https://orcid.org/0000-0002-7034-4322
Figures
Abstract
Cattle were the predominant domestic animal in the Iron Age and Roman Netherlands, yet their management is still incompletely understood. Some aspects of cattle management, such as birth season and the provision of fodder, have received little or no attention so far. This paper is the first to investigate these aspects for the Iron Age and Roman Netherlands, through a case study of the site of Houten-Castellum. The rural settlement of Houten-Castellum was inhabited from the Middle Iron Age to the Middle Roman period, allowing a comparison between the Iron Age and Roman period. Excavations at this site have yielded a large, well-preserved animal bone assemblage. This paper investigates cattle husbandry by using an integrated approach, combining a multi-isotope analysis (oxygen, carbon and strontium) with archaeozoological and archaeobotanical results from Houten-Castellum and comparing the results with archaeobotanical evidence for fodder and evidence for dairy use for the Iron Age and Roman Netherlands in general. While our data set is small and results must therefore be interpreted cautiously, there is convincing evidence for an extended birth season in the Middle Iron Age, as well as the use of fodder.
Citation: Groot M, Albarella U, Eger J, Evans J (2021) Cattle management in an Iron Age/Roman settlement in the Netherlands: Archaeozoological and stable isotope analysis. PLoS ONE 16(10): e0258234. https://doi.org/10.1371/journal.pone.0258234
Editor: Peter F. Biehl, University at Buffalo - The State University of New York, UNITED STATES
Received: July 18, 2021; Accepted: September 21, 2021; Published: October 1, 2021
Copyright: © 2021 Groot et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting information files.
Funding: This work was supported by a Marie Sklodowska-Curie grant funded by the European Union’s Horizon 2020 research and innovation programme (https://ec.europa.eu/programmes/horizon2020/en/home) (Grant agreement No 740394) (MG). We acknowledge support by the Open Access Publication Fund of the Freie Universität Berlin. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Iron Age society in the central part of the Netherlands was based on mixed farming with a strong pastoral component [1]. This ties in with the river landscape, which offered limited space for growing cereals and other crops but plentiful pasture for livestock [2]. Society was egalitarian and self-sufficient in terms of food [1, 3, 4]. Archaeozoological research has shown the predominance of cattle compared to other farm animals [1, 5, 6]. Cattle supported arable farming by traction and manure, provided food in the form of meat and dairy products and raw materials for clothing and artefacts; they also had an important social and status significance [6, 7]. Recent studies have shown that mobility of livestock, including cattle, played a significant role in the study area during the Iron Age [8, 9]. Mobility can be related to the role of cattle as an exchange medium, cattle raiding and the movement of human groups—almost certainly with their livestock [10].
The incorporation of the southern half of the Netherlands into the Roman Empire had a major and widespread impact on farming; this included the introduction of new crops and animal species, a size increase in livestock, intensification of agriculture, a higher degree of specialisation and better organisation of storage and transport of surplus goods [2, 11, 12]. The agrarian economy was transformed from mainly self-sufficient to surplus-producing. An early monetary economy facilitated transactions between the agricultural producers and the urban and military consumers, and a more distinct separation between town and countryside took place. There is evidence that some food was supplied from outside the region [13–15], but most of the meat is assumed to have come from local supply, with cattle as the main meat provider in the Roman towns and army camps. Cattle were also important as transport animals. There is little evidence for mobility towards the rural site of Houten-Castellum during the Roman period, which is not unexpected, as livestock is more likely to have left the settlement to supply army camps or towns [8]. This is supported by strontium isotope results for cattle from the Roman town of Heerlen, which show a high degree of mobility for cattle, which would have supplied the town with meat, but probably also included draught cattle sent for slaughter at the end of their working lives [8].
Although cattle were hugely important in both Iron Age and Roman societies, their day-to-day management is still incompletely understood. The hypsodont structure (high-crowned with enamel extending beyond the gum line) of the herbivore tooth provides the opportunity for high-resolution analysis of seasonal behaviour, which is commonly used in dietary and seasonality studies [16, 17]. Stable isotope analyses in the Netherlands have so far mostly focused on humans [18–21], with a few exceptions [18, 22]. This paper will investigate cattle management in the Iron Age and Roman Netherlands through stable isotope analysis (oxygen and carbon) of cattle teeth from the rural settlement of Houten-Castellum. The results will be integrated with previously published strontium isotope analysis carried out for the same site [8], resulting in a multi-isotope analysis. This will provide insight into aspects of cattle management that are hard to investigate through other methods. Our main questions for this paper are:
- Was there a limited or extended birth season for cattle? What evidence is there for dairying in the site of Houten-Castellum?
- What information do carbon isotopes from cattle tooth enamel provide on seasonal variation in cattle diet? Do they indicate foddering of cattle? Do they indicate the type of pasture?
The results from the stable isotope analysis will be combined with archaeozoological results from Houten-Castellum, which provide further information on cattle exploitation. We will then look at archaeobotanical evidence for cattle diet and fodder and evidence for dairy use (cattle mortality profiles and sex ratios and organic residue analysis), both at Houten-Castellum and more generally, in the Iron Age and Roman Netherlands.
