New RadiocaRboN dates foR the cRiş site of sacaRovca i (Moldova)

Kiosak D., Dergaciov V.A., Szidat S. and Tinner W. 2023. New Radiocarbon Dates for the Criş Site of Sacarovca I (Moldova). Sprawozdania Archeologiczne 75/1, 175-182. Four new AMS radiocarbon dates shed new light on the chronology of one of the easternmost sites of the Criş culture. The conventional dating efforts had yielded indecisive results, while the new results correspond well to the typo-chronological position of the site (Criş IV) and the chronology of other sites with similar finds. The comparison with the nearby para-Neolithic sites demonstrated that the establishment of the para-Neolithic way of life (foragers equipped with pottery) in the region happened several centuries before the spread of early farmers of the Criş culture into Moldova.


InTroDuCTIon
The current consensus for the initial agricultural settlement of Moldova and southwest Ukraine follows the broader, regional model for south-east Europe (Dergaciov and Larina 2015;Larina 1994).It is mostly seen as a phenomenon of the propagation of societies of a particular type that first appeared in Greek Thessaly and later spread out in temporally decreasing steps (Biagi et al. 2005;Whittle 1996).This process was brought to the territory of modern-day Moldova by the people of Criş culture in sixth mill.BCE (Dergaciov and Larina 2015).However, its exact chronology remains unclear in detail.
Sacarovca1 is the reference site for the Criş culture between the Prut and Dniester rivers (Fig. 1: A).At present, it is the only excavated site located at the easternmost fringe of the extension of the Criş culture, which was studied by complex investigation with an application of scientific approaches, thanks to the effort of Olga Larina and Valentin Dergaciov (Dergaciov and Larina 2015).
This paper aims to publish four new AMS radiocarbon dates for the site of Sacarovca I and to shed new light on the chronology of the early Neolithic in Moldova.

MeThoD AnD SAMplIng
The samples were dated in the Laboratory for the Analysis of Radiocarbon with AMS (LARA) at the University of Bern, employing the MICADAS equipment (Szidat et al. 2014).Collagen extraction was performed according to Szidat et al. (2017), which was extended by an additional ultrafiltration step.Radiocarbon dates were calibrated using the online calibration program OxCal 4.4.4(Bronk Ramsey and Lee 2013) using atmospheric data from Reimer et al. (2020).
Samples comprise four Cervus elaphus bones (identifications by A. David and O.P. Siekerska, with thanks): two fragments of metacarpi, a chunk of a femur and a piece of an unidentified long bone.The selected samples come from three features of the site (Fig. 1: B): Pits 21 (1 date), 44 (2 dates), and 46 (1 date).These pits yielded abundant lithic and ceramic assemblages alongside notable archaeozoological and palaeobotanical collections.

reSulTS
Three novel dates (Be-16910, Be-16911, Be-18271) form a consistent group encompassing 5617-5479 calBCE, 2σ, while a single date (Be-16192, 5481-5373 calBCE, 2σ) is slightly later (Table 1, Fig. 2).In order to find out whether certain dates are statistically simultaneous, we used the R_Combine function of OxCal.If they could be combined (the Х 2 meets a certain threshold), we can say that the group of dates is statistically concurrent.It is in this sense that we study the combinations of dates for Saсarovсa 1. Namely, the former three dates can be combined into the time-slot (5613-5482 calBCE, 2σ).The obtained results are consistent with the available radiocarbon dates for Sacarovca 1.The fact that the Berlin date can be successfully combined both with the three earlier AMS dates as well as with the latest date (while these four dates cannot be combined when treated as a group of their own) underlines the increased precision of AMS dating (saying nothing about the Kyiv laboratory's "direct" date on a potsherd, which calibrates to encompass at least 600 years).Another charcoal date (Ki-13899) is a bit earlier than the rest of the dates.It can be explained by a likely "old wood" effect.Thus, it is possible that complexes of Sacarovca 1 were inhabited for a prolonged period.Namely, pit 46 is slightly younger than objects 21 and 44.

