Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia

Abstract

This paper presents a detailed examination of finds of ‘new glume wheat’ (NGW), recognised as a member of the Triticum timopheevii wheat group, at Late Neolithic sites in Croatia. Increasing evidence of this morphotype from prehistoric sites across Europe, as well as comparative studies of modern Timopheev's wheat, provide a range of comparative material. Using morphometrics this study re-examines grains and spikelet bases previously identified as NGW within the late Neolithic settlements of Velištak (Dalmatia), Sopot, and Ravnjaš (Slavonia), and late Neolithic/Eneolithic Slavča (Slovenia) [All data linked to this report can be found at https://osf.io/amryd/)

Full text

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 1 of 7
Archaeobotanical evidence of Triticum timopheevii from late
Neolithic and Copper Age Croatia
Kelly Reed
ARKEN archaeology, United Kingdom; kellyreed@hotmail.co.uk
25 September 2022
Introduction - This paper presents a detailed examination of finds of ‘new glume wheat’ (NGW),
recognised as a member of the Triticum timopheevii wheat group, at Late Neolithic sites in Croatia.
Increasing evidence of this morphotype from prehistoric sites across Europe, as well as comparative
studies of modern Timopheev's wheat, provide a range of comparative material. Using morphometrics
this study re-examines grains and spikelet bases previously identified as NGW within the late Neolithic
settlements of Velištak (Dalmatia), Sopot, and Ravnjaš (Slavonia), and late Neolithic/Eneolithic Slavča
(Slavonia).
Study sites
The archaeological site of Velištak is situated in
the Velim Valley to the north of Vodice in
northern Dalmatia, Croatia (Fig. 1). Radiocarbon
dates suggest that the settlement was founded
sometime after 5000 cal BC and lasted until 4700
cal BC (McClure et al. 2014.1027, T.1). A large
quantity of classic Hvar culture pottery has been
recovered, along with tools made of animal bone,
knapped and polished stone, as well as jewellery
and polished shells (Spondylus gaederopus)
(Podrug 2010; 2013). Archaeobotanical remains
were collected during excavations from a range of
features, including occupations layers, fireplaces
and pit fills (Reed and Podrug 2016). Overall,
seed density was low, and preservation was
generally poor. Cereals were the most recovered
plant remains at the site, accounting for 94% of
the identified assemblage. Of the cereal grains,
93% of the remains were of barley (Hordeum
vulgare ssp. vulgare), totalling 879 grains,
although 838 of these were recovered from a pit.
Emmer (Triticum dicoccum) and einkorn
(Triticum monococcum) grain and chaff were also
identified, along with one NGW spikelet, and
several possible identifications of NGW.
Sopot is situated 3km south-west of Vinkovci, on
the right bank of the river Bosut (Fig. 1). The tell
site is elliptical, measuring 113 x 98m, and is 3m
deep. Three phases of Sopot culture have been
Figure 1. Map showing the sites of Velištak, Slavča,
Ravnjaš and Sopot.
identified at the site dating from 5050–4780 cal
BC to 4340 and 3997 cal BC, as well as an early
Neolithic Starćevo settlement dated to 6060–5890
cal BC (Krznarić Škrivanko 2011; Obelić et al.
2004). Overall, seed density was low, and
preservation was generally poor. Cereals were the
most recovered plant remains at the site,
accounting for 81% of the identified assemblage
(Reed et al. 2017). Of the cereals, einkorn
represented 37% and emmer 28% of the grain
recovered, with only 8% barley. A large number
of chaff remains were recovered, although many
were fragmented making identification difficult.
NGW grains (21 grains) and chaff (1 spikelet

