternarchaeological sites: an anatomical approach. In: Rapp Jr., G., Mulholland, S.C.(Eds.), Phytolith Systematics: Emerging Issues, Advances in Archaeological andMuseum Science. Plenum Press, New York, pp. 129e147.

Shillito, L.-M., 2011. Taphonomic observations of archaeological wheat phytolithsfrom Neolithic Çatalh€oyük, Turkey, and the use of conjoined phytolith size as anindicator of water availability. Archaeometry 53, 631e641.

Shillito, M., Matthews, W., Almond, M., Bull, I., 2011. The microstratigraphy ofmiddens: capturing daily routine in rubbish at Neolithic Catalhoyuk, Turkey.Antiquity 85, 1024e1038.

Stevenson, M.G., 1982. Toward an understanding of site abandonment behavior:evidence from historic mining camps in the southwest Yukon. J. Anthropol.Archaeol. 1, 237e265.

Stoops, G., Marcelino, V., Mees, F. (Eds.), 2010. Interpretation of MicromorphologicalFeatures of Soils and Regoliths. Elsevier, Amsterdam.

Tsartsidou, G., Lev-Yadun, S., Albert, R.M.,Miller-Rosen, A., Efstratiou,N.,Weiner, S., 2007.The phytolith archaeological record: strengths and weaknesses based on a quanti-tative modern reference collection from Greece. J. Archaeol. Sci. 34, 1262e1275.

Twiss, P.C., 1992. Predicted world distribution of C3 and C4 grass phytoliths. In:Rapp Jr., G., Mulholland, S.C. (Eds.), Phytolith Systematics: Emerging Issues,Advances in Archaeological and Museum Science. Plenum Press, New York,pp. 113e128.

Twiss, P.C., Suess, E., Smith, R.M., 1969. Morphological classification of grass phy-toliths. Soil Sci. Soc. Am. Proc. 33, 109e115.

van Zeist, W., Bakker-Heeres, J.A.H., 1985. Archaeobotanical studies in the Levant, 4:Bronze Age sites on the north Syrian Euphrates. Palaeohistoria 27, 247e316.

Verhoehen, M., 1999. An Archaeological Ethnography of a Neolithic Community:Space, Place and Social Relations in the Burnt Village at Tell Sabi Abyad, Syria.Nederlands Historisch-Archaeologisch Instituut Te Istanbul.

Wadley, L., Sievers, C., Bamford, M., Goldberg, P., Berna, F., Miller, C., 2012. MiddleStone Age bedding construction and settlement patterns at Sibudu, South Af-rica. Science 334, 1388e1391.

World Meteorological Organization, 2014. Weather Information for Ganja. WorldWeather Information Service. http://www.worldweather.org/007/c00019.htm(Last accessed 12 March 2014).






















































































http://www.worldweather.org/007/c00019.htm

Kadowaki, S., ..., F. Guliyev, and Y. Nishiaki (2015) Geoarchaeological and palaeobotanical evidence for prehistoric cereal storage at the Neolithic settlement of Göytepe (mid 8th

Documents