ArticlePDF Available

Life histories at stone age Zvejnieki based on stable isotope profiles of tooth dentine

  • Pisces Conservation Ltd
  • Institut of Latvian history, Latvian University


The timing of infant weaning in the past is important for its implications for birth-spacing and infant survival, and hence for population maintenance or growth under different socio-economic regimes. Prior to the adoption of agriculture, breastfeeding is believed to have been more prolonged amongst hunter-gatherers due, at least partly, to the lack of suitable weaning foods that are available to agriculturalists. The introduction of pottery possibly also changed weaning patterns due to shifts in food preparation even prior to the adoption of domesticated foods. Here we apply stable carbon and nitrogen isotope sequential samples on dentine to explore differences in diet relating to weaning age, social roles and food sharing between children and adults in a well preserved Mesolithic/Neolithic population from the cemetery of Zvejnieki, Latvia. We address whether there are differences in diet between the Mesolithic and the Neolithic periods, defined here by the appearance of pottery rather than the adoption of agriculture. Considerable variability in weaning patterns was observed, but in general individuals tended to be breastfed from birth, with the contribution of breast milk declining after the age of 6–12 months, and completely withdrawn by the age of 3 years. We note a difference in δ15N dentine profiles between the Mesolithic and Neolithic, which may be linked to the introduction of pottery. We also assess differences in diets in relation to identities marked in death, specifically the presence or absence of animal tooth pendants. The carbon and nitrogen isotope profiles for sequentially sampled first molars show that adults who were buried without animal tooth pendants as grave goods consumed more freshwater resources during their childhoods than those buried with animal tooth pendants. We conclude that infant and childhood diet reflected different societal roles or identities within the population that continued into adulthood.
Journal of Archaeological Science: Reports 44 (2022) 103496
Available online 11 June 2022
2352-409X/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (
Life histories at stone age Zvejnieki based on stable isotope proles of
tooth dentine
Rowena C. Henderson
, Gunita Zarin¸ a
, Andrea Czermak
, Rick J. Schulting
Peter A. Henderson
, Dardega Legzdin¸ a
, Ilga Zagorska
, Julia Lee-Thorp
School of Archaeology, University of Oxford, 1 South Parks Road, Oxford OX1 3TG, UK
University of Latvia, Kalpaka bulv¯
aris 4, R¯
ıga LV-1050, Latvia
Department of Zoology, University of Oxford, New Radcliffe House, Radcliffe Observatory Quarter, Woodstock Road, Oxford OX2 6GG, UK
Carbon and nitrogen isotopes
Dentine increments
Early life histories
Grave goods
Aquatic foods
The timing of infant weaning in the past is important for its implications for birth-spacing and infant survival,
and hence for population maintenance or growth under different socio-economic regimes. Prior to the adoption
of agriculture, breastfeeding is believed to have been more prolonged amongst hunter-gatherers due, at least
partly, to the lack of suitable weaning foods that are available to agriculturalists. The introduction of pottery
possibly also changed weaning patterns due to shifts in food preparation even prior to the adoption of domes-
ticated foods. Here we apply stable carbon and nitrogen isotope sequential samples on dentine to explore dif-
ferences in diet relating to weaning age, social roles and food sharing between children and adults in a well-
preserved Mesolithic/Neolithic population from the cemetery of Zvejnieki, Latvia. We address whether there
are differences in diet between the Mesolithic and the Neolithic periods, dened here by the appearance of
pottery rather than the adoption of agriculture. Considerable variability in weaning patterns was observed, but in
general individuals tended to be breastfed from birth, with the contribution of breast milk declining after the age
of 612 months, and completely withdrawn by the age of 3 years. We note a difference in δ
N dentine proles
between the Mesolithic and Neolithic, which may be linked to the introduction of pottery. We also assess dif-
ferences in diets in relation to identities marked in death, specically the presence or absence of animal tooth
pendants. The carbon and nitrogen isotope proles for sequentially sampled rst molars show that adults who
were buried without animal tooth pendants as grave goods consumed more freshwater resources during their
childhoods than those buried with animal tooth pendants. We conclude that infant and childhood diet reected
different societal roles or identities within the population that continued into adulthood.
1. Introduction
Sequential stable carbon and nitrogen isotope analysis of tooth
dentine is increasingly used for the investigation of individual ‘life his-
toriesrelated both to diet and to stress episodes during early develop-
ment (e.g., Beaumont and Montgomery, 2016; Eerkens et al., 2011;
andez-Crespo et al., 2018; Henderson et al., 2014; Lillie et al., 2016;
Montgomery et al., 2013; Sandberg et al., 2014; Scharlotta et al., 2018;
Tsutaya et al., 2015). In particular, weaning ages have been explored
based on detailed isotopic analysis of the increments in the dentine of 1st
permanent molars, which develop shortly after birth and for several
subsequent years. Using this approach, episodes of dietary change or
starvation have been detected in Neolithic Shetland (Montgomery et al.,
2013) and in mid-nineteenth century Ireland (Beaumont and Mont-
gomery, 2016). In two Late Neolithic groups practicing different burial
traditions in northern Spain, subtle long-term dietary differences in
children and adults were detected using this method (Fern´
et al., 2020). So far, opportunities to explore the weaning practices of
more mobile forager populations, which may have been distinct from
those of more recent sedentary agricultural populations, have been
limited (Eerkens et al., 2011; Scharlotta et al., 2018), at least partly
because their inhumations are often infrequent (Fern´
andez-Crespo et al.,
The cemetery of Zvejnieki, northern Latvia, is unusual in that it
contains a large number of well-preserved skeletons from the Mesolithic
and succeeding ‘Pottery Neolithicperiods, dened by the use of pottery
* Corresponding author.
E-mail address: (R.C. Henderson).
Contents lists available at ScienceDirect
Journal of Archaeological Science: Reports
journal homepage:
Received 16 February 2022; Received in revised form 25 May 2022; Accepted 25 May 2022
Journal of Archaeological Science: Reports 44 (2022) 103496
rather than by the appearance of domesticates (Liiva and Loze, 1993;
ciauskas et al., 2017). The ‘Pottery Neolithichas been observed as a
widespread phenomenon across the high northern latitudes of Europe
and Asia that has been frequently shown to be associated with the
preparation of aquatic foods (e.g. Isaksson and Gibbs, 2018; Lucquin
et al., 2016). A nal phase, the ‘Late Neolithic, sees the presence of
domestic animals and thus ts more securely into conventional deni-
tions of ‘Neolithic as indicating evidence for a degree of reliance on
agriculture. The site thus offers a rare opportunity to compare changes
in childhood diets and weaning practices in a population with a forager
economy before and after the adoption of pottery from the same loca-
tion. This is important since one of the assumed drivers for the use of
pottery is the expansion of dietary breadth that it facilitates, including
the preparation of soft foods suitable for weaning. Post-weaning, we are
also able to investigate possible movements between the sites inland
location and the coast ca. 50 km to the west, suggested by the presence
of dog tooth pendants with isotopic values characteristic of marine re-
sources, as well as one adult bone collagen value exhibiting moderately
C (Eriksson and Zagorska, 2003). Finally, we explore possible
social differences within the population, since grave goods are associ-
ated with many of the burials (Schulting et al., 2020).
1.1. Site background
Zvejnieki is located on the shores of Lake Burtnieks in northern
Latvia (Zagorskis, 1987) (Fig. 1). It has both an occupation and a cem-
etery site, the latter containing over 300 inhumations making it one of
the largest known hunter-gatherer cemeteries in Europe. The cemetery
was used throughout an extended period of time from the Mesolithic and
Neolithic, into even later periods (c.70003000 cal BC) (Meadows et al.,
2018). A substantial number of radiocarbon determinations have shown
that the cemetery was rst established in the Middle Mesolithic, with the
earliest directly dated burial at 8240 ±70 cal BC (Ua-3634) (Zagorska,
1997). The majority are from the Late Mesolithic (c. 70005200 cal BC),
Early Neolithic (c. 52004200 cal BC) and Middle Neolithic (c.
42003200 cal BC); again, with the Early Neolithic dened by the
regional appearance of pottery c. 5200 cal BC. Denition of these pe-
riods at Zvejnieki is based almost entirely on direct radiocarbon dating
of human remains, which for all periods excepting the Late Neolithic
Corded Ware is complicated by the consumption of aquatic foods
imparting signicance amounts of ‘old carbon, leading to a freshwater
reservoir effect (Meadows et al., 2016; 2018). Dates are likely to be ‘too
old by two to three centuries on average in individuals relying on
freshwater foods (Meadows et al., 2018). We have not corrected radio-
carbon dates here as not all individuals analysed have been directly
dated, and we are only interested in broad period attributions.
The ‘Early NeolithicNarva culture in the Baltic region is dened by
the presence of pottery, rather than of domesticates (Liiva and Loze,
1993; Piliˇ
ciauskas et al., 2017). Thus, there is considerable continuity
between the Mesolithic and Neolithic, as shown by the continued use of
the cemetery, similar burial practices, the continued use of symbolism
such as red ochre and animal tooth pendants within the graves and a
sustained reliance on hunting, gathering and shing (Zagorskis, 2004;
Eriksson, 2006). The exception is a small shift in the Middle Neolithic to
pendants made from the teeth of dogs and seals, versus the earlier pre-
dominance of red deer and wild boar (Larsson, 2006). The situation
changes in the Late Neolithic when domestic animals rst appear in the
region in association with Corded Ware groups (Kriiska, 2003).