Material and methods
Archaeozoological methods
To establish mortality profiles for cattle from Houten-Castellum, both mandibular tooth eruption and wear and epiphyseal fusion data were considered. Grant’s [23] method was used for tooth eruption and wear. Only complete mandibles were included. The mandible wear stages were assigned to the age categories defined by Halstead [24, 25]. Epiphyseal fusion data were assigned to different stages to calculate the percentage of animals killed in five categories: younger than 1 year, 1–2 years, 2–3 years, 3–4 years and older than 4 years [26–29]. Sex ratios were established in two ways. First, the shape of the pubic bone was used to assign fragments to males or females and second, biometrically through plotting the distal breadth of the trochlea of the humerus and assessing the distribution. The results from the Early and Middle Roman period have been combined since the Middle Roman phase only covers the early part of that period (until circa AD 120).
Theoretical principles behind oxygen and carbon
The δ18oxygen values in the carbonate fraction of tooth enamel are determined by the values in drinking water and water in food, which are related to values in precipitation [30]. The isotopic composition of precipitation is determined by several factors, including temperature, distance from the coast, latitude, altitude and the amount of precipitation [31, 32]. Since these factors vary geographically, δ18O values also vary geographically. In addition to geographical variation, δ18O values vary seasonally. In a temperate climate, summers are typically warmer and drier (leading to higher δ18O values) and winters colder and wetter (leading to lower δ18O values). The δ18O values in the carbonate fraction of tooth enamel reflect climatic conditions during tooth formation. The hypsodont cheek teeth of herbivores allow sequential sampling of enamel from top (occlusal surface—where tooth formation starts) to bottom (the enamel-root junction; ERJ—where tooth formation ends), which provides a record of δ18O values over the period during which the tooth formed. This can provide insight into seasonal changes, not just of the δ18O values, but also of conjointly measured δ13carbon values. Sequential sampling of molars is generally expected to show a sinusoidal curve of δ18O values, with the maximum reflecting summer and the minimum winter.
The δ13carbon values in the carbonate fraction of tooth enamel reflect diet, which for cattle consists of vegetation. The δ13C values of plants depends on the type of photosynthetic pathway (C3 or C4 plants), the part of the plant, water availability/aridity, salinity, altitude and season [33, 34]. In addition, the canopy effect leads to lower δ13C values in plants growing in closed forest [33, 35]. The mean δ13C value for C4 plants is around -13‰, while it is around -27‰ for C3 plants [34–37]. The δ13C enrichment from diet to tooth enamel for large ruminants in general and cattle specifically has been estimated at 14.1‰ to 16‰ [38–40]. Sequential sampling for δ13C values in cattle teeth will show seasonal changes in vegetation in free-ranging cattle, with the δ13C curve following the δ18O in temperate climates. Providing cattle with fodder or moving them to different pastures during an annual cycle can change the curve [41–45]. Flat curves suggest a uniform food supply during tooth formation [46] and could suggest the provision of fodder.
Balasse’s method
The δ18O values from sequential sampled tooth enamel can also be used to infer the length of the birthing period or to express birth seasonality for archaeological animals [47–50]. Since the timing of tooth growth is fixed within a species, inter-individual variability in the δ18O curve permits to investigate the distribution of birth in past herds. In order to estimate the birth season of cattle in this study (x0/X), the position (distance from ERJ) of the maximum δ18O value (x0) were normalized to the periodic cycle corresponding to the crown length formed potentially over a year (X). This technique eliminates differences in tooth size between individual animals. To allow an objective description, the sinusoidal δ18O sequences of the sample were modelled using the method developed by Balasse and colleagues [47, 48]. The Pearson correlation coefficient was used to measure the proximity between the modelled and raw δ18O data.
The chemical preparation and isotope analysis of carbon and oxygen in structural carbonate
For the isotope analysis of phosphate carbonate oxygen, approximately 3 milligrams of prepared enamel was loaded into a glass vial and sealed with septa. The vials are transferred to a hot block at 90°C on the GV Multiprep system. The vials are evacuated and 4 drops of anhydrous phosphoric acid are added. The resultant CO2 was collected cryogenically for 14 minutes and transferred to a GV IsoPrime dual inlet mass spectrometer. The resultant isotope values are treated as a carbonate. δ18O is reported as per mil (‰) (18O/16O) normalized to the PDB scale using a within-run calcite laboratory standard (KCM) calibrated against SRM19, NIST reference material and were converted to the SMOW scale using the published conversion equation of Coplen [51]: SMOW = (1.03091 x δ18OVPDB) +30.91. Analytical reproducibility for this run of laboratory standard calcite (KCM) is for δ18OSMOW = ± 0.06‰ (1 σ, n = 40) and δ13VPDB is ± 0.03‰ (1s, n = 40). The reproducibility of the enamel analysis, based on reproducibility of duplicate pairs of samples, for δ18OVSMOW is ± 0.04‰ (1SD, n = 4).