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 2 of 7
base) were identified from 9 of the 114 samples
examined (Reed et al. 2017). The grains had
previously been identified as spelt (Triticum
spelta) grains (Reed 2013), but without diagnostic
spelt glume bases and the identification of a NGW
glume base these identifications were later
changed.
The prehistoric site of Slavča (Nova Gradiška-
Slavča) is located approx. 1.5 km north of the
centre of Nova Gradiška (Fig 1.). The site is a fort
type, on a flat plateau at the point where the
southern slopes of Psunj exceed the Posavina
Plain. At an elevation of 240.61m, it offers a
strategic position commanding the surrounding
area. The site is a multi-layered prehistoric
settlement with Sopot and Brezovljani type Sopot
culture occupation, illustrating the transition from
the late Neolithic to the early Eneolithic (Sopot
IV). The site was also occupied through the
Eneolithic with evidence of the Lasinja, Kostolac
and Vućedol cultures (Skelec 1997). Carbon dates
have been taken from Sopot levels dating to 5210–
4950 cal BC and 4960–4340 cal BC, as well as
4250–4030 cal BC associated with Sopot IV
(Mihaljević 2013a). Overall, seed density was
relatively high, although this was due to large
numbers of chaff remains found in the samples,
but preservation was generally poor (Reed 2017;
Reed et al. 2017). Cereals were the most
recovered plant remains at the site, accounting for
98% of the identified assemblage. However, the
presence of cereal grain was very low, with only a
scattering of grains of emmer, einkorn, and barley.
Interestingly the site had a huge amount of cereal
chaff, many of which could be identified to emmer
and einkorn, with nine glume bases identified to
NGW from two of the samples.
Ravnjaš (Nova Kapela-Ravnjaš) is located on the
upper slopes of the Požega hill, north-west of the
village of Nova Kapela. Excavations have
revealed a phase II Sopot culture tell settlement
and in particular a burnt down house (SJ022). The
rectangular house was oriented north-south,
consisting of two rooms containing a large amount
of burnt material and large quantities of household
items, including millstones pottery and lithics.
Carbon dates indicate a range of 4970 to 4690cal
BC (Mihaljević 2013b). Overall, seed density was
low, and preservation was generally poor (Reed et
al. 2017). Cereals were the most recovered plant
remains at the site, accounting for 96% of the
identified assemblage. Of the cereals, einkorn
represented 31% and emmer 23% of the grain
recovered, with only 3% barley. A large number
of chaff remains were recovered, although many
were fragmented making identification difficult.
Of the NGW, eight grains were identified from
four samples, as well as one spikelet base and
several indeterminate spikelets (Reed et al. 2017).
Previously the grains were identified as spelt
grains (Reed 2013), but without diagnostic spelt
glume bases and the identification of NGW glume
bases these identifications were later changed.
Methods
Recent studies have highlighted the usefulness of
morphometric analysis as an alternative or
addition to traditional archaeobotanical methods
to address differences within and between plant
species and their remains (Portillo et al. 2019).
The application of statistical analyses to size and
shape variables can help refine and enhance
taxanomic resolution and enables analysis of form
in a comparable way. To examine the presence of
NGW at the study sites, grains identified as NGW
or Triticum sp. were photographed in dorsal and
lateral view, and spikelet bases of einkorn,
emmer, NGW and Triticum sp. were
photographed in abaxial view using a
stereomicroscope with a digital camera. For the
grains three measurements were selected: total
length (L), breadth (B) and height (H) (Fig. 2). For
Figure 2. Measured dimensions of grains and spikelet
bases. Total length (L), breadth (B) and height (H),
and width of the disarticulation scar (B) and width of
spikelet base (C).