1.2. Previous isotopic Research at Zvejnieki
The rst dedicated stable carbon and nitrogen isotopic studies on
human and faunal remains from Zvejnieki were undertaken by Eriksson
and colleagues (Eriksson and Zagorska, 2003; Eriksson, 2006). The δ
results for human bone and tooth collagen range from 8.9
to 17.7,
and δ
C from 25.0 to 18.8, suggesting highly diverse diets
during the Late Mesolithic and Early/Middle Neolithic, with a strong
reliance on freshwater resources for some individuals. The isotopic
indication for a freshwater aquatic food diet at Lake Burtnieks is char-
acterised by low δ
C and high δ
N values, while marine foods in
contrast are characterised by relatively high δ
C and high δ
N. Addi-
tional isotope data for bone collagen since obtained by Meadows and
colleagues (Meadows et al., 2016; Meadows et al., 2018) suggests no
major dietary shift with the introduction of pottery. There is, however, a
hint of a minor trend away from high-trophic-level foods in the 6th
millennium BC, a more diverse diet in the 5th millennium BC and a
possible narrowing of the diet in the mid-4th millennium BC. The small
number of Late Neolithic Corded Ware individuals analysed to date tend
to have lower δ
N values on average, indicating a shift in diet consistent
with the appearance of a farming economy (Eriksson, 2006).
The published faunal and human isotope data are summarised in
Fig. 2. Animals with δ
C values indicative of a marine diet include a
mallard (Anas platyrhynchos) and seals (Halichoerus grypus), while the
most negative δ
C values came from freshwater sh and pond turtle
(Emys orbicularis). The most elevated δ
N values came from the aquatic
animals including otters (Lutra lutra), pond turtles (Emys orbicularis),
pike (Esox Lucius), and from badgers (Meles meles), which are terrestrial
omnivores (Eriksson, 2006). Isotopic data for dog tooth pendants asso-
ciated with burials clearly show that some dogs consumed food from
marine resources, indicating the exchange of either live dogs or dog
tooth pendants (Eriksson and Zagorska, 2003). Human adults tend to
have elevated δ
N values, comparable to those of otters, unidentied
sh species and mallard.
An early ‘life historypilot study was undertaken by Eriksson (2007)
for a single individual, a male in Grave 165, with a slightly higher bone
collagen δ
C value compared to other individuals at the site. It was
therefore hypothesised that he may have consumed more marine foods
and perhaps originated on the coast. It should be noted that δ
N values
do not show the typical trend of increasing alongside δ
C values
because freshwater sh are equally enriched in
N. Single samples of
tooth dentine from the rst, second and third molars (M1, M2, and M3,
respectively) representing the age brackets of 24, 510 and 1016 years
were analysed for δ
C and δ
N. The highest δ
N value occurred in the
youngest age period of 24 years (13.5), which was interpreted as
reecting at least in part a nursing signal. Modest variation in δ
C was
noted throughout the individuals life (24 years 22.0, 510 years
20.9, 1016 years 20.6, 2035 years 18.8and 3035 years
), leading the author to suggest increasing marine input into
the diet from childhood through to adolescence and early adulthood,
declining again in later adulthood. Whether the individual made short
regular trips to the coast or longer trips could not be ascertained due to
the low temporal resolution of this sampling method (Eriksson, 2007).
2. Aims
The aim of this study was to apply incremental dentine carbon and
nitrogen isotope analysis to better understand the childhood diets and
weaning practices of individuals who survived into adulthood in a
diachronic sample from Zvejnieki. Hunting and gathering societies are
1 1
The lowest δ
N value of 6.2 for a worn incisor pendant identied as
human is very unusual compared to all other human data and may be a mis-
identied faunal tooth.
2 2
Obtained from bone collagen taken from different bone elements.
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
often thought to prolong breastfeeding, with consequences for infant
health and survival, and lower birth rates (Anderson et al., 1986). This
stands in contrast to generally earlier weaning in agricultural-based
societies (e.g. Bocquet-Appel, 2011), leading to signicant anthropo-
logical interest in weaning ages because of its inuence on infant
survival and birth-spacing, and consequently on population structure
and demography. Prolonged nursing, however, may apply primarily to
highly mobile foragers, rather than to the more sedentary hunter-sher-
gatherers that better characterise southern Scandinavia and the eastern
Baltic. But it is also not clear that sedentism alone is the major inuence
rather than availability of energy-rich foods, since reductions in birth
interval are linked to maternal energy balance (Bocquet-Appel, 2011).
In Zvejnieki, the appearance of pottery in the sequence allows us to
investigate whether or not its adoption led to any isotopically detectable
changes in weaning practices by making novel supplementary foods
available (cf. Orlas et al., 2017)
Here we investigate the life histories of selected Mesolithic and
Neolithic individuals from Zvejnieki as expressed in their isotopic intra-
tooth sequences, to assess the following: 1) changes in weaning age and
the nature of supplementary foods before and after the appearance of
pottery; 2) whether higher resolution sampling can identify excursions
to the coast as suggested by isotopic values for dog tooth pendants and
by hints of enriched
C in one human (Grave 165); and 3) variability in
childhood and adult diets in relation to the provision of animal tooth
pendants as grave goods. We earlier noted differences in adult diets
between those with and without pendants (Henderson, 2015; Schulting
et al., 2020). The aim here is to investigate the age at which these ma-
terial culture-related dietary differences can be detected, and whether
they may have inuenced other aspects of early life history (that are
amenable to stable isotope investigation). Although a variety of grave
goods were recovered, tooth pendants are the most common artefact
class in both male and female graves. They may have represented status
or identity that may have been ascribed from birth due to a familial
relationship or afliation with a particular group. It might thus be
possible to explore issues of status and/or identity as well as any intra-
Fig. 1. Map showing the location of Zvejnieki adjacent to the large freshwater Lake Burtnieks. Map data © OpenStreetMap contributors, Microsoft, Esri Community
Maps contributors, Map layer by Esri, Sources: Esri, USGS, National Geographic, Esri, Garmin, HERE, UNEP-WCMC, USGS, NASA, ESA, METI, NRCAN, GEBCO,
NOAA, increment P Corp., Esri, HERE, Garmin, FAO, NOAA, USGS, Esri, USGS.
Fig. 2. A comparison of the average δ
N and δ
C in humans, compared to
broad faunal trophic groups. Error bars show ±1 standard error. Sample sizes
are: 11 terrestrial carnivores, 37 terrestrial herbivores, 9 aquatic birds and sh,
7 seals, 11 terrestrial omnivores, 5 aquatic mammals, 9 dogs, and 31 adult
humans. Data from Eriksson, 2006, Eriksson and Zagorska, 2003 and Meadows
et al., 2018.
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
population variability or diachronic trends as noted above.
3. Materials and methods
The left mandibular rst molars of 18 individuals were sampled
(Supplementary Data, Table 4), in order to reect approximately the
rst 08 years of development. Some of the teeth exhibited advanced
attrition, which is often a feature of hunter-gatherers consuming coarse,
gritty foods (Smith, 1984). Thus, the full life histories spanning 08
years old were not available in these cases. Where a loss of crown
dentine had occurred, the number of missing segments was estimated.
The approach uses incremental samples of dentine, following
methods developed earlier by Eerkens et al. (2011) and Beaumont et al.
(2013). Teeth were rst cleaned using aluminium oxide air abrasive to
remove adhering sediments allowing for closer inspection of the enamel
surface for the presence of small carious lesions which were avoided.
Each tooth was sectioned using a Buehler low speed saw with a diamond
wafer blade. One half of the tooth was conserved for future analysis or
curation, and the remaining half was stripped of enamel using a hand
drill. This half was placed in 0.5 M HCl at 4 C in order to remove the
mineral phase (Honch, 2008). The acid was changed every two days and
the tooth remained submerged until it felt soft when probed with a
pipette. The demineralisation time was highly variable between sam-
ples, depending on preservation. All tooth halves retained their shape
after demineralisation.
The demineralised tooth halves were sectioned into ten horizontal
incremental samples from the crown to the root using a scalpel, ac-
cording to Henderson et al. (2014). Following this procedure, ve
samples were evenly spaced throughout the crown zone, while the
remaining ve were evenly spaced from the cemento-enamel junction to
the root tip. Age assignments are based on the London Atlas of Tooth
Formation (AlQahtani et al., 2010), as follows. Each samples location
was compared to the stage of dental development of the M1 and assigned
to a putative age by using anatomical locations as temporal markers for
growth rates and the data are plotted as presented in Czermak et al.
(2020). Anatomical areas are dened according to The London Atlas as
crown, prefurcation, middle root, and root tip, and are used as temporal
markers for growth rates (Fig. 3, lower right). The median age of
completion of each developmental stage was used: crown 3 years, pre-
furcation 5 years, middle root 6.5 years, root tip 7.5 years. As dentine
formation occurs to a certain stage by a specic age, any sample from
before this stage represents dietary intake by this age. Sample timespans
are calculated by dividing the time of completion of an anatomical area
with the number of samples taken within an anatomical area. For
example, the crown formation of the rst permanent molar starts around
birth and ceases at approximately three years of age (AlQahtani et al.,
2010). Each of the ve samples taken from this area covers approxi-
mately a fth of that timespan. Then the mid-point of each time-span
was calculated. A midpoint of 0.25 years of age was used for the rst
sample, reecting the average diet consumed between 0 and 0.5 years.