Samples and sample preparation
A total of 10 cattle teeth are included in this study, all from the site of Houten-Castellum (Fig 1). Four left and six right lower third molars from 10 different individuals were selected. S1 Table shows the context information of the selected teeth. Teeth were selected based on the length of the crown and a roughly equal distribution over the different time periods. The 10 teeth are dated as follows: 3 from the Middle Iron Age (500–250 BC), 3 from the Late Iron Age (250–19 BC) and 4 from the Early/Middle Roman period (19 BC-AD 120). The contexts from which the teeth derive were dated based on stratigraphy, typochronology for pottery and metal finds and 14C dating on wood [52]. Two molars were excluded from modelling of δ18O sequences due to the lack of reliable maximum and minimum δ18oxygen values (MG13 and MG24).
Sites mentioned in the text are marked. The dotted line represents the Roman border (Map by Marjolein Haars, BCL Archaeological Support).
After cleaning the surface of the teeth with a dental drill, a strip of enamel was taken from the buccal side of the anterior lobe. Sequential sampling for stable isotope analysis of oxygen and carbon was carried out by taking transversal slices along the length of the strip, from apex to ERJ. These chips of enamel were then ground into powder.
Cattle lower third molars are formed between 9 and 23 months [53], with another 6 months before mineralisation is complete [16]. This means that the samples provide information on the second and early third year of life, with the oldest samples present at the apex of the tooth, which is formed first.
Archaeological background and earlier stable isotope research
Background Houten-Castellum
The archaeological site of Houten-Castellum was a rural settlement inhabited in the Iron Age and Roman period [54]. It is located in the central Dutch river area, within the border of the Roman Empire (Fig 1). The site consists of a residual channel containing settlement refuse and intentional deposits, as well as traces of habitation on the western bank of the channel. It seems to have been abandoned by the end of the 2nd century AD. The large quantity of finds includes ca. 86,000 animal bone remains [55]. Most of the finds come from layers within the residual channel, which can often be dated accurately. Due to the waterlogged conditions, the preservation of the animal bones is excellent. The site of Houten-Castellum was a self-sufficient agrarian community, probably producing a surplus of food in the Roman period. Cattle is the main animal species in all periods, with the highest proportion occurring in the Middle Iron Age [55]. Palynological data suggest that the landscape around Houten was open with scattered groves of trees in the early Middle Iron Age [56]. A decline in tree pollen indicates that the landscape became more open over time and consisted mostly of grassland and arable fields.
Earlier research: Stable oxygen and carbon isotope analysis from cattle tooth enamel in the Netherlands
The only archaeological site in the Netherlands for which sequential δ18O and δ13C analysis of cattle tooth enamel has been published is Neolithic Schipluiden [50]. The data were combined with δ13C and δ15N isotope analysis of cattle bone collagen and archaeozoological analysis with the aim to investigate cattle diet and calving season in one of the first permanently settled sites in the Netherlands. Lower third molars from eight cattle were sampled. For seven teeth, the seasonal curve was complete enough to establish the season of birth. They show that cattle births occurred over a period of 5.5 months. The extended calving season would have resulted in a year-round availability of milk.
The δ13C values for some of the Schipluiden cattle are lower than expected. Since such low values were not found in red deer and suids, they have been interpreted as evidence of the use of leafy fodder rather than grazing on waterlogged pastures [50]. In one animal, the minimum δ13C value coincided with the minimum δ18O value, showing that this value was reached in winter. Overall, the authors of the Schipluiden study believe that some cattle occasionally received leafy fodder in winter to support lactating cows and thereby extending the lactation period.
Archaeological and archaeobotanical evidence for cattle pasture and fodder
To understand cattle diet, we must investigate the potential for livestock pasture in the river area in general and more specifically in the surroundings of Houten-Castellum. Archaeobotanical research for Houten-Castellum and other settlements provide indications for landscape, vegetation and land use, as well as indications for livestock diet and the use of fodder. The use of fodder can also be deduced indirectly from the presence and size of stables and storage structures. The characteristic farmhouse in the Iron Age and Roman Netherlands is the byrehouse, which contains a stable section. In regions where wood has been preserved, stable boxes suggest that it was mainly larger livestock that was stabled. Stabled cattle would have required fodder.
Most natural woodland in the river area had disappeared before the Middle Bronze Age, resulting in an open landscape with only small patches of woodland [57, 58]. Streamridges (former river systems consisting of a river bed and river banks) were the highest parts of the landscape, with the least risk of flooding. They were used for habitation and arable fields. They are unlikely to have offered space for pasture, but cattle could have grazed stubble fields after harvest and fallow fields (weeds and crops grown from spilled seed from previous years). Botanical research for a site close to Houten-Castellum (Houten-Tiellandt) found indications for a rotation system with fallow years [59]. Insect remains from the same site suggest that livestock grazed on stubble fields [59]. Flooding reduced the available pasture in the low-lying flood basins in winter, but in summer, they provided fertile pasture for livestock. The flood basins consisted of reed marshes and damp grassland [2]. The grassland could also be used to collect hay. Pollen from mineralised manure from either cattle or pig from the settlement of Kesteren-De Woerd shows that the manure’s producer grazed on damp grassland or reed marshes, almost certainly in the flood basins [59, 60]. Remains of water plants and grassland plants found in wells in Geldermalsen-Hondsgemet could stem from either fodder or manure [61]. If fodder, then it is most likely that this was fed in summer, since water plants are not suitable for long-term storage [2]. If manure, then the plant remains indicate the type of pasture: marshy areas in the flood basins and dry grassland on streamridges or levees [61]. An additional indication of livestock diet is provided by the high proportions of water and marsh plants (compared to weeds, meadow plants and woodland plants) found in the Roman settlements of Tiel-Passewaaijse Hogeweg, Geldermalsen-Hondsgemet and Kesteren-De Woerd [2], which again can be explained by manure or fodder. Evidence for fodder was found in Wijk bij Duurstede-De Horden, where it consisted of a combination of hay, uncleaned cereals and weeds [62].
