8/7/2019 Ant 0740611

1/13

Palaeoenvironments and economy of Iron AgeSaka-Wusun agro-pastoralists in southeasternKazakhstanARLENE ILLER OSEN,CLAUDIA HANG

& FEDOR AVLOVICHRIGORIEV"A detailed geoarchaeological and environm ental s tu dy of southeastern K azakhstan revealssubtle changes of land use and envir onm ent during the Iron Age. Major eco nom ic changesfro m pastoralism to agriculture over time m a y be reinterpreted fro m these n ew fin din gs .Key-words: Kazakh stan, Cent ra l Asia , geoarchaeology, envi ro nm enta l a rchaeology, phyol i ths , noma ds , agr icul ture

IntroductionThe acquisition of agricultural goods among IronAge pastoralists of Central Asia is a topic cru-cial for understanding the relationship betweenthe nomads and their sedentary neighbours.Many authors have characterized nomadiceconomies as imbalanced and dependent onsettled agriculturalists of the great civilizationsto the east and west of the Eurasian steppes(Lattimore 1979; Khazanov 1984).This represen-tation allows little place for intensive agriculturalpursuits (Kazanov 1984:17),and perceives Cen-tral Asian environments as being largely hos-tile to farmers since the end of the Mid-HoloceneNeolithic optimum (Yablonsky 1995a; 199513).Di Cosmo (1994) challenges this view by dem-onstrating the importance of cultivation amongIron Age pastoral nomads in Inner Asia. He useshistorical, archaeological and ethnographic stud-ies to demonstrate that farming was commonlypractised as a supplementary activity within re-gions dominated by nomadic peoples. The no-mads either grew agricultural products themselvesor dominated small farming communities withintheir domain. Evidence for agricultural pursuitsalso comes from archaeological finds in kurganburial mounds which include cereal grains aswell as iron agricultural implements (Debaine-Francfort 1988; 1989).

Recent archaeological investigations, directedby Chang and Gregoriev at the site of Tuzusaiin southeastern Kazakhstan, have attempted toelucidate the nature of agro-pastoral nomadiceconomies among the Saka-Wusun Period no-madic groups (Chang& Tourtellotte 1998). Herewe present regional geoarchaeological evidencefor climatic amelioration during the Saka (fromc. 750-300 BC) and Wusun Period (c . 300 BC-AD 200) occupations which was favourable toagricultural pursui ts. Phytol ith data at the siteof Tuzusai suggest agricultural intensificationduring this time period as well. The environ-mental shift occurred at a time when histori-cal trends made agricultural intensificationsocially and politically advantageous, demon-strating the importance of converging environ-mental and historical transitions in changingagricultural strategies.Natural settingThe site of Tuzusai is located on the Talgar al-luvial fan, on the northern foothills of the TienShan mountain range in the Semirechye regionof southeastern Kazakhstan (FIGURE ) .Theregion as a whole is characterized by a richlydiverse mosaic of landscapes within a relativelyrestricted area. Today the Talgar area is semi-arid receiving an annual average rainfall of 463

* Rosen. Ins t i tu te of Archaeology, Univers i ty Col lege Londo n, 31-34 G or don S qua r e , L ondon W C l H O PY , E ngla nda.rosen@ucl.ar : .uk(:hang, Department of Anthrop ology, Sweet Br ia r Col lege, Sweet Br ia r VA 24595, IJSA.Grigoriev, Archaeology Divis ion, Kazakh C ent ra l S ta te M us eu m , Almaty, Kazakhstan.Received 20 July 19 99, accepted 9 No vemb er 199 9, r evised 8 M ay 2000.A N T I Q U 1 7 Y 74 (2000 ): 611-23

8/7/2019 Ant 0740611

2/13

612 ARLENE MILLER ROSE N, CLAUDIA CHANG & FEDOR PAVLOVICH GRIGORIEVI

mm per year with heaviest precipitation in thespring and autumn. To the south are the highgranitic mountains of the Tien Shan with peaksreaching over 7000 m. The advance and retreatof alpine glaciers have left a valuable record ofPleistocene and Holocene climatic fluctuations.Silts produced by these glaciers have washedout onto the plain to the north, where they weredeposited in the semi-arid steppe and desertzones. In the cool arid episodes of the latePleistocene and Holocene, these silts were en-trained by the prevailing northerly winds, andredeposited as the wind-blown loess whichblankets the northern Tien Shan foothills alongthe margin of the uplift zone.The transition from mountain to steppe isabruptly defined by the northern Tien Shan faultline which separates the horst features of theTien Shan mountains and the half-graben ofthe Ily River Valley. The Talgar alluvial fan issituated at the juncture between the mountainsand the plain. It is an approximately 30-km widestrip of land composed of distinct sedimen-tary units of coarse gravels, reworked loess andpoorly sorted sandy silts and gravels. Thesesediment units record a rich history of LateHolocene environmental change.