Similarly, the second sample was assigned to a midpoint of 0.75 years of
age reecting the average consumed diet between 6 and 12 months.
Thus samples from the crown are representative of a shorter period of
time (~3 years overall) than those from the root representing ~38
years, meaning that crown segments provide a higher resolution
The demineralized slices were rinsed three times in distilled water,
placed in test tubes in pH3 water and heated at 70 C for two days in
order to denature the collagen. Thereafter, they were cooled, ltered
using Ezee lters (Elkay Laboratory Products), and freeze-dried.
Approximately 1 mg of the resulting dentine ‘collagen (i.e., including
a small proportion of non-collagenous proteins) was weighed into tin
Samples were analyzed on a SERCON 20/22 continuous ow isotope
ratio mass spectrometer coupled to a SERCON GSL elemental analyzer in
the Research Laboratory for Archaeology and the History of Art,
University of Oxford. Analytical reproducibility is ±0.2for both δ
and δ
N, based on multiple replicates of the internal laboratory stan-
dard, alanine, used to calculate instrumental drift. International stan-
dards IAEACH6 and USGS40 and two well-characterised laboratory
standards, COW and SEAL, were intermittently spaced between the
unknowns, in order to further assess the quality of the run and to cali-
brate the results. The δ
C and δ
N measurements were calibrated using
a two-point calibration curve using the internal cow and seal bone
collagen standards (following Coplen et al., 2006).
All statistical analyses were undertaken using the R platform (R Core
Development Team, 2013). Generalized Linear Models (GLMs) were
used to test for the effects of age and pendant presence on carbon and
nitrogen stable isotope ratios. Statistical analysis of the effect of age and
presence of pendants on the stable isotopes requires a nested analysis as
multiple observations were collected from each individual tooth. The
nlme library (Pinheiro et al., 2018) was used to t a mixed effects model
including nested random effects. A term was included for the correlation
between measurements representing the different ages across a single
tooth. The basic model for δ
N is:
N =Pendant +Age +Pendant*Age, random =~1|Individual
with the correlation between measurements for an individual tooth
between ages given by a rst order autoregressive model with a
correlation =0.78.
The basic model for δ
C is similar to that for δ
C =Pendant +Age +Pendant*Age, random =~1|Individual,
with the correlation between measurements for an individual be-
tween ages given by a rst order autoregressive model with a cor-
relation =0.81. Statistical analysis of isotopes on adults was
performed using analysis of variance using the aov R function.
Further to the age assignment methods outlined above, the term
‘young infantis used to refer to individuals up to 1 year of age, ‘older
infantfrom 1 to 3 years, ‘childrenfrom 3 to 8.5 years. To estimate the
maximum duration of the weaning process, two marker points that
represent the last point before and the rst point after weaning are used
(Eerkens et al., 2011; Nitsch et al., 2011; Sandberg et al., 2014). The
sample with the earliest highest δ
N value before the typical decrease in
N, suggesting the introduction of complementary food, was used to
estimate the onset of weaning. The sample with the lowest δ
N value
after the sustained drop represents the end of the transition and the point
at which the child can be considered to be weaned.
The duration of weaning was estimated by subtracting lowest δ
value observed at the ‘troughof a decline from the highest value. Note
that this need not always be the lowest value, since in some cases a
slower decline continued into the mid-childhood ‘dip in δ
N values
seen in some individuals. In such cases there was usually still a marked
change in the slope that was used to infer the completion of weaning (e.
g., Grave 121). The difference between the two values is termed Δ
The rst sequential sample representing age 06 months was given
the value of 0.6 yr, since this slice is likely to be biased towards the later
part of the range. The result of this is that weaning is assumed to begin
only after six months, at least in a way that is detectable, which will
depend in part on the difference in the isotopic values between breast-
milk and the complementary foods. For subsequent increments, the
midpoint of the age ranges was used.
4. Results
Comparison between dentine collagen carbon (δ
C) and nitrogen
N) isotope proles of M1s from the three time periods is shown in
Fig. 3. Sequential δ
N and δ
C data for all individuals are shown in
Fig. 4 and Supplementary Table S1. Averaged data for each period are
summarised in Table 1, while all data and quality control indicators such
as C/N ratios are provided in the Supplementary Material Table S1. Late
Neolithic individuals also differ compared to those from the Early-/Mid
Neolithic period, being signicantly lower in δ
N and higher in δ
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
values during the rst 7 years. The difference in δ
N results between
Mesolithic and Late Neolithic is signicant in the rst 3 years (Fig. 3;
statistical data see Supplementary Table S2 and S3). It is notable that the
most elevated δ
N in infants and children are seen in Middle Neolithic
individuals. A further observation is that while the data for the two
earlier periods vary between individuals, the life-history pathway for
each individual seems to continue along the same trajectory, with
elevated δ
N in infancy, declining into childhood to below adult values,
before rising again into adulthood.
The data for both isotopes are highly variable especially through the
Mesolithic and Early/Middle (Pottery) Neolithic (Table 1). The δ
values for many individuals are low (ca. 23 to 25) suggesting strong
reliance on aquatic foodstuffs. However, this is not always the case, for
instance the individuals in Graves 92 and 170 do not follow this pattern
as their values are closer to 21, which one might expect for an in-
dividual living in a C
terrestrial environment at this latitude. More of
the Late Neolithic individuals, on the other hand, have conventional
C values close to 21(mean value 21.4SD 0.391). In general,
N values declined steadily from those early segments representing the
youngest infants (up to 1 year of age) (mean 14.05 ±0.38), to the
older infants (12.80 ±0.28) (aged 13 years) and children (11.80 ±
0.19) (aged 38.5 years), and nally the adults (11.8 ±0.2)
established from Eriksson (2006). Mean δ
C values on the other hand
do not show any clear corresponding age-trends. There is no difference
between the children and adults (Students t-test, t =0.788, p =0.433),
and no statistically signicant trend in δ
C in relation to age (F =1.345,
p =0.261). Two Early/Middle Neolithic adult males, from Graves 124
and 179 (the former, incidentally, with pendants, and the latter
without), show similar early life histories to one another, distinct from
those of all the other individuals (Fig. 4.). Both exhibit increasing δ
values from 0 to 6 months to a peak at ca. 2 years of age, following which
they decline sharply. The δ
N and δ
C mirror each other in these two
individuals, a pattern which is also seen most clearly in the Mesolithic
individual 154 and the Neolithic individuals 173 and 207. Six in-
dividuals also have adult bone isotope values available from previous
analysis (Meadows et al., 2018, Eriksson, 2006), comprising individuals
123, 170, 124, 313, 154 and 307. Excluding 313, which did not produce
data for the childhood period, these individuals had mean childhood
isotope values of 12.79 ±0.45and 22.43 ±0.26(n =5). The same
individuals had mean adult isotope values of 11.83 ±0.63and 23.31
±0.33(n =6).
5. Diachronic trends in weaning practices through time
The shape of the curves shown in Fig. 3 show that the sequences for
both isotopes are quite variable within each ‘period, and one can
observe that there is a striking difference in the proles of both isotopes
for the (pottery) Neolithic period compared to the preceding periods
(which look similar).
The results of individualsweaning pattern are shown in Table 1. The
overall minimum duration of exclusive breastfeeding was 0.5 ±0.3
years and is fairly consistent with 0.7 ±0.2 years in Mesolithic (n =5),
0.5 ±0.4 years in Early-/Mid Neolithic (n =9) and 0.4 ±0.3 years in
Late Neolithic (n =4) individuals. While there is individual variability,
the average maximum duration of the weaning process was 2.7 ±1.4
years, with 2.5 ±1.2 years in the Mesolithic, 2.7 ±1.3 years in the
Early-/Mid Neolithic and 2.0 ±1.5 years in the Late Neolithic. The
average age at weaning cessation was 3.0 ±1.4 years, with 3.1 ±1.1
years in the Mesolithic, 3.2 ±1.6 years in Early-/Mid Neolithic, and 2.3
±1.5 years in Late Neolithic individuals (see Supplementary Figure S1).
There are no signicant differences in either duration of breastfeeding,
weaning process or age at weaning cessation between the time periods,
though sample sizes are small and all these ndings should be taken as
Grave 206 (child, 714 years, Early-/Mid Neolithic) showed the
longest estimated weaning period (5.2 years), the longest duration of
exclusive breastfeeding (1.3 years) and the oldest age at weaning
cessation 6.5 years. Late Neolithic Grave 137a (4.3 years) and Early-/
Mid Neolithic Grave 207 (3.8 years) also showed longer weaning
duration than the average (2.7 years). The overall shortest weaning
duration was detected in Late Neolithic individuals 137 and 186 (1.1
years). The shortest weaning period during Mesolithic (1.3 years) was
detected in individuals 154 and 170. While the duration of exclusive
breastfeeding was average, the age at weaning cessation was at least 1.4
Fig. 3. Comparison between dentine collagen nitrogen (δ
N) and carbon (δ
C) isotope proles of M1s from the three time periods (Mesolithic: blue, Early-/Mid
Neolithic: green, Late Neolithic: red). Mean values obtained from bone collagen representing individuals at adulthood are shown for reference on the edge of the plot.
