Introduction

The first recognition of ‘new glume wheat’ (NGW) as a distinct prehistoric crop was made based on the archaeobotanical finds from Neolithic and Bronze Age sites in northern Greece (Jones et al. 2000). This was soon followed by the identification of this cereal at Bronze Age Stillfried, Austria (Kohler-Schneider 2003). Subsequently, through new discoveries or revision of previously ambiguous identifications of wheat remains, its presence has been confirmed at a series of prehistoric sites across Europe. However, the taxonomic identity of this wheat remained unresolved.

Several recent publications placed the ‘new glume wheat’ under a new spotlight as they applied advanced methods such as detailed morphological, geometric morphometric and ancient DNA analyses of archaeological specimens, to determine its taxonomic identity and its domestication location and status (Czajkowska et al. 2020; Charles et al. 2021; Badaeva et al. 2022; Roushannafas et al. 2022). They demonstrated that NGW was a tetraploid member of the Triticum timopheevii group, which entered domestication in southwest Asia from the mid/late 9th millennium bce. In comparison to T. timopheevii sensu stricto, the wheat traditionally grown in western Georgia as a component of the Zanduri landrace mixture which is geographically restricted to this region (Jones et al. 2000; Bedoshvili et al. 2021), the ancient T. timopheevii sensu lato (Timopheev’s wheat) had a much wider distribution. This is indicated by recent evidence from western Asia (Bogaard et al. 2017, 2021; Ergun et al. 2018; Ulaş and Fiorentino 2021) and reviews for Europe (Kenéz et al. 2014; Toulemonde et al. 2015).

Now that the identity of Timopheev’s wheat has been resolved, it is necessary to examine its geographical and temporal spread in greater detail, as a means to understand its biogeography, cultivation and uses. The (marked) presence of T. timopheevii s.l. at Neolithic sites in western Asia, together with T. monococcum (einkorn) and T. dicoccum (emmer), demonstrates its status as one of the ‘founder crops’. It was shown that during its long cultivation history at Çatalhöyük, central Anatolia, Timopheev’s wheat was undergoing domestication (Charles et al. 2021; Roushannafas et al. 2022). Ulaş and Fiorentino (2021) suggest that, in parts of western Asia and Thrace, it was becoming adapted to local growing conditions and that, thanks to its ability to withstand cold conditions, it successfully spread across prehistoric Europe.

This paper assembles the existing data on the presence of this crop in selected parts of southern and eastern Europe, in the territories of the modern states of Ukraine, Moldova, Romania, Bulgaria, Greece, North Macedonia, Albania, Kosovo (European Commission 2023), Serbia, Montenegro, Bosnia and Herzegovina, Croatia and Hungary. Much of this region was less known archaeobotanically up to about a decade ago, and some parts of it even until recently. New archaeological investigations have benefited from archaeobotanical field and laboratory work, most of which was carried out by the present authors and has offered a wealth of new data on crop growing and consumption in prehistory. We have taken advantage of our direct access to the recently retrieved study materials, as well as some of those which were previously analysed (and published) and which we have now been (re-)examining for the presence of Timopheev’s wheat.

This paper provides a list of sites with published and unpublished records of Timopheev’s wheat along with, where possible, absolute counts or estimates of the quantity of finds of this wheat per site. In some cases, the analysis is still in progress. As far as the current data allow, we explore spatial, temporal and quantitative patterns in the occurrence of T. timopheevii s.l. and compare them with those of einkorn and emmer. The AMS 14C dates obtained on a selection of the finds have confirmed their expected ages according to the archaeological chronologies of the sites and show that the early finds are indeed early and not intrusions from later occupation layers. They also allow us to estimate the time of the start and possible end of the cultivation of this wheat in our study region. Parts of this area represent the earliest and longest enduring farming zones in Europe and therefore offer an opportunity to take a long-term perspective on the presence of Timopheev’s wheat in places where it first appeared on the continent. This overview aims to track its presence through time in these and other parts of the region where we conducted new analyses or re-examination.