FIGURE.Locationm a p of study area.The vegetation was once a bunch-grass steppe

dominated by Stipa capil lata, S. caucasica andFestuca vaiesiaca),with trees such as elm, mapleand willow growing along the stream banks(ENVIRC 1995). Further north toward the Ilyriver, precipitation drops and the steppe zonerapidly transforms to desert. At the point wherethe toe of the fan descends into the Ily Rivervalley, water tables are high, as indicated bynumerous stands of common reed (Phragmi tessp.) present even today. Under moister climaticconditions the marshes of this area would havehad a greater extent than those of today, pro-viding pasturage and areas possibly suitablefor wetland cultivation. This region has beenan attractive zone for sett lement by agro-pas-toral groups since the Iron Age because of itsgreat environmental diversity within a mere80 km.Background to the site of TuzusaiTuzusai is a Saka and Wusun period settlement( c . 400 BC-AD 100) located on the edge of theTalgar alluvial fan at an elevation of 740 m. Ithas been excavated since the late 1980s by FedorP. Grigoriev and most recently by the Kazakh-American Talgar Project from 1994-1996.

8/7/2019 Ant 0740611

3/13

PALAEOENVIRONMENTS A ND ECONOMY OF IRON AGE SOUTHEASTERN KAZAKHSTAN 613The site consists of four occupation periods

and at least six different archaeological hori-zons. The radiocarbon dates for the earliestperiod of occupation (here referred to as Oc-cupation Period 4 or OP-4) approximately rangefrom 355 to 115 cal BC (2150+30 to 2350+ 40BP). Occupation Period 2 (OP-2) is dated from180 cal BC to 75 cal A D , and the latest Occupa-tion Period 1 (OP-1) is dated from 1650 to 1950

There have been disagreements about thenature of the architecture found at Tuzusai. In1994 Chang identif ied fragments of mudbrickwalls, most likely associated with the latest IronAge occupation of OP-2 ( c. 100 A D ) , and alsoevidence for the existence of semi-subterraneanhouses. However, Grigoriev disagrees with thisinterpretation of pit-houses at the site. The onecomparative example of semi-subterranean orhouse pit dwellings described for the Eurasiansteppe comes from the Odessa area, VarvarovskiI, dating from the 4th to the 3rd centuries BC(late Scythian) (Khazanov 1975: 70). In addi-tion to occupation surfaces at the site, the ex-cavators found a large number of ashy storageand refuse pits as well as hearths.

In the vicinity of Tuzusai and along ancientand current stream beds such as the Tuzusaistream and the Tsiganka River, we identifiedceramic materials and grinding stones on thesurface of the plough zone. The artefact scat-ters and ceramic finds indicate a Saka andWusun spatial pattern of medium to large set-tlements (similar to Tuzusai and often associ-ated with kurgans (burial mounds) of the period),which are located in the steppe zones of richsilts and chernozem-like soils in areas of thehighest agricultural potential.Paleoenvironmental changes and theirinfluence on Saka-Wusun settlement andeconomiesAlthough monocausal and environmentallydeterministic explanations are seldom satisfac-tory for the explanat ion of culture change, it isimportant to examine the effects of environ-mental change as either an enabler or inhibi-tor of other cultural transformations with in achanging social milieu. It is from this perspec-tive that we examine the past environments ofthe Semirechye and their influence on the agro-pastoral pursuits of the Iron Age Saka-Wusunnomadic peoples.

AD.

Palyn ologica1 evidenceMuch of the early literature on Holocene palyno-logical studies from Central Asia is summarizedby Khotinskiy (1984). Due to the agreementbetween the classic European Holocene climaticphases and the Siberian bog cores, Khotinskiycorrelates the pollen sequences and uses the Eu-ropean terminology for climatic phases. For thepurposes of this paper all radiocarbon dates forproxy paleoenvironmental data will be reportedwith their calibrated equivalents in order to fa-cilitate comparison with the archaeological record.

The Mid-Holocene Atlantic period, climati-cally a warm-moist optimum, witnessed thebeginning of Neolithic communities in CentralAsia. Although there is little direct pol len in-formation for the steppe and desert zone it islikely that their overall steppe/desert charac-ter was established during this period. Accordingto Khotinskiy the transition to the Subborealwas marked by three distinct environmentalchanges, a cool-dry early Subboreal from c.4600BP (3357 cal BC ) to 4100 BP (2600 cal BC), a warm-moister middle Subboreal from 4100 BP (2600cal BC) to 3200 BP (1440 cal BC), and another coollate Subboreal episode from 3200 BP (1440 calBC ) to 2500 BP (c.660 cal BC ) (see TABLE ).Thislast cool phase corresponds to the first appear-ance of the Saka peoples in the Semirechye re-gion. Yablonsky (1995a; 1995b) suggested thata dry episode was responsible for an initialmigration southward. Finally, the Siberian pol-len data point to warmer moist condit ions pre-vailing at the beginning of the Subatlantic fromc. 660 BC (Khotinskiy 1984).