The schematic graph illustrates the correspondence between tooth developmental ages and anatomy used to assign age to each dentine sample. Data shown for each
individual (left) and mean values for each period (right). Shaded areas represent the mean values ±1 pooled SDs at different age periods (03, 35, 57 and 78
years). For corresponding data see Supplementary data Table S2 and S3.
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
years earlier than those of the other three Mesolithic individuals and
N values, both pre- and postweaning, were higher. However, no
signicant decrease in δ
N during early life and thus no strong weaning
signal can be seen.
The overall mean preweaning and postweaning δ
N was 14.8 ±
2.3and 11.5 ±1.9, with 14.3 ±1.3and 12.3 ±2.2during the
Mesolithic, 16.6 ±1.4and 11.9 ±1.2in the Early-/Mid Neolithic,
and 12.2 ±1.9 and 9.5 ±1.7 in Late Neolithic individuals. The
strongest decline in δ
N can be seen in individuals 206 (7.7) and 201
(5.5), both Early-/Mid Neolithic, the lowest decline was detected in
Mesolithic individual 170 (0.9). There is a signicant difference in the
decline in δ
N during the weaning process between the Mesolithic and
the Early-/Mid Neolithic (p =0.02, t-test). The average decline is 3.3 ±
1.7, with 2.0 ±1.0 in Mesolithic, 4.5 ±1.7 in Early-/Mid
Neolithic and 2.7 ±1.1 in Late Neolithic individuals (Fig. 4). The
lowest postweaning δ
N showed Early-/Mid Neolithic individual 206
(9.6), and Late Neolithic individuals 137 (9.9) and 137a (7.1).
However, while the decline was below the average of 3.3in 137 and
137a (2.0 and 3.0), it was strongest in individual 206 (7.7). A strong
decline can also be seen in Early-/Mid Neolithic individuals 201 (5.5)
and 208 (4.8), whereas the lowest decline can be seen in Mesolithic
individuals 154 (1.0) and 170 (0.9) (Fig. 5).
5.1. Presence or absence of animal tooth pendants
Statistical analysis of pendants and age versus δ
N in children
showed that both pendant presence and age were signicant predictors
of δ
N (p =0.032 for pendants and p <0.001 for age). Late Neolithic
individuals were excluded from this analysis. There was however no
signicant pendant-age interaction term, allowing a simplied model to
be used to eliminate the interaction term. This leads to the adoption of
the simpler model: δ
N =Pendant +Age, random =~1|Individual.
There is a decrease in δ
N from infancy to childhood and variability in
diet based on the presence or absence of tooth pendants in individuals
between the ages of 38.5 years was found to be signicant (p =0.03).
No statistically signicant difference was found in δ
C between young
infants (01 year) buried with or without pendants (p =0.24). There is a
suggestion of a pendant effect on δ
N values, with children buried with
Fig. 4. Sequential δ
N and δ
C data for all individuals analysed in this study, showing the former as red dotted lines and the latter as blue. Where available the δ
and δ
C data for the adult is shown on the edge of the plot (from Eriksson, 2006; Meadows et al., 2018). The data are shown in columns according to period -
Mesolithic, Early/Mid or Pottery Neolithic (across two columns), and Late Neolithic to the right. Each plot gives the grave number, the sex and the radiocarbon age
where known. The x-axis provides an estimation of approximate age in years and the panel immediately below each plot shows the divisions. This is illustrated by the
tooth section in the lower right of the gure. The radiocarbon dates are from Meadows et al. (2016) and Meadows et al. (2014), and the grave numbers and sex from
Zagorskis (2004).
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
pendants having less elevated δ
N, though it narrowly fails to attain
statistical signicance (p =0.053).
These analyses show that individuals without tooth pendants had, on
average, higher δ
N values. Individual 170 is a clear outlier from the
group with pendants, with an average δ
N of 15.7during childhood.
This male burial dates to the Middle Mesolithic (8159 ±80
C yr BP)
and is notable for its impressive array of grave goods, including 167
animal teeth pendants from elk, wild boar and aurochs, ornaments and a
headdress, suggesting this person may have been an important member
of the community with high status (Zagorskis, 2004). (Note that, based
on the early date, this individual was excluded from the analysis pre-
sented in Schulting et al. (2020)).
The δ
C values in the group with pendants cluster into two distinct
sets, with individuals 170 and 92 having less negative δ
C values than
the remaining individuals. As noted, individual 170 is an anomaly with
elevated δ
N, whereas there is no distinguishing feature of individual
92 which marks it out from the other graves.
The mean adult δ
N values, based on published data from Eriksson
(2006) and Meadows et al. (2016), were found to be signicantly
different between individuals buried with and without tooth pendants,
again excluding the Late Neolithic individuals (analysis of variance, n =
30, F =4.6, p =0.041). No signicant difference in δ
C was identied
(analysis of variance, n =30, F =1.76, p =0.195), though the difference
is in the direction expected given the negative relationship between δ
and δ
C values in freshwater resources in this region (Fig. 2), i.e., those
interred with pendants are higher in δ
6. Discussion
6.1. Diet variability and life histories
There is a clear breastfeeding signal in infancy in most individuals,
with elevated δ
N in the rst year (mean δ
N =14.6 ±0.4). The
small drop in δ
N between the 06 month and 612 month age brackets
suggests other foods began to be introduced to some individuals, but in
general there was a consistent inclusion of breast milk between these
two age brackets. The greater decline in δ
N occurs between the 612
month and 1220 month age brackets, although this is variable. This
more substantial change indicates this was a period when breast milk
began to be withdrawn, and continued to be withdrawn until completion
of weaning around age 2028 months. There is a large degree of vari-
ability between individuals in both δ
C and δ
N values during infancy.
This is either a reection of different infant rearing strategies or a
reection of the mother/carers diet. Individuals 124 and 179
Table 1
Duration of exclusive breastfeeding and weaning process for the 18 individuals with dentine microsampling results. The ‘*denotes graves containing animal tooth
pendants. ‘#denotes that onset of weaning was determined by change in δ
C as marker for the introduction of weaning food. Data from bone are taken from Eriksson
(2006) and Meadows et al. (2018).
Grave Period Age at
Sex Estimated min.
duration of exclusive
breastfeeding (years)
of weaning
Age at
N decline
during weaning
92* Mesolithic Adult Male 0.25 13.0 3.3 3.5 10.5 2.5 12.0
121* Mesolithic Adult Male 0.75 14.3 3.8 4.5 11.2 3.1 10.9
123* Mesolithic Child Male 0.75 13.2 2.8 3.5 10.5 2.7 9.9
154 Mesolithic Adult
Male 0.75 14.8 1.3 2.0 13.8 1.0 13.4
170* Mesolithic Adult Male 0.75 16.3 1.3 2.0 15.4 0.9 12.9
124* Early-/Mid
Adult Male n.d. n.d n.d. 3.6 10.8 11.9
162 Early-/Mid
Male 0.25 15.0 1.5 1.3 12.0 3.0 10.1
164* Early-/Mid
Male 0.25 14.3 2.5 2.7 11.8 2.5 11.7
173 Early-/Mid
Adult Male 0.75 17.0 2.0 2.7 13.2 3.8 12.7
179 Early-/Mid
Adult Male n.d n.d. n.d. 3.5 11.5 11.1
201* Mid
Child Unknown 0.25 17.3 2.5 2.7 11.8 5.5 12.2
206 Mid
Unknown 1.3 17.3 5.2 6.5 9.6 7.7 12.9
207 Mid
Unknown 0.75 16.9 3.8 4.5 12.8 4.1 13.3
208 Mid
Adult Male 0.25 18.4 1.5 1.3 13.6 4.8 13.6
137 Late
Adult Male 0.25 11.9 1.1 1.3 9.9 2.0 9.7
137a Late
Unknown Female 0.75 10.1 4.3 4.5 7.1 3.0 n.d.
186 Late
Child Unknown 0.25 12.1 1.1 1.3 10.4 1.7 10.9
307 Late
Juvenile Unknown 0.25 14.8 1.8 2.0 10.7 4.1 11.0
Mean (all) 0.4 14.8 2.7 3.0 11.5 3.3 11.8
0.3 2.3 1.3 1.4 1.9 1.7 1.3
Mean Mesolithic (n =5) 0.7 14.3 2.5 3.1 13.2 2.0 11.8
0.2 1.3 1.2 1.1 2.2 1.0 1.4
Mean Early-/Mid Neolithic (n =9) 0.5 16.6 2.7 3.2 11.9 4.5 12.2
0.4 1.4 1.3 1.6 1.2 1.7 1.1
Mean Late Neolithic (n =4) 0.4 12.2 2.0 3.0 11.5 2.7 11.8
0.3 1.9 1.5 1.4 1.9 1.1 1.3
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
conversely show an elevation in δ
N at this age, possibly implying the
changes in these individuals had less to do with breastfeeding and more
to do with switching to a diet much richer in aquatic foods. Other than
breastmilk, aquatic foods are the most likely source of
N enriched
values at this site.