Materials and methods

We compiled all records of T. timopheevii s.l. grain and chaff (glume bases and spikelet forks) in the study region that are known to us to date; the majority have been identified by us and they derive from 89 site phases (ESM 1, 2). Some of the records have already been published, often with details of the quantities found and archaeological contexts. Others come from the assemblages that are currently being analysed and so full information is not yet available. Examples of the assemblages considered here include those recovered in recent or ongoing large, long-running, systematic projects (EUROFARM, PlantCult, EXPLO and CRC 1266), as well as those from older excavations that have now been checked and the finds of Timopheev’s wheat in them confirmed, such as several collections from Bulgaria and Serbia. We note that, ideally, all assemblages from old excavations should undergo re-examination using the identification criteria specified in the publications presenting the first discoveries of Timopheev’s wheat (Jones et al. 2000; Kohler-Schneider 2003) and those that followed. Our paper reflects the varied pace of progress in this direction in our study region. Of note is that, in the re-assessments done so far, remains of Timopheev’s wheat have often been encountered among remains initially identified as ‘einkorn/emmer’, ‘Triticum spelta (spelt)’ or ‘emmer’.

Charring is the main form of archaeobotanical preservation for the study region, but mineralised and waterlogged remains also occur, such as from lake shore sites in North Macedonia; we considered all of these types of preservation. All of the sites with Timopheev’s wheat which are known for the region benefited from a careful analytical procedure. They were all archaeobotanically sampled; the samples were processed by bucket or machine flotation, using sieves with small mesh sizes (down to 0.5−0.25 mm). The information on the numbers of analysed samples or contexts and the soil sample volumes helps to distinguish between sites which are ‘poor’ in Timopheev’s wheat, those with a high number of samples or large soil volume processed and very few finds of this wheat, as opposed to sites ‘rich’ in Timopheev’s wheat, of which some yielded large, dense deposits of it.

For some of the sites, only presence/absence information is currently available. For a subset of sites with detailed quantitative data, we considered the amounts of einkorn, emmer and Timopheev’s wheat. These represent the most common glume wheats found at prehistoric sites in the region; they required the same type of processing for consumption and, whether grown individually or in some combinations, they may have been processed together. To gain a first impression of how prominent Timopheev’s wheat is among the three glume wheats, we calculated their relative proportions in the site level assemblages containing a minimum of 30 remains of this wheat. In the future, it will be important to do this for individual archaeological contexts, notwithstanding various taphonomic factors, and use it to make inferences on the status of each glume wheat, its combined or separate cropping, storage, processing and discard.

Recognising the remains of Timopheev’s wheat is challenging, particularly when dealing with small numbers of specimens, or if they are poorly preserved. Some of the remains we could identify only tentatively (‘cf.’), especially in the case of grains, and these determinations should be treated with caution. As was the case in the initial study (Jones et al. 2000), we also find the chaff of Timopheev’s wheat easier to differentiate from that of other hulled wheats, unlike the grains. Therefore, for some of the assemblages included here, the initial identifications of grain as spelt, emmer or two-grained einkorn are yet to be fully evaluated. In this overview, we consider only the remains that we are confident as (most likely) originating from T. timopheevii s.l. Some of the absolute counts presented here may change in the future, as the (re-) analyses continue.

The attribution of sites to archaeological periods, phases and cultures follows relevant archaeological information published or otherwise available to the archaeobotanists who studied these assemblages. The start/end and duration of the identically or similarly named periods and phases inevitably lack synchrony within the large region covered here. This is because the cultural developments identifying them, such as the emergence of agriculture (Gronenborn et al. 2021), took place at different times across the region; other reasons are differences in the local histories of research and in archaeological interpretations. These are often confined to modern administrative territories and we therefore present and discuss the results within subregions or states and cultural periods/phases as they are defined locally. The remains from which we obtained radiocarbon dates were selected according to their immediate availability and their attribution to ‘early’ (Neolithic) or ‘late’ (late Iron Age) periods according to the archaeological chronology. Additionally, dates on the remains from two sites (Feudvar and Stolniceni) result from general archaeological programmes of radiocarbon dating of these sites.

In order to consider the finds of Timopheev’s wheat identified in our study region within their broader geographical context, we also compiled its occurrences in other parts of Europe and in Asia, on the basis of archaeobotanical reports accessible to us at the time of writing.