An environmental sequence for the Holocenehas also been outlined by Krementski (1997) forKazakhstan and Western Siberia based on pol-len from lake cores (TABLE). e points out thatthe scenario is roughly consistent with the se-quence of transgressions and regressions fiom LakeBalkhash (Venus 1985;Khrustalev & Chernousov1992, both cited by Krementski 1997), yet hisinterpretation for the period beginning at 650BC conflicts wi th that of Khotinskiy.

In addition to these conflicts in interpreta-tion, one must use caution when comparingpollen data to archaeological phases. It is dif-ficult to interpolate calendar dates on pollensequences because of variations in sedimenta-tion rates. One also must consider the lag-timebetween actual climatic change and the colo-nization by new vegetation communities.

8/7/2019 Ant 0740611

4/13

61 4 ARLENE MILLER R OSEN. CLAUDIA CHAN G & FEDOR PAVLOVICH GRIGORIEV:alendar Date!

1500 AD14001300120011001000900800700600500400300200100

0 BCIAD100BC200

30040050060070080090010001100120013001400150016001700180019002000

mul

algarFa n-Unit 2Semi-Arid

-~Erosion

LandscapeStabilitysot1 2

VarmlMoist

~ ~~

Unit 3AlluvialSiltCoollMoist

TABLE . Graphic representat ion ofselected palaeo env ironm enta l reconstruct ions. Data are derivedfrom the fo l lowing sources: * Savosku l 6. olomina 1996; + K rem en t sk i 1997; ho t i nsk i y 1984; O V e n u s1985; Khrustalev & C hernousov 1992. The calendar dates com e from calibrated +Cdates .Geomorph ological evideri ceEvidence for later Holocene climatic change canbe found in Alpine glacial advances and re-treats in the Tien Shan range. Savoskul &Solomina (1996)used radiocarbon dates of soilson end-moraines and lichenometric dating oftill deposits to establish a sequence of glacialadvances and retreats in the alpine areas whichreflects distinct climatic changes within theHolocene (see TABLE ) .Two soils identifiedby Savoskul & Solomina (1996) are indicativeof warm-moist intervals between the glacialadvances and can be tentatively correlated withtwo soil horizons identified on the Talgar fan.The soil from Savoskul& Solominas(1996)Site5 yielded two radiocarbon dates, 1740k70 BP(calibratedas 264,281 or 32 9 AD), and 1710i240BP (347, 360 or 374 AD). A later soil from Site

6 is dated to 380+30 (1483AD). When calibratedthis last date immediately precedes the LittleIce Age alpine glaciation.Geoarch aeological investigations on theTalgar alluvial fanGeoarchaeological reconnaissance was under-taken in the region of the Talgar alluvial fan inthe interest of three main goals:1 to find evidence for landscape evolution

which would provide information on cli-matic and micro-environmental changesduring the Late Holocene period:to assess the landscape potential for agri-cultural and pastoral pursuits through time;and

3 to look for evidence of human manipula-tion and impact on the landscape.

2

8/7/2019 Ant 0740611

5/13

PALAEOENVIRONMENTS AND ECONOMY O F IRON AGE SOUTHEASTERN KAZAKHSTAN 6

FIGURE. Locationm a p of sec t ions onth e Talgar alluvialf a n .

Along the extent of the Talgar fan we wereunable to locate evidence for Holocene depos-its that were earlier in date than the Late Holo-cene. A number of sections along the fanheadand niidfan revealed approximately 4 m of al-luvium dating from c. 8th century BC-AD 17thcentury. At the base of these sections there wasa thick gravel deposit (Unit 4) which i s mostlikely dated to the Middle or Late Pleistocene,based on our finding of a handaxe fragment ona similar deposit to the east of the Talgar fan.This sequence of Late Holocene sediments re-peats itself in all of the sections we observedand described from the area (see FIGURE). Asuitable type-section for this sequence is theTsegonka River section of GS-VI described inTABLE and illustrated in FIGURE. These de-posits form a fill-terrace adjacent to the mod-ern course of the stream. To summarize, theLate Holocene Unit 3 deposits are dated by ar-tefacts to the Saka-Wusun period. They indi-cate a slow accumulat ion of fine silty alluviumin the form of stream overbank deposits or fine-

grained sheetwash. The slow accumulation ofsilty sediments over the fan suggests fine-grainedoutwash from advancing alpine glaciers, im-plying a cool environmental regime as alsoproposed by Khotinskiy (1984) from the pol-len data.