A large degree of variability within groups has been observed across
a series of weaning studies, although the methods used also vary and
might contribute to some of the differences. For instance methods using
bone from aged babies and children suffers from the dual problems of
aging and that individuals who died early may not be representative of
those who survived to adulthood (Sandberg et al., 2014). Waters-Rist
et al. (2011) analysed long bone samples from non-adult individuals of
various ages amongst Neolithic and Early Bronze Age hunter-gatherers
at Lake Baikal. In many ways Neolithic Baikal presents a comparable
population to that of Zvejnieki, with both sharing a strong focus on
freshwater aquatic resources (though the isotope ecology of Lake Baikal
differs). For the Neolithic group, it was found that isotopically visible
weaning started late (1.01.5 yr) and was not complete until the age of
3.54.0 years. However, this is subject to the dual problems mentioned
above. Using bone collagen of adults and bone and dentine collagen of
children, Howcroft et al. (2014) found in their study of infant feeding
practices at the Neolithic Pitted Ware Culture site of Ajvide on the island
of Gotland a large variability in supplementary foods and the timing of
their introduction, possibly related to season. Infants appear to have
been weaned late at Ajvide, in the 3rd or 4th year of life (Howcroft et al.,
2014). Based on isotopic composition of infants with approximately
known age of death, Clayton et al. (2006) calculated breastfeeding for at
least 1.5 years and weaning between 2 and 4 years in hunter-gatherer
individuals with ample food sources at the Matjes River Rock Shelter
in South Africa.
Large variability in weaning practices was observed though
sequential dentine sampling in a Central Californian hunter-gatherer
population (ca. 30004000 years old), with some individuals weaned
between 1 and 2, while others were not weaned until 5 or 6 years old
(Eerkens and Bartelink, 2013). Scharlotta et al.s (2018) results using
sequential tooth dentine from Neolithic Baikal (including one of the
same sites analysed by Waters-Rist et al. (2011)) found that the majority
of individuals (60%, 18 of 28) showed what they termed ‘abrupt
weaning, completed by age 2.5 yr. The remaining 10 individuals,
however, showed a ‘gradual weaning pattern extending to age 45 yr,
with a subset of these showing a more complicated pattern of an initial
rapid drop followed by a slower decline in δ
N values. As discussed by
Scharlotta et al. (2018: 588), the interpretation of this latter group
comparable to that seen in some individuals at Zvejnieki is not
The pattern of increasing δ
N from 0 to 6 months to 2 years in in-
dividuals 124 and 179 at Zvejnieki does not appear to present the usual
‘weaning signal and suggests instead the early introduction of high-
trophic-level supplementary foods. The δ
C values show an inverse
‘mirroringtrend, which is entirely consistent with the consumption of
freshwater aquatic resources. In these two cases, there may be a link
between pottery and aquatic resources as a weaning food. This negative
correlation between δ
N and δ
C values, also shown clearly in in-
dividuals 154, 173 and 207, is presumably tracking variable consump-
tion of freshwater sh.
Relatively high δ
N values for individuals dating from the Meso-
lithic to Middle Neolithic suggest a variable, but overall important
contribution of aquatic foods. This is further conrmed by the presence
Fig. 5. Diachronic comparison of pre- and postweaning δ
N and δ
C data: Pre- and postweaning dentine and bone δ
N and δ
C of Mesolithic (blue), Early-/Mid
Neolithic (green) and Late Neolithic (red) individuals (left). Change in δ
N and δ
C during weaning (in ) (δ
– δ
) (right).
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
of a freshwater reservoir offset in the radiocarbon dates on humans
(Meadows et al., 2016). The results from previously analysed individual
165 indicate that after weaning the highest δ
N was from the M3
(representing the 1016 year age bracket, 13.4), becoming less
elevated in the adult skull bone (representing 2035 years, 12)
(Eriksson, 2007). This individual, however, is a known outlier so that
this pattern is unlikely to be typical of the population.
One other observation is the general pattern of elevated δ
N in early
infancy, dropping to below adult values in childhood, before once again
rising. This trend has been seen across multiple populations from
different time periods and varying socio-economic backgrounds (Beau-
mont et al., 2012; Eerkens et al., 2011; Fern´
andez-Crespo et al., 2018;
Henderson et al., 2014; Trueman et al., 2006), suggesting a physiolog-
ical cause. The decrease in δ
N during childhood may reect the higher
nutrient requirements of growing children compared to adults, that may
in some circumstances lead to a nitrogen decit and thus reduced
isotope fractionation (Schurr, 1998). If nitrogen (protein) is limiting,
fractionation of
N is less expressed, and thus the bodys nitrogen
pool has lower δ
N. This also raises a question about how weaning is
interpreted, if the lowest point of this decline in δ
N is interpreted as a
sign of late weaning rather than physiological causes.
6.2. Dietary change over time
The results show no difference in the age of the onset of weaning or
its cessation before and after the introduction of pottery. What variation
there is, does not appear to be related to time or cultural period
(Table 1). Hence there is no evidence that the introduction of pottery in
the Early Neolithic led to a shortening of the weaning period, nor indeed
is there any clear impact from the introduction of domestic plants and
animals in the Late Neolithic in this regard. The limited difference in
N between the Mesolithic and Early/Middle Neolithic, suggests little
change in the kinds of supplementary foods used following the intro-
duction of pottery. This is particularly interesting in that it runs contrary
to the idea that the use of ceramics in hunter-gatherer contexts was
primarily related to the more efcient exploitation of aquatic resources
(Gibbs et al., 2017; Lucquin et al., 2016; Craig et al., 2013) and facili-
tated the preparation of weaning foods (Buikstra et al., 1986; Molleson,
1994). An ethnographic study of preindustrial populations that explored
the hypothesis that weaning occurred earlier in agricultural and pastoral
populations suggested that breastfeeding was longer in hunting and
gathering groups, but that this was not explained by a lack of appro-
priate weaning foods (Sellen and Smay, 2001).
Considering childhood after the weaning period, the results indicate
that, excluding the Middle Mesolithic individual, the post-weaning
mean δ
N for the Mesolithic is lower than for the Early/Mid
Neolithic. The high δ
N indicates they may have been consuming
higher trophic level foods, likely to reect a diet with a greater contri-
bution of sh. This elevation is also reected in the infants, which given
they were consuming breastmilk is likely to reect the
foods consumed by their mothers or carers, typical of the community as
a whole. By comparison the Late Neolithic children have the least
elevated δ
N, reecting a shift towards subsistence practices based on
agricultural production rather than hunting and gathering. They also
show less variability post-weaning, consistent with an isotopically more
homogeneous diet.
6.2.1. Grave goods and diet
Statistical analysis of the data show that there is a link between the
distribution of δ
N, which differ signicantly between those with and
without pendants in their graves. Individuals without tooth pendants
consumed diets with a greater contribution of freshwater foods during
their childhoods than those who were buried with pendants. The pre-
dominance of freshwater animals in the associated settlement, including
8427 sh remains from species such as perch, bream, pike and tench
ougas, 2006), indicates these foods were readily available and
freshwater resources might have been consumed by those with a lower,
or at least different, social status. This suggests children were born into a
family group, or perhaps a stratum of society, which inuenced what
foods were available to them. Whatever factor caused this dietary dif-
ference stayed with the individual throughout his or her life, inuencing
the artefacts with which they were interred.
One possibility is that children had different diets depending on the
role their parents had in the group. For example, children whose parents
were skilled shermen might be more likely consume freshwater sh,
while children whose parents relied strongly on hunting would be more
likely to consume larger quantities of elk, game and foraged foods. The
different roles of adults within the community would therefore have
affected childhood experiences by association. The difference between
infants buried with and without pendants could be explained by the
contribution of milk from their mothers. The mothers in those families
that emphasised hunting and those that emphasised shing could have
produced milk which reected their diets, which in turn would have
inuenced their infantsdiets. Hunting and shing are being emphasised
here as they would have contributed far more protein than the plant
foods that no doubt were gathered by all families, and hence dominate
the isotopic signals. That said, the fact that the differences are seen
mainly in δ
N rather than δ
C could suggest that a contribution from
plant foods is blurring the freshwater signal, though the isotopic dif-
ference between aquatic and terrestrial foods is also greater for the
former than the latter (Fig. 2).
The presence or absence of pendants, and the observations that their
isotopic distributions differ signicantly, may reect different roles
within the population, as hunters may have worn the trophies of their
hunts as symbols of their hunting prowess. In this scenario, individuals
from hunting families were buried with tooth and bone pendants, while
families of individuals whose main skill or role was in shing were
buried without such pendants. While freshwater sh remains were not
recovered frequently in the cemetery, they were present in vast numbers
in the settlement area. Faunal remains recovered from the settlement are
thought to represent species consumed, while those from the cemetery
had a symbolic meaning (L˜
ougas, 2006). This suggests terrestrial ani-
mals and hunting were more highly valued, perhaps suggesting people
who hunted also had a different standing within the community
(Schulting et al., 2020). It cannot be argued that those lacking animal
tooth pendants had low status, as other grave goods have been found in
some of their graves, including spearheads, arrowheads, tools, amber,
pottery and lithic akes. Most graves also contained red ochre.
A further consideration is that children may have inuenced their
own diets from an early age. Children in hunter-gatherer groups often
collect food, help adults prepare foods, or take a role in adult subsistence
activities (Bird and Bliege Bird, 2000). It also seems clear from the care
taken in their graves that children were valued by their communities, as
they were buried alongside adults in the cemetery, laid out in their ne
clothes, with their own grave goods (Larsson, 2006).
7. Conclusion
We analysed carbon and nitrogen stable isotopes in rst molar
dentine sequences from individuals recovered at the Mesolithic/
Neolithic cemetery of Zvejnieki, Latvia in order to investigate childhood
diet and weaning age, population variability, changes over time and
whether there was a correlation between grave provisions and diet. We
found that in general, individuals were ‘exclusivelybreastfed from birth
until about 612 months (i.e., lacking isotopic evidence of weaning),
whereafter breastmilk was completely withdrawn by the age of 3 years.