The large storage capacity in the settlement of Tiel-Passewaaijse Hogeweg (much larger than the local arable fields could supply) has been taken as an indication for the possibility of the storage of fodder [2]. The systematic use of branches and leaves as fodder is unlikely since woodland was scarce. Livestock is also unlikely to have been fed with cereals. The area available for growing cereals and other crops was limited, and any surplus is more likely to have been sold [2].
Archaeobotanical research (pollen and macro-remains) in Houten-Castellum indicates that the landscape was partly open with patches of woodland at the start of the Middle Iron Age and became progressively less wooded and more intensively used during the Middle and Late Iron Age [56]. Grasslands were an important landscape element, increasing over time. There are also indications for arable fields, which—in the Roman period at least—were on fertile soil and contained damp parts. The latter means that cereals are likely to have been grown as a summer crop, leaving stubble fields as a possible livestock food source in winter.
Indications for milk: Organic residue analysis
The use of milk can be attested through organic residue analysis of pottery, distinguishing between lipids from milk fat and body fat. Unfortunately, such research has not yet been carried out systematically for the Iron Age and Roman Netherlands. One exception, where animal protein could stem from milk, is represented by the Early Roman site of Utrecht-Leidsche Rijn 60, where organic residues from cooking pots and bowls were analysed [63]. A combination of animal protein, animal or vegetable fats and green plant parts was interpreted as either cheese or fermented milk products with herbs, or fermented fish products with leaf vegetables. Organic residue analyses for other sites in the region (Late Iron Age/Roman Tiel-Medel-Afronding and Roman Boxmeer-Sterckwijk) have not found evidence for dairy use [64, 65]. Organic residue analysis of ceramics from the Roman-period site of Uitgeest-Groot Dorregeest has identified charred animal products that are likely to be milk [66]. In summary, organic residue analysis shows evidence for dairy, but it does not seem to have been ubiquitous. However, research so far has been limited, and further research may change this impression.
Archaeozoological evidence for dairy use in the Iron Age and Roman Netherlands
Cattle mortality profiles for the Iron Age and Roman Netherlands show that few sites focused on the exploitation of milk (Fig 2). Dairy exploitation is found most often in the western part of the Netherlands. Dairy exploitation decreases in the Roman period, which is not just suggested by mortality profiles, but also by an increase in the proportion of adult male cattle. Nevertheless, even when exploitation is aimed more at other cattle products, this does not exclude the use of milk; it only means that this was not the primary exploitation aim.
n = number of sites for which the type of exploitation could be established. Mortality profiles based on tooth eruption and wear and epiphyseal fusion were analysed to establish the type of exploitation.
Results
Archaeozoology: Mortality profiles and sex ratios of cattle
Mortality profiles for cattle at Houten-Castellum, based on mandibular tooth eruption and wear, show that the largest slaughter peak occurred between 18 and 30 months (Fig 3). Mortality of young calves (0–1 months) was low: 5–8%. A higher proportion (12–17%) was killed between 8 and 18 months. There is an increase in adult and older cattle over time.
Mortality profiles based on epiphyseal fusion are only available for the Middle Iron Age and Roman period. They show mortality in the first two years of life of 10% or less, higher slaughter rates between 2 and 4 years and high survival into adulthood (Fig 4).
Sex ratios based on the shape of the pubic bone show both a predominance of females and a higher female to male ratio in the Middle Iron Age than in the Roman period (Fig 5; unfortunately, there are no data for the Late Iron Age). Measurements of the distal humerus also show a predominance of females in all periods, and although the sample size is small, the proportion of males seems to increase in the Roman period (Fig 6).
Cattle: breadth of the trochlea of the humerus.
Stable isotope analysis
Of the 10 teeth, two were identified as coming from non-local cattle by strontium isotope analysis: MG31 (Middle Iron Age) and MG04 (Late Iron Age) [8]. The other eight samples were consistent with the local signal. The results of the stable isotope analysis are summarised in Table 1 and plotted in Fig 7. All δ18O and δ13C values per tooth are reported in S2 Table. Non-local MG04 has δ18O and δ13C values that deviate from the other teeth, while MG31 does not. The δ18O values vary between 22.72 and 28.23‰ with an intra-tooth amplitude of variation of 1.62 to 2.91‰ (Table 1). The δ13C values vary between -13.86 and -11.36‰, with an intra-tooth amplitude of variation between 0.21 and 0.92‰ for eight animals and two higher values of 1.22‰ and 2.27‰ (both non-local animals) (Table 1).