Ultimately, however, the meaning of Unit 3for land use is important regardless of its cli-matic implications. The deposition of fine-grained sediment for an extended period of timeover the surface of the fan would provide a fer-tile medium for both agricultural and pastoralpursuits. The fields would be moist for longperiods of time within a given year and the siltswould be continually renewed, thereby elimi-nating the problem of soil exhaustion.In many exposures Unit 3 is truncated, withan abrupt erosional contact between it and theoverlying Unit 2. However, at GS-VI the top ofthe unit i s capped by a weakly developed soilA-horizon (Soil 2 ) (FIGURE) . Soil 2 constitutesimportant evidence for hundreds of years ofrelatively warm moist environmental stability

8/7/2019 Ant 0740611

6/13

616 ARLENE MILLER ROSEN, CLAUDIA CHANG & FEDOR PAVLOVICH GRIGORIEVunit depth crn colour texture structure

(moist)0-5 3 black silt loam platy soil

(10YR 2/11 stru ctu re53-80/147 dark brown loam; with cut-

(10YR 3/31 and -fill cha nne ldeposits consistingof moderatelywell-sorteddiscoidal an dspherical, well-rounded to sub-round ed gravelsup to 3 c m80/147-215 brown well-com pacted massive

(10 YR 4/31 silt with smallsilt; YO sand;very dark silt loamgreyish A-horizonbrown at top of unit(10 YR 3/21A-horizon

4 215-263 large we ll- grade drounded, beddingdiscoidal chan nelgravel up to 18cm in length,grading upwardsto 2-5 cmspherical an ddiscoidal,rounded towell-rounded gravels

inclusions boundary depositionalenvironment

rootlets, wavy & modern soil 1 A-horizonsparse snail abrup tshellsgley, Mn &Fe stains, abruptirregular & fine-nrained alluviumanimal tooth,small cut-and-fill channelsc. 10 cm deepx 40 cin long

-with small gravel cut-and-fill channel deposi tsoccurring from 108 cm to147 cm. Equivalent toUnit 2 in GS-VII a n d VIII.Possibly Medieval in date

Saka period wavy &sherds, snail graded toshells, bone, abruptgley and som eFe stains,truncatedburiedsoil A-horizonat the top of theunit (-80 cm )granite notgravels visible

fine silty alluvium,overbank or sheetwashdeposits. Approximatelydated to the Saka period;Soil 2 ormed at the top ofthis unit in a period oflandscape stability

alluvial fan bedloaddeposi t date unknow n;equivalent to Unit 4 inGS-VII and GS-VIII

TABLE. Description of sediment sequence from geological section GS-VI.at some time between the later Saka-Wusunperiod and the Medieval period. It is possiblethat it corresponds to the highland soil datedby Savoskul & Solomina (1996)to c. 300 AD.The boundary between Unit 3 and Unit 2 isdistinctly unconformable arid irregular, repre-senting an intensive erosional phase that prob-ably occurred in post-Wusun, early Medievaltimes. Since an alluvial fan is primarily anaggradational feature, it is possible that the ero-sional episode can be attributed to neo-tectonicsubsidence of the Ily basin and a subsequentdrop in base level, or perhaps drier climaticconditions in which lower water tables led tostream incision on the fan. An apparent increasein channel density on the Talgar fan duringMedieval times resulted in broad erosion chan-

nels forming in places where no streams hadexisted before. The end result of this episodewas a patchwork of surfaces of different LateHolocene ages over the face of the fan.At the end of the erosional episode whichmarks the Unit 3 / 2 boundary, the hydrologicalregime returned to an aggradational episoderather than one of stability. This suggests thateither climatic change or human-induced ero-sion upstream might have been involved. Thedeposits are typical of fan sediments in semi-arid environments composed of many smallscour-and-fill structures. These deposits filledin the erosion channels which had previouslydissected Unit 3.This massive deposition in-dicates that the sediment yield had increased.Such a depositional situat ion could be caused

8/7/2019 Ant 0740611

7/13

PALAEOENVIRONMENTS AND ECONOMY OF IRON AGE SOUTHEA STERN KAZAKHSTAN 61 7by several variables working in concert. Onefactor could be a decrease in vegetation upstreamdue to drier climatic conditions, human defor-estation, or both. Another important factor mayhave been a change in rainfall regime from themore stable situat ion of gentle rains evenlydistributed throughout the year to more ero-sive conditions of heavy rainfall events at fewerintervals in a given year. The latter situation isoften associated with an overall decrease inannual rainfall and a more arid climate.Unit 2 is stratigraphically dated to the Me-dieval period, being bracketed by the Saka-Wusun ceramic dates of Unit 3 below, and aradiocarbon date of c. 1500AD (calibrated) fromUnit 1 at GS-VII. Additionally, Medieval pe-riod (c. loth-12th-century A D ) archaeologicalpits appear in Unit 2 on the plain to the southof Medieval Talgar City (Rosen 1997a). A pre-liminary assessment of the early Medieval en-vironmental situation near Talgar is a drierclimate with a less stable rainfall regime thanin the Saka-Wusun period. Landscape degra-dation might have been exacerbated by defor-estation on the hillslopes and overgrazing byflocks belonging to the urban residents of thearea. This would seem to contradict evidencefrom the Balkhash Lake cited by Krementski(1997) [see TABLE),but different environmentalvariables control lake levels than alluvial fansediments.