There was a large degree of variability in weaning patterns within the
population, suggesting either different weaning strategies or differences
in the mothersdiets. Childhood diet, in the earlier phases of life at least,
was derived from high-trophic-level food sources which had a terrestrial
and freshwater component. We found no difference across the cultural
periods in the onset or cessation of weaning, either with the introduction
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
of pottery or of domestic animals, though the high degree of inter-
individual variability means that a larger sample size would be
required to assess this more robustly. The limited differences in δ
between the Mesolithic and Early/Middle Neolithic do not provide
strong support for the idea that the introduction of freshwater aquatic
resources as a weaning food in the latter periods depended on the
introduction of pottery.
There was a difference in the childhood and adult diet of those who
were buried, as adults, with and without bone and tooth pendants. While
all individuals at Zvejnieki consumed freshwater resources, those buried
with animal tooth pendants had lower δ
N values on average, sug-
gesting that they consumed proportionally more terrestrial game
compared to those interred without such pendants. The status or identity
of people at Zvejnieki appears to have been ascribed from birth and
particular economic roles within society, emphasising shing or hunt-
ing, might have been lled by people born into families specialising in
these activities.
CRediT authorship contribution statement
Rowena C. Henderson: Writing original draft, Investigation.
Gunita Zarin¸ a: Resources. Andrea Czermak: Investigation, Writing
review & editing. Rick J. Schulting: Writing review & editing. Peter
A. Henderson: Formal analysis. Dardega Legzdin¸ a: Investigation. Ilga
Zagorska: Resources. Julia Lee-Thorp: Writing review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
The authors would like to thank Peter Ditcheld for his technical
assistance, Debra Lewis for preparing the mapping and anonymous re-
viewers for very helpful feedback on an earlier version of this
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
AlQahtani, S.J., Hector, M., Liversidge, H.M., 2010. The London atlas of human tooth
development and eruption. Am. J. Phys. Anthropol. 142 (3), 481490.
Anderson, J.E., Becker, S., Guinena, A.H., McCarthy, B.J., 1986. Breastfeeding effects on
birth interval components: a prospective child health study in Gaza. Stud. Family
Plann. 17 (3), 153160.
Beaumont, J., Gledhill, A., Lee-Thorp, J., Montgomery, J., 2013. Childhood diet: A closer
examination of the evidence from dental tissues using stable isotope analysis of
incremental human dentine. Archaeometry 55 (2), 277295.
Beaumont, J., Geber, J., Powers, N., Wilson, A., Lee-Thorp, J., Montgomery, J., 2012.
Victims and survivors: stable isotopes used to identify migrants from the Great Irish
Famine to 19
century London. Am. J. Phys. Anthropol. 150 (1), 8798.
Beaumont, J., Montgomery, J., 2016. The great Irish famine: identifying starvation in the
tissues of victims using stable isotope analysis of bone and incremental dentine
collagen. PLoS One 11 (8), e0160065.
Bird, D.W., Bliege Bird, R., 2000. The ethnoarchaeology of juvenile foragers: shellshing
strategies among Meriam children. J. Anthropol. Archaeol. 19 (4), 461476.
Bocquet-Appel, J.-P., 2011. When the worlds population took off: the springboard of the
Neolithic demographic transition. Science 333 (6042), 560561.
Buikstra, J.E., Konigsberg, L.W., Bullington, J., 1986. Fertility and the development of
agriculture in the prehistoric Midwest. Am. Antiquity 51 (3), 528546.
Clayton, F., Sealy, J., Pfeiffer, S., 2006. Weaning age among foragers at Matjes River
Rock Shelter, South Africa, from stable nitrogen and carbon isotope analyses. Am. J.
Phys. Anthropol. 129 (2), 311317.
Coplen, T.B., Brand, W.A., Gehre, M., Gr¨
oning, M., Meijer, H.A.J., Toman, B.,
Verkouteren, R.M., 2006. New guidelines for δ
C measurements. Analyt. Chem. 78
(7), 24392441.
Craig, O.E., Saul, H., Lucquin, A., Nishida, Y., Tach´
e, K., Clarke, L., Thompson, A.,
Altoft, D.T., Uchiyama, J., Ajimoto, M., Gibbs, K., Isaksson, S., Heron, C.P.,
Jordan, P., 2013. Earliest evidence for the use of pottery. Nature 496 (7445),
Czermak, A., Fern´
andez-Crespo, T., Ditcheld, P., Lee-Thorp, J., 2020. An anatomically
sensitive method for sequential dentine microsampling and age-alignment for stable
isotope analyses of human teeth. Am. J. Phys. Anthropol. 173 (4), 776783.
Eerkens, J.W., Berget, A.G., Bartelink, E.J., 2011. Estimating weaning and early
childhood diet from serial micro-samples of dentin collagen. J. Archaeol. Sci. 38
(11), 31013111.
Eerkens, J.W., Bartelink, E.J., 2013. Sex-biased weaning and early childhood diet among
Middle Holocene hunter-gatherers in Central California. Am. J. Phys. Anthropol. 152
(4), 471483.
Eriksson, G., 2006. Stable isotope analysis of human and faunal remains from Zvejnieki,
in: Larsson, L., and Zagorska, I. (Eds.), Back to the Origin: New Research in the
Mesolithic-Neolithic Zvejnieki Cemetery and Environment, Northern Latvia. Acta
Archaeologica Lundensia Series 80, 183-215.
Eriksson, G., 2007. Immigrant, returnee or commuter?, in: Hårdh, B., Jennbert, K.,
Olarsson, D. (Eds), On the Road: Studies in Honour of Lars Larsson. Acta
Archaeologica Lundensia Almquist and Wisksell International, 4.
Eriksson, G., Zagorska, I., 2003. Do dogs eat like humans? Marine stable isotope signals
in dog teeth from inland Zvejnieki, in: Larsson, L., Kindgren, H., Knutsson, K.,
Loefer, D. and Akerlund, A., Mesolithic on the Move: Papers Presented at the Sixth
International Conference on the Mesolithic in Europe, Stockholm 2000. Oxbow
Books, Oxford.
andez-Crespo, T., Snoeck, C., Ordo˜
no, J., de Winter, N.J., Czermak, A., Mattielli, N.,
Lee-Thorp, J.A., Schulting, R.J., 2020. Multi-isotope evidence for the emergence of
cultural alterity in late Neolithic Europe. Science. Advances 6 (4), eaay2169.
andez-Crespo, T., Czermak, A., Lee-Thorp, J.A., Schulting, R.J., 2018. Infant and
childhood diet at the passage tomb of Alto de la Huesera (north-Central Iberia) from
bone collagen and sequential dentine isotope composition. Int. J. Osteoarchaeol. 28
(5), 542551.
Gibbs, K., Isaksson, S., Craig, O.E., Lucquin, A., Grishchenko, V.A., Farrell, T.F.G.,
Thompson, A., Kato, H., Vasilevski, A.A., Jordan, P.D., 2017. Exploring the
emergence of an ‘aquaticNeolithic in the Russian Far East: organic residue analysis
of early hunter-gatherer pottery from Sakhalin Island. Antiquity 91 (360),
Henderson, R., 2015. Early life histories: a study of past childhood diet and health using
stable isotopes and enamel hypoplasia. University of Oxford. D.Phil thesis.
Henderson, R.C., Lee-Thorp, J., Loe, L., 2014. Early life histories of the London poor
using δ
C and δ15N stable isotope incremental dentine sampling. Am. J. Phys.
Anthropol. 154 (4), 585593.
Honch, N.V., 2008. The palaeodietary implications of amino acid stable isotope analysis:
developments in the application of compound specic isotope techniques to
archaeological bone collagen. University of Oxford Doctoral Thesis.
Howcroft, R., Eriksson, G., Lid˙
en, K., 2014. Infant feeding practices at the Pitted Ware
Culture site of Ajvide, Gotland. J. Anthropol. Archaeol. 34, 4253.
Isaksson, S., Gibbs, K., Jordan, P. 2018. Maritime nomads of the Baltic Sea: ceramic
traditions, collective identities and prehistoric cuisine. In Gibbs, K. & Jordan, P.
(Eds.) 2018. Ceramic in Circumpolar Prehistory. Technology, Lifeways and Cuisine.
Archaeology of the North Series, Cambridge University Press. Pp. 63-80.
Kriiska, A., 2003. From hunter-sher-gatherer to farmer-changes in the Neolithic
economy and settlement on Estonian territory. Archaeologia Lituana. 1126.
Larsson, L.,2006. A tooth for a tooth. Tooth ornaments from the graves at the cemeteries
of Zvejnieki. Larsson, L. and Zagorska, I. (Eds), in: Back to the Origin. New Research
in the Mesolithic-Neolithic Zveknieki Cemetery and Environment, Northern Latvia.
Acta Archaeologica Lundensia Series 80.
Liiva, A., Loze, I., 1993. Mesolithic and Neolithic habitation of the eastern Baltic.
Radiocarbon 35 (3), 503506.