Modelling of δ18O sequences measured in cattle third molars
Results for the best fit for combined variation of the equation parameters are shown in Table 2. The location of the maximum δ18O value in tooth crown (xo) varies from 0.5 to 22.5 mm from ERJ. Pearson’s r-values (between 0.87 and 0.98) confirm the similarity between the δ18O values and the modelled data. Middle Iron Age values are distributed between 0.03 and 0.74 (range of 0.71) in three specimens, born within 71% of a yearly cycle, corresponding to a birthing season of approximately eight and a half months (Fig 8). For MG06, the location of the maximum δ18O value in tooth crown (xo) is exceptional compared to all other samples (indicating a different birth season); however, although Pearson’s r-value is lower than that for the other samples, it is still high enough for this tooth to be included. Late Iron Age values are distributed between 0.01 and 0.28 (range of 0.27) in two specimens, born within 27% of a yearly cycle, corresponding to a birthing season of approximately three months. Roman values are distributed between 0.09 and 0.24 (range of 0.15) in three specimens, born within 15% of a yearly cycle, corresponding to a birthing season of approximately two months.
Discussion
Birth season
Free-ranging cattle in northwestern Europe would naturally have a limited birth season of 2–3 months, from May to June/July [47]. Our results show a birth season of 8.5 months in the Middle Iron Age, 3 months in the Late Iron Age and 2 months in the Roman period. The number of teeth sampled per period is very small, so that these figures should be seen as minimum figures. An extended season of birth suggests a deliberate extension of the period during which cows can be milked.
Organic residue analysis shows that milk or milk products were used in the Iron Age and Roman Netherlands, but dairy does not seem to have been ubiquitous. Mortality profiles for cattle in the Iron Age and Roman Netherlands show only a few sites with specialised milk production. Mortality profiles for Houten-Castellum do not show a focus on dairy products (the ones based on epiphyseal fusion even less so). The main slaughter peak between 18–30 months represents slaughter for meat of cattle whose growth had started slowing down. Mortality of calves younger than 1 month old is consistent with natural mortality without any need to resort to dairying as an explanation. The presence of this age category, as well as the presence of bones from foetal and neonatal animals, suggests that calving took place within or around the settlement. Cattle killed between 8 and 18 months (12 to 17%) could represent post-lactation slaughter. The increase in older cattle over time is a phenomenon that has been observed for this region before and is likely to be related to intensification of arable farming and not to an increase of dairy cows [12]. Sex ratios show a dominance of females and a slightly higher female to male ratio in the Middle Iron Age than in the Roman period. The higher proportion of calves killed in the Middle Iron Age compared to a slightly higher proportion of cows fits with the longer birth season and can tentatively be understood as a stronger emphasis on milk in the Middle Iron Age compared to the later periods. The lack of clear signs of specialised milk production can be explained by the fact that herds were kept in a mixed farming, subsistence system. The use of milk occurred at a small scale (household level) and was one of several products cattle provided. Spreading out births ensured that milk was available year-round. In non-specialised subsistence farming, sequential analysis of δ18O from carbonate in tooth enamel may be a favourable indicator for the targeted use of milk.
Cattle diet
The variety in carbon isotopic composition for the ten cattle that were included in this study is low. Excluding non-local MG04, which has much lower values, as well as the single deviating value from MG13, the range of values is 1.5‰. Even when these lower values are included, the range is 2.5‰, which is much less than that reported for eight cattle teeth from a Neolithic site in the west of the Netherlands, where the range is 3.5‰ [50]. The intra-tooth amplitude of variation is also low: the eight animals consistent with local origin have amplitudes of 0.2 to 0.9‰, while only the two non-local cattle have higher amplitudes of 1.2 and 2.3‰. The amplitudes for the local animals are comparable with those found in third molars from Chillingham cattle, which are feral cattle living outside year-round and are fed hay in winter [67]. While a reduced amplitude of intra-tooth variation in cattle compared to sheep seems to be typical and may be related to the longer duration of mineralization in cattle [68], some cattle do show significantly higher amplitudes (e.g. our non-local animals, but also cattle from Neolithic and Chalcolithic sites in the Netherlands, Greece and Romania [50, 68–71]).
Most δ13C values fall within the expected range of -13.2 to -11.3‰. The two exceptions are MG04 and one value for MG13. The range translates to calculated plant values with a mean of -26.2‰ (excluding MG04 and the low value for MG13; based on +14.1‰ δ13C enrichment between diet and enamel bioapatite [38]). This corresponds clearly to a diet dominated by C3 plants. It was not expected that C4 plants would have contributed to cattle diet, as there are few indigenous C4 plants in the Netherlands. The C4 crop millet was grown in the Netherlands, but only in small quantities compared to other crops and is unlikely to have been used as fodder. The results confirm that C4 plants are insignificant to cattle diet in our region during the Iron Age and Roman period.
There is also no evidence for a canopy effect, which occurs when animals graze in dense forest or are fed leaf fodder collected in dense forest. Again, this is not unexpected, since archaeobotanical research has shown that there was only limited forest in our region. Although non-local MG04 and the lowest value for MG13 have lower δ13C values than the other teeth (minima of -13.86‰ and -13.76‰, respectively), these values are nowhere near the lowest values reported by Balasse and colleagues for Neolithic Bercy, France (-14.7‰) [47].