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 3 of 7
the spikelet bases two measurements were
selected: the width of the disarticulation scar (B)
and width of spikelet base (C) (Fig. 2).
Measurements were collected using ImageJ
v.1.8.0, an open-source image analysis program
(https://imagej.nih.gov/ij/download.html).
Measurements were repeated three times and the
average taken to try and control for human error.
Measurements were also collected from published
data as comparison. This included the average
dimensions of einkorn, emmer, and NGW spikelet
bases from late Bronze Age Stillfried, Austria
(Kohler-Schneider 2003). Average dimensions of
NGW grains recently measured from Neolithic
Çatalhöyük, Turkey (Roushannafes et al. 2022),
and emmer, einkorn and NGW spikelet bases
from Çatalhöyük were measured from
illustrations in Bogaard et al. (2013). The average
dimensions of emmer, NGW and spelt grains from
Bronze Age Feudvar, Serbia, were included (Kroll
2016). Spelt grains from Iron Age Gomolava were
also measured from figures in Medović et al.
(2021) for comparison. Finally, measurements
were taken of the illustrations of NGW identified
by Jones et al. (2000) at three Bronze Age sites in
northern Greece. However, all comparative data
measurements taken from published figures will
be viewed with a certain amount of caution due to
possible errors in measurements.
Results
NGW grains were identified from Sopot and
Ravnjaš in the past, however, many of the grains
were slightly fragmented and did not allow
measurements to be accurately taken. From Sopot
six grains identified as NGW/spelt were
successfully measured, and nine grains identified
as either NGW/spelt, or NGW/emmer were
measured from Ravnjaš. At Velištak, two grains
identified as emmer were also selected as they had
a flatter appearance. Plotting the L:H and L:B
values of the grains, there was a clear clustering
of the Feudvar, Stillfried and Çatalhöyük NGW
grains to the right of the graph, while emmer and
spelt plotted to the left (Figure 3). The grains from
Ravnjaš and Velištak cluster with the emmer and
spelt grains, however, for Sopot one grain clusters
with the NGW and two grains plot away from the
Figure 3. Graph showing the length:height and length:breadth values (mm) of emmer (Triticum
dicoccum – T.D), NGW (Triticum timopheevii – T.T) and spelt (Triticum spelta – T.S) grains from
comparative sites and the grains measured from Velištak, Ravnjaš and Sopot.

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 4 of 7
group towards the top of the graph. These two
grains are wider than the comparative NGW but
have a similar flat or slightly concave ventral side
(Figure 4, 5), while the grain that plots near the
comparative NGW grains is slimmer but looks
more like emmer.
(a)
(b)
Figure 4. Carbonised (a) Triticum cf. timopheevii
grains from Sopot (Block 5, SJ6,

H6), (b) Triticum
dicoccum grains from Velištak (Sample U-185). Scale
= 1 mm
The comparative spikelet measurements were
plotted to determine where the clusters would be
in the plot (Figure 6). It showed that einkorn
spikelets cluster to the top/middle left, emmer to
the bottom right and NGW centre right. However,
when we plot the spikelets identified previously as
einkorn and emmer from Sopot, Slavča and
Velištak with the spikelets that were identified as
possible NGW we see three groups forming
(Figure 7). As with the comparative material
einkorn clusters to the top/middle left. In contrast,
the emmer spikelets cluster to the bottom centre
of the plot, while grains that cluster to the top right
are similar to the NGW comparative material.
This suggests that emmer spikelet bases at the
Croatian sites are generally narrower than the
comparative sites, while the NGW spikelets from
Croatia are wider than the emmer from the same
region.
Looking at the Croatian material (Figure 7), we
see that Sopot has one spikelet and for Slavča two
spikelet bases identified as possible NGW,
clustering in the NGW area of the plot. We also
see two spikelet bases from Slavča previously
identified as emmer (Figure 8) clustering in the
NGW area of the plot. For Velištak, we see one
spikelet cluster with einkorn, two with emmer,
and one with NGW. For Ravnjaš, we see two
spikelets cluster with einkorn, one with emmer,
and two with NGW.
Figure 5. Triticum timopheevii grains from Bronze Age Feudvar (W3063). Scale = 1 mm

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 5 of 7
Figure 6. Graph plotting the width of the disarticulation scar (B) and width of spikelet base (C) (mm) of emmer
(Triticum dicoccum – T.D), einkorn (Triticum monococcum – T.M) and NGW (Triticum timopheevii – T.T)
spikelets from comparative sites.
Figure 7. Graph plotting the width of the disarticulation scar (B) and width of spikelet base (C) (mm) of emmer
(Triticum dicoccum – T.D), einkorn (Triticum monococcum – T.M) and NGW (Triticum timopheevii – T.T)
spikelets measured from Velištak, Slavča, Ravnjaš and Sopot and the comparative site Stillfried.