Lillie, A., Henderson, R., Budd, C., Potekhina, I. 2016. Factors Inuencing the
Radiocarbon Dating of Human Skeletal Remains from the Dneiper River System:
Archaeological and Stable Isotope Evidence of Diet from the Epipaleolithic to
Eneolithic Periods.
ougas, L., 2006. Animals as subsistence and bones as raw material for settlers of
prehistoric Zvejnieki. Larsson, L. and Zagorska, I. (Eds), in: Back to the Origin. New
Research in the Mesolithic-Neolithic Zveknieki Cemetery and Environment,
Northern Latvia. Acta Archaeologica Lundensia Series 80, 75-89.
Lucquin, A., Gibbs, K., Uchiyama, J., Saul, H., Ajimoto, M., Eley, Y., Radini, A., Heron, C.
P., Shoda, S., Nishida, Y., Lundy, J., Jordan, P., Isaksson, S., Craig, O.E., 2016.
Ancient lipids document continuity in the use of early hunter-gatherer pottery
through 9,000 years of Japanese prehistory. PNAS 113 (15), 39913996.
Meadows, J., B¯
erzin¸ ˆ
s, V., Brinker, U., Lübke, H., Schm¨
olcke, U., Staude, A., Zagorska, I.,
Zarina, G., 2016. Dietary freshwater reservoir effects and the radiocarbon ages of
prehistoric human bones from Zvejnieki, Latvia. J. Archaeol. Sci. Rep. 6, 678689.
Meadows, J., B¯
erzin¸ ˇ
s, V., Legzdin¸ a, D., Lübke, H., Schm¨
olcke, U., Zagorska, I., Zarin¸ a, G.,
2018. Stone-age subsistence strategies at Lake Burtnieks, Latvia. J. Archaeol. Sci. 17,
Meadows, J., Lübke, H., Zagorska, I., B¯
s, V., Cerina, A., Ozala, I., 2014. Potential
freshwater reservoir effects in a Neolithic shell midden at Rin¸ n¸ ukalns, Latvia.
Radiocarbon 56 (2).
Molleson, T., 1994. The eloquent bones of Abu Hureyra. Sci. Am. 271 (2), 7075.
Montgomery, J., Beaumont, J., Jay, M., Keefe, K., Gledhill, A., Cook, G., Dockrill, S.,
Melton, N., 2013. Strategic and sporadic marine consumption at the onset of the
R.C. Henderson et al.
Journal of Archaeological Science: Reports 44 (2022) 103496
Neolithic: increasing temporal resolution in the isotope evidence. Antiquity 87
(338), 10601072.
Nitsch, E., Humphrey, L., Hedges, R., 2011. Using stable isotope analysis to examine the
effect of economic change on breastfeeding practices in Spitalelds, London, UK.
American Journal of Physical Anthropology 146 (4), 619628.
Orlas, E., Lucquin, A., L˜
ougas, L., T¯
orv, M., Kriiska, A., Craig, O.E., 2017. The adoption of
pottery by north-east European hunter-gatherers: evidence from lipid residue
analysis. J. Archaeol. Sci. 78, 112119.
ciauskas, G., Jankauskas, R., Piliˇ
e, G., Craig, O.E., Charlton, S., Dupras, T.,
2017. The transition from foraging to farming (7000500cal BC) in the SE Baltic: a
re-evaluation of chronological and palaeodietary evidence from human remains.
J. Archaeol. Sci. 14, 530542.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., 2018. nlme: Linear and nonlinear moxed
effects models. R package version 3.1-137.
R Core Development Team, 2013. R: A Language and Environment for Statistical
Sandberg, P.A., Sponheimer, M., Lee-Thorp, J., van Gerven, D., 2014. Intra-tooth stable
isotope analysis of dentine: a step towards addressing selective mortality in the
reconstruction of life history in the archaeological record. Am. J. Phys. Anthropol.
155 (2), 281293.
Scharlotta, I., Goude, G., Herrscher, E., Bazaliiskii, V.I., Weber, A.W., 2018. Mind the
gap-assessing methods for aligning age determination and growth rate in multi-
molar sequences of dietary isotopic data. Am. J. Hum. Biol. 30 (5), 32163. https://
Schulting, R.J., Henderson, R.C., Czermak, A., Zarina, G., Zagorska, I., Lee-Thorp, J.,
2020. Unequal in death and in life? Linking burial rites with individual life histories.
Moreau, L. (Ed), in: Social inequality before farming? Multidisciplinary approaches
to the study of social organization in prehistoric and ethnographic hunter-gatherer-
sher societies, p.279-291.
Schurr, M.R., 1998. Using stable nitrogen-isotopes to study weaning behaviour in past
populations. World Archaeol. 30 (2), 327342.
Sellen, D.W., Smay, A.B., 2001. Relationship between subsistence and age at weaning in
‘preindustrialsocieties. Hum. Nat. 12 (1), 4787.
Smith, B.H., 1984. Patterns of molar wear in hunter-gatherers and agriculturalists. Am. J.
Phys. Anthropol. 63 (1), 3956.
Trueman, C.N., McGill, R.A.R., Guyard, P.H., 2006. The effects of growth rate on tissue-
diet isotopic spacing in rapidly growing animals: an experimental study with Atlantic
salmon (salmo salar). Rapid Commun. Mass Spectrometry 20 (17), 2634.
Tsutaya, T., Ishida, H., Yoneda, M., 2015. Weaning age in an expanding population:
stable carbon and nitrogen isotope analysis of infant feeding practices in the Okhotsk
culture (5th-13th centuries AD) in northern Japan. Am. J. Phys. Anthropol. 157 (4),
Waters-Rist, A.L., Bazaliiski, V.I., Weber, A., Katzenberg, M.A., 2011. Infant and child
diet in Neolithic hunter-sher-gatherers from Cis-Baikal, Siberia: intra-long bone
stable nitrogen and carbon isotope ratios. Am. J. Phys. Anthropol. 146 (2), 225241.
Zagorska, I., 1997. The rst radiocarbon datings from Zvejnieki Stone Age burial ground,
Latvia. ISKOS 11, 4246.
Zagorskis, F., 2004. Zvejnieki (Northern Latvia) Stone Age Cemetery. Archaeopress,
Zagorskis, F.A., 1987. Zvejnieku akmens laikmeta kapulauks, Zin¯
R.C. Henderson et al.
... In particular, the resolution of such longitudinal studies has been refined in the past 20 years by increasing the number of intra-tooth samples from ~ 3 to > 20 sections 22,33,34 . Thus, isotopic analysis of serial dentine is now widely accepted as a routine method for the investigation of past breastfeeding, weaning and childhood dietary practices 1,3,4,6,7,22,31,[33][34][35][36][37][38][39][40] . Here, building on our previous research of plants, human bone, hair, muscle and scalp samples from the Yingpan cemetery site in Xinjiang, China 41,42 , a detailed investigation of the breastfeeding, weaning and childhood dietary practices of this Silk Road population is carried out with the application of isotopic analysis of serial dentine of human teeth (Fig. 1). ...
Full-text available
Childhood is a unique phase in human life history, in which newborns are breastfed and weaned, and are progressively familiarized to adult diets. By investigating dietary changes from infancy to adolescence, valuable information regarding past cultural behaviors and aspects of human lives can be explored and elucidated. Here, in conjunction with published isotopic results of serial dentine (n = 21) from Yingpan Man, new δ¹³C and δ¹⁵N results are obtained from 172 samples of incremental dentine from 8 teeth of 8 individuals of the Yingpan cemetery, located in Xinjiang, China. The δ¹³C values range from – 18.2 to – 14.6‰ with a mean ± SD value of – 16.3 ± 0.9‰, and the δ¹⁵N results range between 13.4 and 19.9‰ with a mean ± SD value of 16.0 ± 1.4‰. This indicates that the childhood diets were mixtures of C3 and C4 dietary resources and were clearly influenced by breastfeeding and weaning practices. In particular, the findings indicate that there were significant inter-individual differences in terms of the timing and duration of breastfeeding and weaning practices as well as childhood dietary practices at Yingpan. For instance, three individuals were exclusively breastfed after birth, while, two individuals and Yingpan Man were not. In addition, the post-weaning diets of most Yingpan individuals were relatively stable, but one individual and Yingpan Man displayed clear evidence of increased consumption of C4 foods, likely millet, during late and post-weaning periods. Further, 7 individuals had unique dietary changes between 9 to 14 years old. Potential factors related to this are presented from the perspective of changes in social roles that might be caused by their early participation in the social division of labor or puberty and marriage.
Full-text available
The results of traceological (technological and use-wear) studies of osseous artefacts discovered in the Mesolithic burial no 57 and Neolithic burial no 164 from the cemetery in Zvejnieki, Latvia are discussed. This research identified significant differentiation in the techniques used to make animal tooth pendants deposited in the Mesolithic grave and, by contrast, identified relative uniformity in the methods used to make those from the Neolithic grave. In addition, use-wear traces were observed on many of the Mesolithic pendants and were absent from almost all Neolithic ones, indicating the Neolithic examples were probably made for funerary purposes. This research also identified that the bone points from both graves were made in a similar manner and had been used. The technological and functional similarities and differences between the artefacts from the Mesolithic and Neolithic burials are discussed in the context of their probable origin and on observations made on other collections of similar chronology in the region.