Unit 2 grades gradually into Unit 1,whichis capped by Soil 1 . The radiocarbon-datedhearth from Unit 1 in GS-VII is 1447-1634 calAD (RT-2333, 380+90 BP). This would place itdirectly into the time-period of the Little IceAge event (with a cooler environment prevail-ing), and with Soil 1 signifying a return to amore stable rainfall and vegetation regime anda cessation of aggradation. The modern erosionalregime has led to renewed channelization ofthe fan. This, however, might be due to the di-version of much of the river water for wide-spread irrigation, thus artificially lowering thebase flow of the streams.To summarize, geoarchaeological evidencefrom the Talgar fan suggests that the buildupof alluvial silts during the earlier part of theSaka-Wusun period created an attractive localityfor pastoral and agricultural exploitation. How-ever, cool climates and short growing seasonswould have limited yields of wheat and en-couraged the cult ivation of millet by semi-no-

GS a1 l - - - - - i

loo

@ OIL A-HORIZONS l l T

S A N D

GRAVEL

UNIT 1

UNIT 2

UNIT 3

UNIT- 4

S N A I L S

GCE Y

SHERDS

FIGURE. Schematic geological section drawing ofGS - V I .madic peoples in the region. In the later phasesof the Saka-Wusun occupation, warming cli-mate and the subsequent development of sta-ble land surfaces with a rich topsoil allowedmore intensive cultivation of cereals.Phytolith evidence for agricultural activitiesof the Saka-Wusun communitiesPrevious evidence for agriculture among Saka-Wusun period peoples has come from finds ofiron agricultural implements and small amountsof grain, primarily wheat, within burial kurgans(Debaine-Francfort 1989).These finds afford usonly a limited view of the diversity of Saka-

8/7/2019 Ant 0740611

8/13

618

lab no.

Ty -95-2Ty-95-3Ty-95-4Ty-95-5Ty-95-6Ty-95 7Ty-95-8Ty -95 9Ty -95 15-95-11Ty 9 1Ty-95-13Ty-95-4Ty-95-15Ty 9 1Ty -95 1Ty-95-185-96-5-96-3Ty-96-7Ty-96-14Ty-96-15Ty-96-16

ARLENE MILLER ROSEN , CLAUDIA CHAN G & FEDOR PAVLOVICH GRIGORIEV

B 6 7v5 135A3 135BQ 60v9 65v9 32A4 225

G9IV8 -V8 80-100B6 100- 15- -- -- -- -- -- -

10-D 60-100D-10 100-120D-10 100-1 2012-G 120-34012-G 120-140

145 c m o bottom

quadrat depth cm occupation context TABLE.surface Archaeologicalproveniences off rom Tuzusai .2 lens of ash in bricky fill phytoli th samples

FIGURE (below).Percentagefrequencies ofpanicoid versus pooidgrass phyto l i ths . T hesamples ore arrangedin generalchronological orderfrom o ldest (left) toyo u n g es t (right).

occ. 435 ,30

25

20%15

10

5

0

3411141232333333444444

ashy middenashy middenash from floorash from floorash from pit basebase of pitpit filldebris on f loorpit fillhearth fillpit #17 illpit #18 illpit #17 illpit #18 illpit #17 illpit #18 fillpit #24afill from pit #24bpit #24bpit #27a;west halfpit #28bpit #29

Tuzusai 199511996o c c . 3

- 1- -- . I;IITotal Panicoid L. ._I.Total Pooid I

w m . -

o c c . 2 occ . 1

- - i

Sample No.

8/7/2019 Ant 0740611

9/13

PALAEOENVIRONMENTS AND ECONOMY OF IRON AGE SOUTHEASTERN KAZAKHSTAN 61 9

Tuzusai 199541996occ.4

0.90.8

0.7

0.60 0.52

0.4

0.3

0.2

0.1

0

Wusun agriculture. During the 1995 and 1996seasons at Tuzusai, samples were collected fromliving floors, pits and hearths from within thefour occupation phases at the site (TABLE )(Rosen 1996; 1997b) (see Rosen 2000 for a de-tailed descrip tion of phytolith methodology).The results of these analyses testify to agricul-tural activities and imply variations in the in-tensity of agricultural pursuits throughout theperiod of site occupation.A general summary of trends suggests a strongdominance of panicoid grasses in the earliestoccupation period (OP-4)(FIGURE). This grasssub-family includes the cultivated millets aswell as weed grasses which favour warm dryenvironments. There is a much larger contri-bution of weed grasses to the assemblage of OP-4 than in later periods (FIGURE). Cereals wereidentified according to methods outlined inRosen 1992. Although the cereal componentof OP-4 included both millet (Setaria sp.) andwheat (Triticumsp.), he percentage of millet

was significantly higher in this level than inOccupation periods 3 and 2 (FIGURE). Thedominance of millet is consistent with, althoughnot limited to, a semi-nomadic settlement pat-tern under cool climatic conditions since mil-let is produced in a short growing season. Wheatremained relatively constant in all periods witha slight increase in later phases.