The deviating low values of MG04 can be explained by the fact that this animal was raised in a different location (as strontium isotope analysis has shown [8]) and must have consumed a different diet and/or lived in a different environment. The strontium isotope ratio for MG13 is consistent with the local signal. The sample for strontium was taken in the same location of the tooth where the deviating δ13C value was found. However, the area with this particular strontium ratio is large and it is possible that the animal was living at another location with a similar geology but different vegetation at the time of formation of this part of the tooth.
In cattle grazing in the same pasture year-round, δ13C sequences are expected to co-vary with δ18O sequences, as vegetation will show seasonal variety in δ13C values. A lack of correlation between δ13C and δ18O sequences suggests that cattle did not graze in the same location year-round; they were either moved between pastures and/or received fodder, as was suggested for cattle from Pool, Orkney and the Upper Rhine Valley [43, 44]. Feeding animals 13C-enriched fodder (enriched because collected in summer) during winter would lead to higher values during the winter month, disturbing normal seasonal variation in δ13C sequences [41, 42].
For five of the teeth from Houten-Castellum (MG04, MG08, MG33, MG38 and MG40), the δ13C sequences seem to co-vary with the δ18O curves, although they are much less pronounced. One tooth from Houten-Castellum has a δ13C sequence that hardly fluctuates (MG24, amplitude of variation of 0.21‰). Unfortunately, we have no modern data on free-ranging cattle in our region to compare with, so we do not know how much seasonal variation to expect. δ13C curves for most of the cattle from Neolithic Schipluiden are also much less pronounced than the δ18O curves [50], so perhaps this is typical for the Netherlands. The dampened δ13C curves may indicate that these cattle received some kind of fodder during the period when the tooth was formed (second year of life), but at the moment we cannot exclude the possibility that moving between different pastures would have the same result. Grazing cattle in flood basins (including marsh vegetation) in summer may have led to a 13C-depleted diet, while grazing cattle on stubble and fallow fields in winter may have led to a 13C-enriched diet, which theoretically could cancel out naturally occurring differences in δ13C values in vegetation.
Tooth MG06 has a δ13C sequence that appears to show natural seasonal changes in vegetation by following to some extent the δ18O curve, except for one sample location, where the δ13C value is high while the δ18O value is low. This could indicate that the animal was given 13C-enriched fodder (such as hay) in winter. In tooth MG26, the δ13C sequence does not follow the δ18O curve. Provision with fodder may have caused the lack of correlation. Indeed, one of the lowest δ18O values coincide with one of the highest δ13C values, possibly suggesting that this animal received 13C-enriched fodder in the winter.
Tooth MG31 also has a δ13C sequence that does not follow the δ18O curve. This individual has a non-local strontium signal at the beginning of M3 tooth formation. Unfortunately, we do not know when the animal moved to Houten-Castellum. If this occurred during the second year of life, then perhaps the change to a different environment with different vegetation can explain the lack of correlation between the δ13C and δ18O sequences. Finally, MG13, the tooth with one very low δ13C value, does not have a complete δ18O curve, but it is striking that the lowest δ13C value is found in the location with the highest δ18O value. As mentioned above, perhaps at the time when this part of the tooth was formed this animal lived in another part of the area with the ‘local’ strontium ratio and consumed vegetation depleted in 13C compared to the vegetation consumed by most of the cattle in Houten-Castellum.
These results tentatively suggest that cattle in Houten-Castellum received fodder during their second year of life. However, we need comparative data from modern free-ranging cattle or other archaeological assemblages to confirm whether the lack of seasonal variation is typical for our region or whether it represents foddering. Lacking such data but considering the archaeological and archaeobotanical indications for the use of fodder, we can conclude that foddering was practised in the Iron Age and Roman Netherlands.
Cattle drinking water
Sequential sampling of δ18O along the tooth crown was expected to show a sinusoidal curve, reflecting seasonal variation in precipitation, with the highest values occurring in the summer and the lowest in the winter [32, 69, 72–77]. The amplitude of the δ18O curves in the cattle teeth from Houten-Castellum is relatively low. While the lowest amplitudes are comparable to the lowest values found in Schipluiden [50], more of the teeth from Houten-Castellum have low amplitudes. The teeth with the highest amplitudes include two teeth that are from non-local cattle, so this may say nothing about local conditions.
Depressed amplitudes can have different causes. The first—movement to higher altitudes or other areas with relatively lower δ18O values in the summer [78–80]–is highly unlikely for the Netherlands. The second—a homogenised water source that is not controlled by seasonal rainwater fluctuations (such as an aquifer or large lake) [73, 74]–also seems highly unlikely for our region. The availability of drinking water for livestock would not have been a problem in the river area. Cattle could have drunk from rivers, standing water in the lowest parts of the flood basins and residual channels such as the one adjacent to the settlement of Houten-Castellum. Cattle hoof prints on the muddy banks of a residual channel in Geldermalsen-Hondsgemet suggest that such channels were used to water livestock [81]. Groundwater shows reduced δ18O seasonal changes because it averages δ18O values from precipitation over a longer time period [77]. Wells are common in Iron Age and Roman settlements in our region and perhaps we should consider the possibility that drinking water for stabled cattle was drawn from wells.