Full-text available
Conclusions: One of the main tasks of archaeologists is pattern recognition. In this paper we have focused on one particular case study, the prehistoric hunter-gatherer cemetery of Zvejnieki in northeastern Europe (Henderson et al. in prep.). A very striking and persistent pattern was highlighted, linking the presence or absence of animal tooth pendants in graves with long-term dietary histories, over millennia of use of the site. But recognizing a pattern is relatively straightforward; the real challenge lies in its interpretation, including a full and honest consideration of the possible alternative explanations. We have presented a plausible case for early socioeconomic inequality before farming in prehistoric Europe. It is not definitive, but the implications are nevertheless worth pursuing. One of these is that we are likely seeing only the proverbial tip of the iceberg. This is partly because isotopic studies of prehistoric hunter-gatherers have generally been concerned with characterizing a group’s overall diet, or at most with comparing age and sex (for an exception see Scharlotta et al. 2016). But probably a more important issue is that, even if they exist, many dietary distinctions will not be visible isotopically, since they involve foods with similar isotopic signatures (e.g., different cuts of meat). This makes cases like Zvejnieki all the more important, as it is unlikely that this was a unique situation. There are a number of large cemeteries in the European Mesolithic, with a strong tendency to be situated on the coast, lakes, and major rivers with access to aquatic resources. Finally, it is worth commenting on the apparent resilience of the social-ecological system at Zvejnieki (cf. Folke 2006). This was maintained (e.g., there is little evidence for high hunting pressure on large game) in a way that suggests that it was very stable, and apparently did not lead to attempts by those of putatively higher status to expand their control, or, if it did, then they were unsuccessful. Exploring the reasons for this falls outside the scope of this paper, but are likely to relate to the ecological restrictions on intensifying hunting, as well as to social levelling mechanisms. Social inequality always implies a tension between competing interests, one that in the case of Zvejnieki seems to have reached a long-lasting, stable balance that did not lead to escalating inequality.
Full-text available
The coexistence of cultural identities and their interaction is a fundamental topic of social sciences that is not easily addressed in prehistory. Differences in mortuary treatment can help approach this issue. Here, we present a multi-isotope study to track both diet and mobility through the life histories of 32 broadly coeval Late Neolithic individuals interred in caves and in megalithic graves of a restricted region of northern Iberia. The results show significant differences in infant- and child-rearing practices, in subsistence strategies, and in landscape use between burial locations. From this, we posit that the presence of communities with distinct lifestyles and cultural backgrounds is a primary reason for Late Neolithic variability in burial location in Western Europe and provides evidence of an early “them and us” scenario. We argue that this differentiation could have played a role in the building of lasting structures of socioeconomic inequality and, occasionally, violent conflict.
Full-text available
The Neolithic in north-east Asia is defined by the presence of ceramic containers, rather than agriculture, among hunter-gatherer communities. The role of pottery in such groups has, however, hitherto been unclear. This article presents the results of organic residue analysis of Neolithic pottery from Sakhalin Island in the Russian Far East. Results indicate that early pottery on Sakhalin was used for the processing of aquatic species, and that its adoption formed part of a wider Neolithic transition involving the reorientation of local lifeways towards the exploitation of marine resources.
Full-text available
Our knowledge of the timing and completeness of the transition from foraging, fishing and hunting to food production in boreal northeastern Europe is far from clear. Here, we present new bone collagen AMS 14C dates, and δ13C and δ15N isotope values for 20 humans and 17 animals from a 6500-year period dating from the Late Mesolithic to the Bronze Age in Lithuania. AMS 14C dates revealed large discrepancies in comparison to previously obtained radiocarbon dates, thus highlighting the need to re-date all prehistoric human remains where chronology was based on 14C dating of bone collagen. Stable isotope data indicate that inland Mesolithic-Subneolithic hunter-gatherers (7000–3000 cal BC) relied on a balance of freshwater food and game animals with regard to protein intake. The coastal Subneolithic groups (ca. 3000 cal BC) relied heavily on lagoon fishing, while seals and forest game were of lesser importance. Animal husbandry, most likely of sheep or goats, was a main source of protein for Neolithic Corded Ware Culture people (2900–2400 cal BC), although a significant contribution of freshwater food is also evident. Significant intra-individual variation in stable isotope values may demonstrate that a highly flexible subsistence strategy was adopted by the CWC people. Unusually high δ13C values indicate that millet had been already introduced into the farming economy of the Late Bronze Age around 1000 cal BC.
Archaeological investigations over the past 50 years have challenged the importance of domestication and food production in the emergence of institutionalized social inequality. Social inequality in the prehistoric human past developed through multiple historical processes that operate on a number of different scales of variability (e.g. social, economic, demographic, and environmental). However, in the theoretical and linguistic landscape of social inequality, there is no clear definition of what social inequality is. The lifeways of hunter-gathererfisher societies open a crucial intellectual space and challenge to find meaningful ways of using archaeological and ethnographic data to understand what social inequality exactly is with regard to variously negotiated or enforced cultural norms or ethoses of individual autonomy. This interdisciplinary edited volume gathers together researchers working in the fields of prehistoric archaeology and cultural and evolutionary anthropology. Spanning terminal Pleistocene to Holocene archaeological and ethnographic contexts from across the globe, the nineteen chapters in this volume cover a variety of topics organized around three major themes, which structure the book: 1) social inequality and egalitarianism in extant hunter-gatherer societies; 2) social inequality in Upper Palaeolithic Europe (c. 45,000–11,500 years ago); 3) social inequality in prehistoric Holocene hunter-gatherer-fisher societies globally. Most chapters in this volume provide empirical content with considerations of subsistence ecology, demography, mobility, social networks, technology, children’s enculturation, ritual practice, rock art, dogs, warfare, lethal weaponry, and mortuary behaviour. In addition to providing new data from multiple contexts through space and time, and exploring social diversity and evolution from novel perspectives, the collection of essays in this volume will have a considerable impact on how archaeologists define and theorize pathways both towards and away from inequality within diverse social contexts.
Objectives: Creating multi-tooth sequences of micro-sampled stable isotope (SI) analytical data can help track 20+ years of individual dietary history. Inferences about individual and population level behavioral patterns require cross-calibration of the timing of dietary changes recorded by each tooth. Dentin sections from contemporaneous tissues (eg, in M1 and M2) reflect dietary signature for the time of growth. Contemporary sections should produce similar values, allowing alignment of temporally overlapping portions of teeth into multi-tooth sequences. Published methods for determining the ages of incremental sections do not provide guidance for adjustment when poor alignment between individual tooth sequences is encountered. The primary objective is to address this problem; examine cause(s), assess the effects of the standard growth-model on available age-assessment techniques, and provide a viable solution. Methods: Investigating difficulty in aligning a 3-molar sequence at Shamanka II, an Early Neolithic (7000-5700 BP) Kitoi hunter-gatherer cemetery in Cis-Baikal, Siberia, we employed 10 age assessment models and 13 variants of 2 published growth rate methods on 3 individuals of different age and sex. Results: At Shamanka II, dentin initiation and/or growth rates were different from the mostly European, reference populations used to create published age-estimation/growth rate models. Initiation ages for M2 and M3 were delayed. Root formation rates were on the rapid end of known development parameters. Conclusions: Age-assessment methods customized to dentin initiation ages and growth parameters of Siberian populations produced a hybrid growth rate model for dentin section ages and improved alignment for multi-tooth SI sequences over published models.
The Rioja Alavesa region of north‐central Spain witnessed great demographic pressure and social unrest, manifested as widespread violent conflict, during the Late Neolithic (ca. 3500‐2900 cal BC). Drawing upon the ethnographic literature, it is possible that this situation impacted upon child‐rearing practices, both through food shortages and differential parental investment, favoring male infants. Here, carbon (δ13C) and nitrogen (δ15N) stable isotope measurements from bone collagen of 17 juveniles and from 163 serial microsamples of dentine from first and second molars of seven adults from the site of Alto de la Huesera are used to examine breastfeeding, weaning and childhood diets. Bone and dentine collagen δ13C and δ15N isotope values both decrease from infancy to early childhood and increase slightly towards adolescence, but dentine provides a more time‐sensitive means of monitoring dietary changes. High δ13C and δ15N values compatible with exclusive breastfeeding are detected up to ca. 1 year, with a significantly shorter duration among males, suggesting differential sex‐related parental strategies from infancy. This is tentative given the small number of individuals being compared, but does suggest that further work would be worthwhile. A gradual decline in both δ13C and δ15N, compatible with the weaning process, is then observed up to ca. 4 years in both sexes. This delayed cessation of nursing is interpreted as a possible response to food shortage. With regard to post‐weaning patterns, shifts to lower δ13C and δ15N values in females at around age 9‐11, and a general progressive increase in both isotope values from childhood to adolescence, are detected. These could be linked either with differential protein intake due to social age‐related nutritional practices or to physiological demand. The comparison between bone and dentine values shows differences between survivors and non‐survivors in both isotopes, so that assessments based on deceased children may be biased by their potentially compromised health status.
Pottery was adopted by hunter-gatherers in the Eastern Baltic at the end of the 6th millennium cal BC. To examine the motivations for this cultural and technological shift, here we report the organic residue analysis of ceramic vessels from the earliest pottery horizon (Narva) in this region. A combined approach using GC-MS, GC-C-IRMS and bulk IRMS of residues absorbed into the ceramic and charred surface deposits was employed. The results show that despite variable preservation, Narva ceramic vessels were preferentially used for processing aquatic products. We argue that pottery was part of a new Late Mesolithic subsistence strategy which included more intensive exploitation of aquatic foods and may have had important implications, such as increased sedentism and population growth.