FromOP-4to OP-3there was a shift in phytolithpercentages which may be indicative of a chang-ing focus in the agro-pastoral economies at thesite. In OP-3 there was a sharp decrease in theweed grasdcercal ratio (FIGURE) indicating alarge increase in cereals at the expense of weedgrasses. This trend continued also in OP-2. Per-centages of silica skeletons from wheat (Triticumsp.) increased and millet (Setaria sp.) declinedin importance (FIGIJRE). This could indicate alarger investment in cultivation.

Another notable observation was the chang-ing relationship between weed grasses and ce-reals (FIGURE). In OP-4 there was no correlation

8/7/2019 Ant 0740611

10/13

62 0 ARLENE MILLER ROS EN, CLAUDIA CHAN G & FEDOR PAVLOVICH GRIGORIEV

UE0QQv,Q

2neh

Tuzusai I99511996

20 1

16

14

12

10

a6

4

2

0

. . . . . . . . . .. . . . . . . .. . . . . .. . . .. . . . . .. . . . . .. . . . .. . . . . .. . . . . . .. . . . .. . . . . .11 . . . . . . . . . . .. . . . . . . . .. . . . .. . .. . . . .. . . . . .. . . . . .. . . . .. _ . . _. . . . .. . . . . .. . . . . .. . . . . .. . . . . .

. ~ . .............CJAverage % MilletElAverage% Rice. . . .. . . . . .. . . . . .

4 3 2 1Occupation PeriodFIGURE . Average percentages of si l ica skeletons from cereals by occupa t ion phase .between the two, strongly suggesting that thesources of cereal and weed grass phytolithsdiffered. Thus weed grasses came to the sitefrom different locations from the cereals andat different rates, perhaps indicating again thatthe weed grasses were from the dung of ani-mals grazed outside agricultural areas. In OP-3 this relationship began to change and in OP-2and OP-1 there was a distinct coincidence be-tween the presence of cereals and the occurrenceof weed grasses. These relationships imply thatthe weed grasses came to the site as field weedsor as dung from animals grazed on agriculturalfields. This can be further evidence for an in-crease in agricultural intensification and suggeststhat herds were kept close to the settlement.

Several of the silica skeletons were from thehusks of cultivated rice (Oryza sa t iva) (FIGURE8).Other phytolith forms (fan-shaped keystones)(FIGURE) were derived from the leaves andstems of the rice plant (Pearsallet al. 1995; Zhao

et a l . 1998).The presence of rice is rare in IronAge sites of this region and poses the interest-ing question of how the rice was obtained,through trade or on-site cultivation. The per-centages of these husk phytoliths were low inthe earliest OP-4 phase, and greatly increasedin OP-3 and OP-2 with a drop-off again in OP-1 (FIGURE). This was the opposite distribu-tion from that of millet. In the last phases ofoccupation in OP-1, millet increased in impor-tance again.

Throughout the entire occupation of the sitethere was a notable quantity of phytoliths fromsedges (Cyperaceae) and horse-tail rushes(Equ i se tum) ndicating the presence of marshyareas and high water tables. This suggests agenerally moist climatic regime throughout theduration of occupat ion.Given the environmental conditions of theSaka-Wusun period discussed above, it mighthave been possible for the occupants of Tuzusai

8/7/2019 Ant 0740611

11/13

PALAEOENVIRONMENTS AND ECONOMY OF IRON AGE SOUTHEA STERN KAZAKHSTAN 62 1Tuzusai 1995/1996

Sample No.FIGURE. Graph showing the relationship between phytoliths from weed grasses a n d those fromcultivated whent.to produce a rice crop on the alluvial fan un-der conditions that were moister than those ofthe present. However, the Talgar fan settlementsare relatively close to the Fergana valley wherecontemporary Kushan towns had sophisticatedirrigation systems and were themselves produc-ing rice among other crops (Mukhamedjanov1994: 274). Given the political connection ofthe Wusun peoples with the Yueh Chih andother peoples to the west in the 1st and 2ndcenturies B C , it is also possible that rice couldhave been obtained from there.Generally, the trends in the phytolith datasuggest an agro-pastoral adaptation in OP-4 witha shift to a more intensive agricultural focuswithin the two middle occupation periods, OP-3 and OP-2. The latest occupation (OP-1) ap-pears to represent a period of somewhat lessintensive agricultural activity.

DiscussionWith the above archaeological, environmentaland agricultural information we can suggest ascenario for Saka-Wusun settlement andeconomy. In the Early Iron Age, Saka tribesmoved into the area from the north (Yablonski1995b) at a time corresponding to a cool-moistSubboreal period. The first settlers at Tuzusaihad an agro-pastoral economy in which ani-mals were grazed at least part of the year onthe panicoid steppe grasses to the north ofTuzusai, and agriculture was characterized bya dominance of millet.