DISCuSSIon
The chronology of the Starčevo-Körös-Criş cultural complex is defined by over 400 relevant dates.Early work in the inner Balkans put initial settlement activities by agricultural colonists associated with this cultural complex at ca. 6200 y.BCE, but a recent reanalysis showed this event likely occurred no earlier than ca.6050 y.BCE (Krauss 2016, 212).This event is viewed as a rapid initial settlement represented by a homogenous material culture across the region (Biagi et al. 2005).The territory of Moldova was settled by these early farmers relatively late in the course of their expansion, namely during the III-IV stages of the Criş culture (Dergachev and Dolukhanov 2007;Larina 1994).These oc- For captions and references: see Table 2 cupations and their associated material culture had first been labelled the "Buh-Dniester Neolithic" (Markevich 1974;Yanushevich 1989), but then their affinities with actual Criş were extensively demonstrated (Larina 1994).
The final Criş sites are poorly represented in the radiocarbon dataset.The novel dates, when compared to the existing data, show that Sacarovca 1 is neither the latest nor it is exceptionally early.It fits nicely in the designated time-slot for the late Criş culture (Fig. 2).Taking into account that the Starčevo-Körös-Criş cultural complex is unlikely to have survived much longer than 5400 BCE (Meadows 2019), the chronology of the Sacarovca 1 complexes seems reasonable and expected from a general historical view of the development of early farming communities in the region.Another essential issue to consider is the comparative chronology of the easternmost Criş sites and the sites of local foragers equipped with pottery (the "Buh-Dniester" para-Neolithic).The proponents of the Balkan impulse for the Neolithization of Ukraine insisted that the "Buh-Dniester culture" sites had arisen as a "barbaric periphery" of the Criş culture during its III and IV stages (Zaliznyak 1998).In particular, ceramics of so-called Pechera style were thought to emerge under the influence of the Criş ceramic technology.Early work on the direct dating of potsherds seemed to indicate the appearance of the ceramic vessels in the forest-steppe Eurasian belt by the mid-7th mill.BCE (Zaitseva et al. 2009).However, it was mostly based on totalling the organic content of a sampled potsherd (TOCC approach; Meadows 2020) for "direct" radiocarbon dating.Averaging carbon content in organic remains of different origins can be highly misleading.Recent redating programs moved the first appearance of pottery in the forest-steppe of east Europe into 6th mill.BCE (Courel et al. 2021).In the light of a revision of the Rakushechny Yar site sequence (Dolbunova et al. 2020), the sites in Moldova appeared to be the crucial link between early potters of Eastern Europe and the ceramists from the Balkans.
The new dates of Sacarovca 1 firmly put its existence in the 57-55 th centuries BCE.It is partially later than a single conventional date on charcoal for the Soroca-II para-Neolithic site, definitely later than radiocarbon dates obtained for the para-Neolithic stratigraphic unit of Melnychna Krucha (Kiosak et al. 2021) in the Southern Buh river valley, some 200 km to the east (Fig. 2).The "direct" dates on "Buh-Dniester" potsherds yielded divergent results.The only two consistent dates (from the lower layer of Gard, Ki-14790 and Ki-14789; Fig. 2) encompass 5719-5232 calBCE.However, they are in reverse stratigraphic order with the dates obtained for the upper layer of the same site (Tovkailo 2014).The "direct" dating of Criş import in the Buh-Dniester site of Hlynske I yielded unacceptably old results (TKA-21090; Haskevych et al. 2019), while dating of another potsherd with some analogies in Criş materials from Bazkiv Ostriv (TKA-20831, 6625±25 BP) returned a result comparable with the chronology of Sacarovca 1 -5622-5483 calBCE, 2σ, although this potsherd was attributed not to the Pechera style but to the earlier, Skybyntsi style of para-Neolithic pottery (Haskevych et al. 2019).
Thus, we can conclude that para-Neolithic sites in the valleys of the Southern Buh and Dniester were settled prior to the expansion of early farmers of the Criş culture during its stages III and IV into the territory between Dniester and Prut.Probably, there were also para-Neolithic sites, which could be contemporaneous with this colonisation event.This chronological framework questions the commonly accepted interpretation of the origin of Pechera-style pottery, namely its origin under the influence of Criş culture potters.Due to their high standard deviations, conventional dates can be highly misleading when solving the issue of the contemporaneity of two samples.The higher precision of AMS dates is needed in order to fine-tune the chronological picture of the first half of 6 th mill.BCE in the North-West Pontic region.
fig. 1. A: location of Sacarovca 1 on the map of europe, topo: natural earth; B -sampled objects on the plan of Sacarovca 1 (after Dergaciov and larina 2015, fig.11 with modifications)

fig. 2 .
fig. 2. Comparison of selected dates for Cris III-IV and Buh-Dniester para-neolithic with novel dates.For captions and references: see Table2

table 1 .
novel radiocarbon dates for the site of Sacarovca 1