Results and discussion

Temporal and spatial distribution

The current evidence shows the presence of T. timopheevii s.l. across central and western Asia and large parts of Europe (Fig. 1). Its occurrence in different biogeographical zones and a range of local ecologies suggests that it could tolerate varied growing conditions, though the extent of its tolerance may have changed between different environments. The growing of Timopheev’s wheat in prehistory seems to have been a widespread practice and one that persisted through millennia, along with the cultivation of a range of other crops.

Fig. 1
figure 1

Map of sites with currently known records of Timopheev’s wheat in Eurasia (attribution to periods based on the information in the source publication; see ESM 1 for site names and location details)

Full size image

In our study region, there are numerous sites with finds of this wheat, altogether 89 site phases known so far (Fig. 2, ESM 1, 2). The earliest finds originate from the early Neolithic and the latest are from the Hellenistic and La Tène periods, demonstrating that T. timopheevii s.l. was present in the area for nearly six millennia, from the late 7th to the late 1st millennia bce. Much of the evidence comes from the earlier part of this time span, principally Neolithic sites, partly because of the greater research focus on the earliest farming in the region. As noted above, archaeological chronologies and terminology differ between parts of our study region; therefore, we rely on rough time spans attributed to major periods in very broad subregions (Table 1). We use archaeological periods only as standardised markers of the cultural developments that define them and do not attempt a reconstruction of the chronology of occurrence of Timopheev’s wheat at the subregional level.

Fig. 2
figure 2

Map showing sites with finds of Timopheev’s wheat in our study region (attribution to periods based on the information in the source publication; see ESM 1 for site names and location details)

Full size image
Table 1 Approximate beginning and end of the four main periods of later prehistory in different parts of the study region (see ESM 1 for refined chronological attribution of the sites); Central and W Balkans: North Macedonia, Albania, Montenegro, Kosovo (European Commission 2023), Serbia, Bosnia and Herzegovina, Croatia; E Balkans: Bulgaria, W Romania; North Pontic region: E Romania, Moldova, Ukraine. All dates are bce
Full size table

New archaeobotanical investigations in northern Greece, where Timopheev’s wheat was first recognised, as well as in areas to the north and west of it (Albania, North Macedonia, Serbia), and the ongoing re-analysis of the material from sites in Bulgaria, demonstrate that it was one of the first cereal crops in the region, along with einkorn and emmer, and that it was ubiquitous (Fig. 2, ESM 1, 2). Either dated archaeological materials or direct radiocarbon dates place the first occurrences of T. timopheevii s.l. in the last two centuries of the 7th or the first several centuries of the 6th millennium bce, the Neolithic (Table 1; Fig. 3). New information on its temporal and spatial distribution adds to the growing evidence that there was high crop diversity from the very beginnings of agricultural activity in our study region (Valamoti and Kotsakis 2007; Bogaard and Walker 2011; Marinova and Valamoti 2014; Motuzaite Matuzeviciute 2020; de Vareilles et al. 2022; Sabanov et al. 2023).

Fig. 3
figure 3

AMS radiocarbon dates obtained directly on the remains of Timopheev’s wheat from the region (images of the dated specimens in ESM 3); calibrated with OxCal (Bronk Ramsey 2021) using IntCal20 (Reimer et al. 2020)

Full size image

At the subregional scale, a detailed and quantitative assessment of T. timopheevii s.l. finds through time has recently been completed for Neolithic Greece, charting the history of the growing and use of this wheat in the earliest farming societies in Europe (Valamoti et al. 2022a). We build on this by looking at Neolithic and later sites elsewhere in the region that offer more detailed information from already finalised and/or published analyses.

Compared to the Neolithic, there are fewer known finds of Timopheev’s wheat from later periods, the reason for which may lie in the lack of archaeobotanical (re-)analyses, certainly in the Balkans north of Greece and in the North Pontic area (eastern Romania, Moldova and Ukraine). As an example, it is only recently that Timopheev’s wheat has been detected at the Chalcolithic mega-sites of the Trypillia culture in Ukraine and Moldova (Kirleis et al. 2023). Nonetheless, the remains that have been recorded came from different parts of the study region, illustrating the wide distribution of this crop after the Neolithic (Fig. 2). The latest remains, dating from the 4th-2nd centuries bce, originate from the early Hellenistic town of Kabyle in Thrace (southeast Bulgaria) and a La Tène farmstead at the site of Kanjiža in Vojvodina (north Serbia).