Third, differences between species in physiology, metabolism, diet and enamel mineralization rates may result in dampening of isotopic sequencing in cattle compared to other species [16, 68, 70, 73, 74, 78, 82, 83]. Low amplitudes have also been found in archaeological and modern cattle teeth from Great Britain [17, 43, 84, 85], while higher amplitudes occur in cattle teeth from archaeological sites in Greece and Romania as well as in northwestern Europe (Great Britain, France and Germany [44, 45, 47, 67, 84]). This suggests that low amplitudes are not typical for cattle in general.
Conclusion
Although the sample size of ten teeth for three time periods is small, the stable isotope analysis has yielded interesting results. Sequential sampling of oxygen isotopes suggests a prolonged birth season in the Middle Iron Age. There are hints that the birth season was shorter in the Late Iron Age and Roman period but this will need to be confirmed by analysing additional samples. By combining the isotopic evidence with data from archaeological, archaeozoological and residue analysis, we can conclude that milk was used, and that milk was more important in the Middle Iron Age than in the later periods. Milk was, however, only one of several products that cattle were exploited for, which is why clear archaeozoological signals for specialised milk production are rare and organic residue analysis has not found dairy in all sites studied. Dairy exploitation was not the main focus of cattle husbandry, but rather occurred at household level within a self-sufficient farming system. If there was indeed a decreased prominence of dairy exploitation in the Late Iron Age and Roman paper, this can perhaps be related to an increased emphasis on other products, such as meat and traction, stimulated by the demand for food by the Roman urban and military markets. Sequential sampling of carbon isotopes suggests that the provision of fodder to subadult cattle was practised in the Iron Age and Roman period. This confirms results from archaeological and archaeobotanical research. The data set in this study is small but the current dearth of parallel evidence from the Netherlands means that, although the results are tentative, they have the potential to stimulate further research in this area.
Supporting information
S1 Table. Context information on the specimens used in this study.
https://doi.org/10.1371/journal.pone.0258234.s001
(PDF)
S2 Table. Intra-tooth oxygen and carbon isotope ratios of enamel from cattle teeth from Houten-Castellum.
Wear stages after Grant (1982) [23].
https://doi.org/10.1371/journal.pone.0258234.s002
(PDF)
Acknowledgments
We would like to thank Hilary Sloane (NERC Isotope Geosciences Laboratory, British Geological Survey) for running the samples. We would also like to thank Mirella de Jong (Provinciaal Depot voor Bodemvondsten Utrecht) for permission to sample teeth and Martijn van Haasteren, Edda Wijnans and Jan van Renswoude (VUhbs) for their help in finding the teeth.
References
- 1.
Brinkkemper O, Van Wijngaarden-Bakker L. All-round farming. Food production in the Bronze Age and the Iron Age. In: Louwe Kooijmans LP, Van den Broeke PW, Fokkens H, Van Gijn AL, editors. The prehistory of the Netherlands. Amsterdam: Amsterdam University Press; 2005. pp. 491–512.
- 2. Groot M, Kooistra LI. Land use and the agrarian economy in the Roman Dutch River Area. Internet Archaeology. 2009;27. https://intarch.ac.uk/journal/issue27/groot_toc.html.
- 3.
Van den Broeke P. Increasing diversity: synthesis. In: Louwe Kooijmans LP, Van den Broeke PW, Fokkens H, Van Gijn AL, editors. The prehistory of the Netherlands. Amsterdam: Amsterdam University Press; 2005. pp. 683–692.
- 4.
Schinkel K. Hamlets on the move. Settlements in the southern and central parts of the Netherlands. In: Louwe Kooijmans LP, Van den Broeke PW, Fokkens H, Van Gijn AL, editors. The prehistory of the Netherlands. Amsterdam: Amsterdam University Press; 2005. pp. 519–541.
- 5. Van Dijk J. Iron Age animal husbandry in the wetlands of the western Netherlands. Environmental Archaeology. 2016;21(1): 45–58.
- 6.
Roymans N. Man, cattle and the supernatural in the Northwest European plain. In: Fabech C, Ringtved J, editors. Settlement and landscape. Proceedings of a conference in Århus, Denmark, May 4–7 1998. Højberg: Jutland Archaeological Society. 1999. pp. 291–300.
- 7.
Van Dijk J, Groot M. The Late Iron Age-Roman transformation from subsistence to surplus production in animal husbandry in the Central and Western parts of the Netherlands. In: Groot M, Lentjes D, Zeiler J, editors. Barely surviving or more than enough? The environmental archaeology of subsistence, specialisation and surplus food production. 2013. Leiden: Sidestone Press. pp. 175–200.
- 8. Groot M, Evans J, Albarella U. Mobility of cattle in the Iron Age and Roman Netherlands, Journal of Archaeological Science: Reports. 2020;32: 102416.
- 9. Kootker LM, Geerdink C, Van den Broeke PW, Kars H, Davies GR. Breaking traditions: An isotopic study on the changing funerary practices in the Dutch Iron Age (800–12 BC). Archaeometry. 2017;60(3): 594–611.
- 10.
Roymans N. Ethnic identity and imperial power. The Batavians in the Early Roman Empire. Amsterdam: Amsterdam University Press (Amsterdam Archaeological Studies 10); 2004.