Due to a short growth period, millet is fa-voured by semi-sedentary pastoralists becausethe crop can be planted and harvested whilemaintaining a semi-sedentary seasonal round.This is especially significant given the shortlength of the summer growing season and in-

8/7/2019 Ant 0740611

12/13

6 2 2 ARLENE MILLEK ROSEN, CLAUDIA CHANG & FEDOR PAVLOVICH GRIGOKIEV

crease in frost-days during the Subboreal pe-riod. Millet would have been better adapted tothis environmental regime than wheat.

In the vicinity of Tuzusai the early Iron Agelandscape was an actively aggrading alluvialfan. The alluvial silt originating from the out-wash of extended alpine glaciers would haveprovided a rich agricultural environment withhigh water tables, aggrading streams, and freshsilts deposited in moist fields each spring.The middle and later phases of occupationat Tuzusai coincided with the onset of thewarmer Subatlantic period. The agricultural

FIGURE.Photomicrograph of apossible rice-husksilica skeleto n fr omTuzusai .Scale bar = 20 m m .

FIGURE.Photomicrograph of afan-shaped bu l li formphyto l i th f romTuzusai which isprobably derivedfr om the leaf of a ricep la n t .Scale = 10 mm.

growing season would have been longer thanthat of the preceding Subboreal phase. Therewas a cessation of alluviation on the fan, a sta-bility in the landscape and the beginning ofsoil formation as indicated by Soil 2.

Historically, it was an interval in whichWusun peoples increased their presence in theregion due to political and economic pressurefrom newly consolidating tribes of the Xiongnuto the east and alliances with the Yueh Chih tothe west (Enokie ta l . 1994; Zadneprovskiy 1994;Ishjamts 1994).There were elite segments whichcommanded large armies (Ishjamts 1994) and

8/7/2019 Ant 0740611

13/13

PA L A E O E N V I R O N M E N T S A N D E C ON O M Y OF I R O N A G E SO U T H E A ST E R N K A Z A K H ST A N 6 2 3i t follows that they would find it necessary tohave some measure of control over agriculturalproduction and surplus. The emergence of com-plex nomadic confederacies might have providedthe 'push' factor leading to the intensification ofthe agricultural sector at the same time that en-vironmental amelioration offered a 'pull' factor.

The phytolith remains from Tuzusai appearto represent a small but significant shift from amore pastoral orientation to a more intensiveagricultural focus, aided by both environmen-tal and historical factors.ReferencesCHANG, . & P.A. TOURTELLOTTE.998. Th e role of agro-pasto-ralism in the evolution of steppe cu lture in the Semirech ye

area of southern Kazakhstan dur ing the Saka /Wu sun pe-riod (600 BC-AD 400), in V. Meir, A.H. Zhim in & E. Kuzmina(ed.) , Bronze an d ear ly Iron Age archaeology of easternCentral Asia: 264-79. Phi lade lphi a (PA): University ofPennsy lvan ia Museum.DAVIS-KIMBALL,., V.A. BASHILOVL.T. YABL ONSK Yed.) . 1995.Nomads of the Eurasian steppes in the early Iron Age.Berkeley (CAI: Zinat.DEBAINE-FRANCPORT,. 1988. Archkologie d u Xinjiang desor igines aux Han: Premiere par t ie, Paleorient 14: 5-29.1989. ArchBologie du X injiang des or igines aux Han: Hemepar t ie , Palkorierit 15: 183-213.D1 COSMO,N. 1994 . Ancien t Inner As ian nomads : Thei r eco-nom ic basis and i ts s ignif icance in Chinese H istory, Th e

Journal of Asian Studies 53: 1092-1126.ENVIRC. 1995. Characterization of ecosystems of archaeologicalexcavations area in Talgar River basin. Almaty: Environ-ment Research Centre.HARMATLA,. (ed.) 1994. The development of sedentary andnom adic civilization. Paris: UNESCO.ISHJAMTS,. 1994. Nomads in eastern Central Asia, in Harmatta(ed. ): 151-69.KHAZANOV,A.M. 1975. Sotsial'naya istoriya skifov[Social historyof the Scythians). Moscow.1984. Nomads and the outside world. Cambridge: CambridgeIJniversity Press.KHOI ' INSKIY,N.A. 1984 . Holocene vegeta t ion h is to ry , in A .Velichko, H.E. Wright & C.W. Barnosky (ed .) ,Late Qua-ternary environments of th e Soviet Union (English edi -tion): 179-2130. Minneapolis (MN):University ofMinne sotaPress.KHKUSTALEV,Y.P. & S.Y. CHEKNOUSO V.992 . On the Holocenehistory of the Balkhash la ke, Izvestia of Russian Geogrophi-cal Society 124 : 164-71.KREMENTSKI, .V. 1997. Th e Late Holocene e nvironm ental an dcl imate sh i ft in Russ ia a nd sur rounding lan ds , in H .N.Dalfes, G. Kukla & H. Weiss ( ed . ) , Third millennium BCclimate change ond Old World collapse: 351-70. Ber lin:Springer-Verlag.LATI'IMORE,.1 97 9. Herders, farmers, urban culture, in L'EquipeBcologie et anthropologie des societbs pastorales (ed.),Pastoral production and society Proceedings of th e in-ternational meeting on nomadic pastoralism, Paris, 1-3