To our knowledge, no traces of Timopheev’s wheat have been found at sites dated to later periods in our study region, to classical antiquity, medieval and modern times. However, the available archaeobotanical evidence from these periods is very limited compared to prehistoric finds, therefore future research might alter the current picture. The records of Timopheev’s wheat from elsewhere in Eurasia are also confined to prehistoric periods (Fig. 1, ESM 1, 2). In Georgia for instance, where cultivation of T. timopheevii s.str. as part of a glume wheat maslin was practised in recent history, there is just one record, found in the deposits dated to the middle to second half of the 1st millennium bce at the site of Atskuri (Bieniek and Licheli 2007). Our observations for the regions considered here should be compared with other parts of Europe.

Archaeological contexts and finds

All of the records we have compiled come from settlement sites, which are of different types (tells, flat-extended, lakeside settlements). The contexts of the finds include features and deposits such as pits, house floors, pottery, ovens and hearths, clay storage bins, cultural layers, ditches, or waste disposal areas. They indicate that the remains of Timopheev’s wheat are a result of domestic activities involved in crop storage, processing, consumption and discard. Most of the remains are fragments of chaff (glume bases or spikelet forks), but this does not necessarily mean that Timopheev’s wheat is chiefly represented by processing by-products. Whereas many sites do indeed contain only chaff remains of this and other wheats, for some sites these remains reflect the level of identification possible, given that the identification is more difficult for Timopheev’s grain than chaff; or the progress of re-examination of previously analysed assemblages, where normally chaff is the first to be checked for its presence. Short descriptions of the (unpublished) evidence of T. timopheevii s.l. from a subset of sites are given in ESM 4, serving as an illustration of the types of sites and contexts from which some of the finds considered here derived.

Cultivation status

The study that first recognised T. timopheevii s.l. in the archaeological record was also the first to recognise the potentially combined cultivation of Timopheev’s and einkorn wheats in some cases, based on their proportional representation in individual samples and contexts (Jones et al. 2000). Applied to some other sites and assemblages, this approach raised the possibility of the cultivation of Timopheev’s wheat as a (nearly) pure crop. This may have been the case at Neolithic Apsalos and Kyparissi in northern Greece, where it is the dominant component of certain samples (Valamoti et al. 2022a). There are also instances of rich concentrations of remains of this wheat from elsewhere in Eurasia, providing evidence of its cultivation as a dominant or pure crop. Some examples are Neolithic Çatalhöyük and Yenikapı in Turkey and the Bronze Age sites of Stillfried (Austria), Lucone (Italy) and Feudvar (Serbia) (Kohler-Schneider 2003; Perego 2015; Kroll 2016; Bogaard et al. 2017; Ulaş and Fiorentino 2021).

Sites that yielded hundreds or thousands of T. timopheevii s.l. remains, from either individual samples and contexts or overall site assemblages, suggest a marked degree of growing and use of this wheat at these sites in the respective periods. Many of the sites with Timopheev’s wheat in our region produced remains of other glume wheats too, mostly einkorn and emmer (Table 2, ESM 4). Here, remains of two or three of these glume wheats are frequently mixed in archaeological deposits. They point to a possibility that Timopheev’s wheat was a component of a ‘maslin’ crop composed of different glume wheats. In other cases, it could have been a contaminant, a tolerated or desired admixture in fields of other wheats.

Table 2 Abridged information on the representation of Timopheev’s wheat, einkorn and emmer (summary counts) for sites with some quantitative data
Full size table