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 6 of 7
Kohler-Schneider (2003), based on the spikelet’s
identified from Stillfried, suggested an
identification index, calculated as the percentage
of the width of the articulation scar and the width
of the spikelet fork. It’s unclear if this index
suggests spikelets are NGW if they fall between
the einkorn (47.2%) and emmer (31.4%) values or
whether the closer you are to the value the more
likely the identification. Nevertheless, if we apply
this to the spikelet’s measured here (Table 1), we
see generally that einkorn spikelets range from
60.6% to 47.2%, although three possible NGW
spikelets are found within this range. For emmer
the spikelets generally range from 28.2% to
44.1%, although again we see three possible
NGW spikelets between 40.1% and 40.5%. The
spikelets identified as NGW range from 38.3% to
49.5%.
Conclusion
Grains and spikelets previously identified as
NGW at four Late Neolithic sites in Croatia have
been re-assessed using morphometrics. This
method proved useful in understanding the
assemblages, as the identification of cereal grain
and chaff can be difficult. For the grain’s
examined from Sopot, Ravnjaš and Velištak, it
was clear that most of the grains were probably
either spelt or emmer. Two outlier grains from
Sopot had some similarities with NGW but were
wider in breadth than the comparative material,
putting a question mark on the identification of
NGW. The grain from Sopot that plotted closer to
the comparative NGW looked like emmer, being
wider than the comparative NGW grains, but was
slightly longer. Again, a question mark should be
put on this on the identification of NGW from the
grains alone.
The spikelet bases plotted into three clusters,
suggesting that eight spikelets had similarities
with NGW. At Ravnjaš, although no grains
correlated with NGW, two spikelet bases
clustered with the NGW comparative sites. These
came from a pit feature (sample 32, SJ28,  J/24),
which contained a large amount of fragmented
glume bases that were difficult to identify to
species. At Velištak only one spikelet base from
U49, a pit feature, was identified as NGW, while
the remaining spikelets clustered with either
einkorn or emmer. Again, no grains were
identified as NGW from Velištak. For Slavča, no
NGW grains were identified, but four spikelet
bases were identified as NGW; one from sample
U129 (SJ7), a pit feature dating to the late
Neolithic Sopot culture, two from sample U169
(SJ123), a pit feature dating to the Sopot/Kotolac
culture and one from sample U9 (SJ(22)23) a pit
feature that dates to the middle Eneolithic Lasinja
culture. Overall, the results suggest that it is
possible that NGW was present at the Croatian
sites, however, these identifications should be
viewed with caution as the number of remains are
so low.
(a) (b)
Figure 8. NGW spikelet base from (a) Ravnjaš, and (b) Slavča. Scale = 1 mm