Acknowledgements. W e w o u l d l i k e t o t h a n k K a r l M.B ai p ak o v , D i r ec t o r o f t h e I n s t i t u t e of A r ch aeo l o g y , K azak hN a t i o n a l A c a d e m y of Sc i en ces , M i n i s t r y of S c i e n c e , a n dT am a r a V. Sav e l i ev a , Sen i o r R es ea r ch e r , I n s t i t u t e of Ar-ch aeo lo g y , A l m a t y , K azak h s t an f o r f ac i l i t a t in g o u r r e s ea r chi n t h e f i e ld . P h y t o l i th a n a l y s e s w e r e c o n d u c t e d w i t h t h eas s i s t an ce of g r an t s f r o m t h e I r en e L ev i Sa l a C A R E A r ch aeo -l o g i c a l F o u n d a t i o n a n d t h e W e i z m a n n I n s t . C e n te r for Co-o p e r a t io n b e t w e e n N a t u r a l S c i e n c e s a n d A r ch a e ol o gy . T h ep h y t o l i t h a n d g e o a rc h a e o l o g i c al r e s e a r c h w a s p a r t i a l l yf u n d e d b y g r a n t s f r o m t h e N a t i o n a l G e o g r a p h i c S o c ie t y a n dt h e N a t i o n a l S c i e n c e F o u n d a t i o n . N e i l R o b er ts a n d S t e v eR o s e n k i n d l y c o n t r i b u t e d e d i t o r ia l c o m m e n t s o n a n e a r -l i e r d r a f t of t h i s p ap e r .

December 1976: 479-90. Cam bridge: Cam bridge Univer-si ty Press .M I J K H A M E U J A N O V ,.R. 1994. Economy and social system inCentral Asia in th e Kushan Age, in Harmatta (ed.) : 265-

90.PEARSALL,.M., D.R. PIPERN O,.H. DINAN,M. UMLAUP',. ZHAO& R.A.J. BENPER. 995. Dist inguishing r ice (Oryza sativaPoaceae) from wild Oryza species through phytolith analy-sis: Results of prel iminary r esearch , Economic Botany49: 183-96.ROSEN,A.M. 199 2 . Pre l iminary iden t i f ica t ion of silica skel-etons from Near Eastern archaeological sites: an anatom icalapproach , in G.J. Rapp & S.C. Mulholland, Phytolith sys-tematics: emerging issu es: 129-47. New York (NY): le -n u m .1996. Phyto li th analysis at Tuzu sai , Kazakhstan, 1995 sea-son . unpubl i shed r epor t on f i le in Depar tment of A n-thropology, Sweet Briar College, Sweet Briar (VA).1997a. Geoarchaeology of Mediev al Talgar City, Kazak hstan.Unpu blished repor t on f ile in Depar tment of Anthropo l -ogy, Sweet Briar College, Sweet Briar (VA).1997b. Phytoli th analys is at Tuzusai, Kazakhstan 1996 sea-son . Unpubl i shed r epor t on f i le in Depar tment o f An-thropology, Sweet Br iar College, Sweet B r iar (VA).

2000. Phyto li th analysis in Near Eastern Archaeology, in S.Pike & S. Gitin (ed.) , Th e practical impac t of science onAegean and Near Eastern archaeo logy 86-92. London:Archetype Press .SAVOSKUL,, s .& O.N. SOLOMINA.996. Late-Holocene gla-cier var iat ions i n the f rontal an d inner ranges of the TianSha n . Cen t ra l As ia , The Holocene 6: 25-35.VENUS, .G. 1985. Features of the lake basin evolution in thehum id and arid belts: Palaeolimnologyin the humid andarid belts. Moscow: Nauka.YABLONSKY, .T. 1995a. Some ethnogenetical hypotheses, inDavis-Kimball et al. (ed.): 241-52.

1995b. The mater ial culture of the Saka and histor ical re-constru ction, in D avis-Kimball et al. (ed. ): 201-40.ZADNEPROVSKIY,.A. 1994. The nomads of nor thern CentralAsia af ter the invasion of Alexander , in Harmat ta ( e d . ) , :457-72.Distinguishing rice (Oryzasativa Poaceae) from wild Oryzaspecie s through ph ytoli th analysis, 11: Finalized method,Economic Bo tany52 : 134-45.

ZHAO,Z., D. PEARSALL,.A.J. BENFER& D.R. P IPERN O. 998.