Previous archaeobotanical studies used the relative proportions and frequencies of occurrence of glume wheats to explore the cultivation status of T. timopheevii s.l. (Jones et al. 2000; Toulemonde et al. 2015; Obradović 2020; Valamoti et al. 2022a). Here, we take a general look at the relative proportions of Timopheev’s wheat, einkorn and emmer. Currently, the counts of grains or chaff of these wheats are available for 53 sites in our study region (Table 2). We explore the quantitative relationships for 31 sites that have provided at least 30 remains of Timopheev’s wheat (Fig. 4). At three sites (Feudvar, Hódmezővásárhely-Gorzsa and Apsalos), the percentages of Timopheev’s wheat exceed those of einkorn and emmer and the assemblage from Hódmezővásárhely-Gorzsa is dominated by it, being more abundant than einkorn (no emmer was reported). Only chaff of Timopheev’s wheat was mentioned for this site, but this may change as the final report on the analysis is pending (Medović and Horvath 2012). The same is true of other sites undergoing re-evaluation. For instance, the recently re-analysed site assemblage from Mursalevo is composed of 50% or more of emmer (Fig. 4, ESM 4), but storage deposits comprising mainly Timopheev’s wheat have also been noted. Similarly, although einkorn constitutes more than half of the Kapitan Dimitrievo site assemblage, discrete stores of cereals dominated by Timopheev’s wheat were found there too, demonstrating that this wheat was the main crop in some cases. Revision of the large cereal deposit from Dabene-Sarovka, initially reported as composed of spelt, revealed the presence of Timopheev’s wheat in it (ESM 4). The assemblage from Feudvar considered here originated from a single feature, a large pit containing fragments of a structure made of daub and a large dense deposit of grain and chaff of Timopheev’s wheat (Fig. 5). This deposit convincingly reflects harvest(s) made up entirely or primarily of this wheat. It also emphasises the need to consider individual samples and contexts and not only site level assemblages; for instance, the assemblage from the Feudvar site as a whole is vastly dominated by einkorn (Kroll 2016, p. 71).

Fig. 4
figure 4

Relative proportions of the three hulled wheats in the assemblages from 31 sites with 30 or more remains of grain and chaff combined of Timopheev’s wheat in our study region; numbers in brackets denote site IDs as listed in Table 2 and indicated in Figs. 1 and 2 (E, Eneolithic; V, Vinča culture; LN, Late Neolithic)

Full size image
Fig. 5
figure 5

Feudvar, Serbia, Bronze Age settlement: large pit (top) containing fragments of a daub feature and a large dense deposit of grain and chaff of Timopheev’s wheat (bottom) (top photo by Frank Falkenstein; grain and chaff photos by Dragana Filipović)

Full size image

There are some recently fully analysed site assemblages where Timopheev’s wheat is present (almost) only in the form of chaff, such as the one from Apsalos, where only glume bases and spikelet forks were found, but no grain, as was also the case with einkorn and emmer at this site (Valamoti et al. 2022a, Table 1.1). Further north, the Timopheev’s wheat component of the site assemblage from Pavlovac-Gumnište was primarily made up of chaff, most of which was discovered in outdoor hearths, along with the chaff of einkorn and emmer, suggesting combined processing of the three glume wheats (Obradović 2020). A mixture of chaff of Timopheev’s wheat, einkorn and emmer in a large pit associated with a mega-structure at Stolniceni may also reflect this practice (Kirleis and Dal Corso personal communicaton). There are numerous similar examples across the study region, demonstrating the complexity of the evidence and the difficulty in tracing it back to the cultivation status of Timopheev’s wheat. Whether grown as a pure crop, a component of maslin, or as a random (tolerated) inclusion may have varied even between households and sowing seasons (see below).

In their ethnographic work, Jones and Halstead (1995) noted that the proportions of maslin crop components vary between fields and harvests, and between farmers and their choices. The purpose of planting different cereals together is that in some years (or locations) one cereal may do well, but in the next season (or a different place), another one may provide higher returns. Growing maslins is a farming strategy intended to counter the risk of crop failure or underperformance (cf. McAlvay et al. 2022). The great regional variety of the site level proportions of the three wheats that we considered here could indeed be a reflection of this. Even relatively pure deposits of a single cereal could represent the result of growing a maslin, in which this particular cereal outperformed the others. Ethnographic work at Kolofana (on the island of Amorgos, Greece) revealed that cereals were primarily grown as a maslin of Hordeum vulgare ssp. vulgare (hulled barley), Triticum aestivum (bread wheat) and T. durum (durum wheat), with the composition of harvests from different fields ranging from what could be described as ‘monocrops’ of either barley or wheat to a more or less evenly balanced (ca. 50:50) mixture of both (Jones and Halstead 1995). Also, harvests did not necessarily (or even usually) return cereals in the same proportions that had been sown. Because barley is more drought tolerant than the wheats, it would tend to become the dominant cereal in the maslin over time, and farmers sometimes manipulated the balance between wheat and barley in the maslin to enrich the wheat component of subsequent seed corn (Jones and Halstead 1995). What we are seeing archaeologically could therefore be a variable spectrum of maslin harvests of einkorn, emmer and Timopheev’s wheat. In other words, the variation in the composition of the site assemblages may reflect a single practice, that of growing a mixture of glume wheats.