Reed K (2022) Archaeobotanical evidence of Triticum timopheevii from late Neolithic and Copper Age Croatia. osf.io/amryd
Page 7 of 7
Bibliography
Bogaard, A., Charles, M., Livarda, A., Ergun,
M., Filipović, D., Jones, G. 2013. The
archaeobotany of mid-later occupation levels at
Neolithic Çatalhöyük. In: Hodder, I. (ed.)
Humans and Landscapes of Çatalhöyük: Reports
from the 2000-2008 Seasons: 93-128. London-
Los Angeles: British Institute at Ankara (BIAA
Monograph 47) and Costen Institute of
Archaeology Press (Monumenta Archaeologica
30)
Jones, G., Valamoti, S., Charles, M. 2000. Early
crop diversity: A “new” glume wheat from
northern Greece. Veget. Hist. Archaebot. 9:133–
146
Kohler-Schneider, M. 2003. Contents of a
storage pit from late Bronze Age Stillfried,
Austria: another record of the “new” glume
wheat. Veg. Hist. Archaeobot. 12:105–111
Kroll, H. 2016. Die Pflanzenfunde von Feudvar.
In: Kroll, H. and Reed, K. (eds.). Die
Archäobotanik. Feudvar III. Würzburger Studien
zur Vor- und Frühgeschichtlichen Archäologie.
Band 1. Würzburg University Press, pp. 37–194
Krznarić Škrivanko, M. 2011. Radiokarbonski
datumi uzoraka sa Sopota. In Botić, K.,
Kovačević S., Dizdar, D. (eds.), Panonski
prapovijesni osviti: Zbornik radova posvećenih
Korneliji Minichreiter uz 65. obljetnicu života.
Institut za arheologiju. Zagreb: 209–226
McClure S., Podrug E., Moore A.M.T., Culleton
B.J., Kennett D.J. 2014. AMS 14C Chronology
and Ceramic Sequences of Early Farmers in the
Eastern Adriatic. Radiocarbon 56(3): 1019–1038
Medović, A., Marjanović Jeromela, A., Mikić, A.
2021. An update to the La Tène plant economy
in northern Serbia. Ratar. Povrt. 58(2):53–65
Mihaljević, M. 2013. Sopotska kultura u
zapadnoj Slavoniji s posebnim osvrtom na
nalazište Slavča Nova Gradiška. Unpublished
PhD thesis. University of Zagreb. Zagreb
Obelić, B., Krznarić Škrivanko, M., Marijan, B.,
Krajcar Bronić, I. 2004. Radiocarbon dating of
Sopot culture sites (Late Neolithic) in Eastern
Croatia. Radiocarbon 46(1):245–258
Podrug, E. 2010. Čista Mala – Velištak: prve tri
istraživačke kampanje na nalazištu hvarske
kulture. Diadora 24:7–25
Podrug, E. 2013. Neolithic immovable finds in
the Šibenik area.Diadora 26/27:185–212
Portillo, M., Ball, T.B., Wallace, M., Murphy,
C., Pérez-Díaz, S., Ruiz-Alonso, M., Javier
Aceituno, F., Antonio López-Sáez, J. 2020.
Advances in Morphometrics in Archaeobotany.
Environmental Archaeology 25(2):246–256
Reed, K. 2013. Farmers in transition: the
archaeobotanical analysis of the Carpathian
Basin from the Late Neolithic to the Late Bronze
Age (5000–900 BC). Unpublished PhD thesis.
School of Archaeology and Ancient History,
University of Leicester
Reed, K. 2017. Agricultural change in Copper
Age Croatia (ca. 4500–2500 cal B.C)? Archaeol.
Anthropol. Sci. 9:1745–1765
Reed, K., Podrug, E. 2016. Reconstructing late
Neolithic plant economies at the Eastern Adriatic
site of Velištak (5th millennium cal BC).
Documenta Praehistorica 43:399–41
Reed, K., Škrivanko, M. K., Mihaljević, M.
2017. Diet and subsistence at the late Neolithic
tell sites of Sopot, Slavca and Ravnjas, eastern
Croatia. Documenta Praehistorica 44:326–337
Roushannafas, T., Bogaard, A., Charles, M.
2022. Geometric morphometrics sheds new light
on the identification and domestication status of
‘new glume wheat’ at Neolithic Çatalhöyük.
Journal of Archaeological Science 142 (2022)
https://doi.org/10.1016/j.jas.2022.105599
Skelec G. 1997. Prapovijesno nalaziΠte Slavča.
Opvscvla Archaeologica 21:217– 233