As the archaeobotanical evidence of Timopheev’s wheat in the region continues to grow, we will soon be able to explore in detail whether or not there was a change through time or geographical variation in its presence as has been done for eastern France, for instance (Toulemonde et al. 2015), and its cultivation status. Some preliminary impressions can be gained from the relative proportions of Timopheev’s, einkorn and emmer wheat in site level assemblages for the 31 sites with at least 30 remains of T. timopheevii s.l. (Fig. 6). What we notice is the apparent variation in the proportions of the three glume wheats within the subregions represented by more than just a few sites (namely the northern Aegean, the central–eastern Balkans and the Pannonian Plain). When comparing the subregions, einkorn and Timopheev’s wheat appear more prominent in the south of the study region (northern Aegean) than in the north (central and northern Balkans), where there is more emmer. This first glimpse further emphasises the potential of future detailed investigations of individual samples and contexts to evaluate the importance of Timopheev’s wheat at local and regional scales.

Fig. 6
figure 6

Map showing relative proportions of the three glume wheats in the 31 site assemblages with 30 or more remains (grain + chaff) of Timopheev’s wheat (Table 2)

Full size image

From a founder crop to its abandonment

We have not come across records of Timopheev’s wheat which are later than the final centuries of the 1st millennium bce, the end of the Hellenistic period in the study region. There are altogether only a few finds from the preceding period, with the finds from the mid 1st millennium bce (late Iron Age) comprising only a few remains per site assemblage. It seems that Timopheev’s wheat was gradually abandoned, but more data and re-analysis of previously studied material will test this impression. Archaeobotanical analyses of Roman, Byzantine and medieval sites in the region do not report the presence of Timopheev’s wheat, but we should probably revisit these assemblages too, before we entirely exclude this possibility.

The disappearance of Timopheev’s wheat from the cereal spectrum would have happened in the context of the socioeconomic transformations following the Roman conquest of much of the region in the last two centuries bce. The Roman period that ensued saw as a whole significant changes in agriculture, most notably a growth in the scale of crop growing and a decrease in its intensity compared to previous periods (cf. Kreuz 2005). This unfolded against the backdrop of increasingly hierarchical societies and a developing market economy, and it apparently resulted in a focus on crops with higher productivity (Nesbitt and Samuel 1996, p. 86). Although both einkorn and emmer occur in the assemblages from Roman sites, they are found in much smaller quantities than spelt or free-threshing wheat (Gyulai 2010; Slavova 2013; Reed and Ožanić Roguljić 2020). Other cereals, Hordeum (barley), Panicum miliaceum (millet), Secale cereale (rye) and Avena (oats) became more prominent in the late Roman and early Byzantine periods (Baron et al. 2019). The evidence from the region is, however, sparse and is rarely from rural contexts, where the local (autochthonous) population may have continued cultivation of ‘traditional’ crops.

Ancient writings and inscriptions offer equally limited information for rural areas. These tell of a growing demand for food and local food production to supply newly founded urban and military centres. This requirement led to an expansion of state land ownership at the expense of small land holders, and to the cultivation of monocrops on extensive fields (Ilić 2012, pp. 44ff), with an apparent (perhaps prescribed) emphasis on large- or multi-grained wheats, as indicated by the archaeobotanical evidence. Textual evidence for the historic periods until the 2nd century ce suggests that ancient Greek authors like Theophrastus, Hippocrates and Galen knew of three glume wheats (einkorn, emmer and spelt), giving further support to the theory of the abandonment of this crop (Valamoti et al. 2022b). However, the story told from ancient texts may differ from the archaeobotanical record, which for prehistoric Greece and surrounding regions from this period includes four glume wheats: einkorn, emmer, Timopheev’s wheat and spelt (Valamoti et al. 2022b). It is possible that T. timopheevii s.l. continued to be grown in parts of the study region in the common era (ce), something that only systematic archaeobotanical research on later periods can reveal.

In thinking about the reasons why its cultivation ceased, the recent history of the cultivation of T. timopheevii s.str. in Georgia (called Chelta Zanduri there) offers some useful clues (Jorjadze et al. 2014; Bedoshvili et al. 2021; Nocente et al. 2022). For many centuries, this wheat was grown as part of a mixed Zanduri crop together with einkorn (known as Gvatsa Zanduri) and T. zhukovskyi, and this Zanduri crop was grown in the mountainous Lechkhumi and Racha provinces of Western Georgia at 600–900 m on thin or poor soils where no other wheat was sown. In contrast, the fertile soils of the Lechkhumi plain were used for growing T. aestivum (bread wheat). In 1965, by the decree of the Soviet government, which was enforced across parts of Asia and eastern Europe, small land owners cultivating the mountain slopes were forbidden to sow their own wheat, because cheap wheat was being imported from the USSR. After the 1970s, legumes and potatoes replaced endemic glume wheats, which were additionally neglected due to the war with Russia and associated economic crisis in the 1990s. Since the early 2000s, Zanduri cultivation has slowly been restored thanks to enthusiastic farmers; however they now grow single glume wheats obtained from botanical collections rather than mixtures as previously. Nowadays, T. timopheevii s.str. is grown on several farms in the lowlands and highlands, on both good and poor soils in various parts of Georgia (M. Mosulishvili personal observation; Fig. 7). Curiously, a recent attempt at growing Georgian T. timopheevii s.str. on fertile soil on a plot near Kraków (Poland) failed, in contrast to a successful growing experiment carried out on a low mountain slope at 283 m on a relatively poor soil substrate near Tarnów (Poland) (A. Mueller-Bieniek personal observation).

Fig. 7
figure 7

T. timopheevii s.str. plants growing in a field in Georgia in July 2020 (photo by Gulnari Chkhutiashvili)

Full size image

These insights instruct us to look for reasons for the abandonment of Timopheev’s wheat cultivation in the last centuries bce in a range of contexts—ecological, socioeconomic, political; this, again, should be done on a smaller spatial and temporal scale than the one used in this overview. Furthermore, we should not rule out the possibility that cultivation of T. timopheevii s.l. persisted for a while after the final centuries bce, for instance in remote, agriculturally less attractive areas of the Roman Empire or outside its borders. On current evidence, this crop seems to have been ‘lost’ but we should not lose sight of it when analysing archaeobotanical assemblages from historical periods.

Conclusions and future prospects

This paper summarises the current state of knowledge on the archaeobotanical occurrence of T. timopheevii s.l. in large areas of southern and eastern Europe. The early and recent discoveries demonstrate its presence from the beginning of farming in the region; it arrived during the process of Neolithisation as one of the founder crops that originated in southwest Asia. It was present in various parts of the region until at least the last centuries bce.

Timopheev’s wheat regularly occurs in site level assemblages together with einkorn and/or emmer, indicating their likely combined cultivation as a mixed crop (maslin). The available evidence shows considerable variation in time and space within the region in the proportions of Timopheev’s wheat relative to the other two glume wheats. This is consistent with the likelihood that the components of a mixed crop contributed different amounts to the harvests, which is a tendency observed ethnographically. The differences, dictated by environmental and cultural factors, may have arisen on an annual basis, creating a continuum from ‘pure’ monocrop to ‘balanced’ maslin harvests, potentially discernible in archaeobotanical assemblages from individual samples or contexts. The first study dealing with Timopheev’s wheat already demonstrated the need for and benefit of carefully examining the composition of each sample and context when investigating the cultivation status of various crops and their possible combinations in the field. Such analysis has been done or is under way for many of the sites included in this overview.

Future research into the (pre)history of T. timopheevii s.l. should try to revisit early identifications of spelt and ambiguous wheat identifications, such as ‘einkorn/emmer’ and ‘emmer/spelt’ and even two-grained einkorn. It should also look for patterns in the composition and distribution of assemblages containing Timopheev’s wheat in relation to the type of archaeological context and site, and the cultural period. This will help to elucidate the uses and importance of this crop, and perhaps point to possible reasons for its abandonment.