- 11. Groot M, Heeren S, Kooistra L, Vos W. Surplus production for the market? The agrarian economy in the non-villa landscapes of Germania Inferior. Journal of Roman Archaeology. 2009;22: 231–252.
- 12.
Groot M. Livestock for sale. Animal Husbandry in a Roman Frontier Zone. Amsterdam: Amsterdam University Press (Amsterdam Archaeological Studies 24); 2016.
- 13. Kooistra LI. The provenance of cereals for the Roman army in the Rhine delta, based on archaeobotanical evidence. Beihefte der Bonner Jahrbücher. 2009;58(1): 219–237.
- 14.
Kooistra LI. Food for soldiers: farm deliveries from Germania Inferior in the second and third centuries AD. In: Ivleva T, De Bruin J, Driessen M, editors. Embracing the provinces: society and material culture of the Roman frontier regions. Essays in honour of Dr Carol van Driel-Murray. Oxford: Oxbow; 2018. pp. 111–120.
- 15. Pals JP, Hakbijl T. Weed and insect infestation of grain cargo in a ship at the Roman fort of Laurium in Woerden (Province of Zuid-Holland). Review of Palaeobotany and Palynology. 1992;73: 287–300.
- 16. Balasse M. Reconstructing dietary and environmental history from enamel isotopic analysis: time resolution of intra-tooth sequential sampling. International Journal of Osteoarchaeology. 2002;12: 155–165.
- 17. Towers J, Jay M, Mainland I, Nehlich O, Montgomery J. A calf for all seasons? The potential of stable isotope analysis to investigate prehistoric husbandry practices. Journal of Archaeological Science. 2011;38: 1858–1868.
- 18.
Kootker LM. Isotopenanalyse. Paleomobiliteit en paleodieet. In: Habermehl D, Van Kampen J, Van Renswoude J, editors. Opgravingen te Tiel-Medel-Hazenkamp en -De Reth. Twee grafvelden uit de Romeinse tijd en nederzettingssporen uit de Late IJzertijd en vroeg-Romeinse tijd. Amsterdam: VUhbs (Zuidnederlandse Archeologische Rapporten 80); 2019. pp. 608–620.
- 19. Smits E, Van der Plicht J. Mesolithic and Neolithic human remains in the Netherlands: physical anthropological and stable isotope investigations. Journal of Archaeology in the Low Countries. 2009;1(1): 55–85.
- 20. Smits E, Millard AR, Nowell G, Pearson DG. Isotopic investigation of diet and residential mobility in the Neolithic of the Lower Rhine Basin. European Journal of Archaeology. 2010;13(1): 5–31.
- 21.
Waters-Rist AL, Palmer J. Report of strontium and stable oxygen isotope results of human remains from the Roman fort at Velsen, The Netherlands. Leiden: Laboratory for Human Osteoarchaeology, Leiden University; 2016.
- 22. McManus E, Montgomery J, Evans J, Lamb A, Brettell R, Jelsma J. “To the land or to the sea”: diet and mobility in Early Medieval Frisia. The Journal of Island and Coastal Archaeology. 2013;8(2): 255–277.
- 23.
Grant A. The use of tooth wear as a guide to the age of domestic ungulates. In: Wilson B, Grigson C, Payne S, editors. Ageing and Sexing Animal Bones from Archaeological Sites. Oxford: BAR (BAR British Series 109); 1982. pp. 91–108.
- 24.
Halstead P. A study of mandibular teeth from Romano-British contexts at Maxey. In: Pryor F, French C, Crowther D, Gurney D, Simpson G, Taylor M, editors. Archaeology and environment in the lower Welland valley, volume 1. Cambridge: Fenland Project Committee (East Anglian Archaeology Report 27); 1985. pp. 219–224.
- 25.
Hambleton E. Animal Husbandry Regimes in Iron Age Britain: A Comparative Study of Faunal Assemblages from British Iron Age Sites. Oxford: BAR (BAR British Series 282); 1999.
- 26.
Chaplin RE. The Study of Animal Bones from Archaeological Sites. London: Seminar Press; 1971.
- 27. Uerpmann HP. Animal bone finds and economic archaeology: a critical study of ‘osteoarchaeological’ method. World Archaeology. 1973;4: 307–22. pmid:16468221
- 28.
Silver I. The ageing of domestic animals. In: Brothwell D, Higgs E, editors. Science in Archaeology. 2nd ed. Bristol: Thames and Hudson; 1969. pp. 283–302.
- 29.
Habermehl KH. Die Altersbestimmung bei Haus- und Labortieren. Berlin: Parey; 1975.
- 30. Longinelli A. Oxygen isotopes in mammal bone phosphate: A new tool for paleohydrological and paleoclimatological research? Geochimica et Cosmochimica Acta. 1984;48(2): 385–390.
- 31. Dansgaard W. Stable isotopes in precipitation. Tellus. 1964;16: 436–468.
- 32.
Rozanski K, Araguás-Araguás L, Gonfiantini R. Isotopic patterns in modern global precipitation. In: Swart PK, Lohmann KC, McKenzie J, Savin S, editors. Climate change in continental isotopic records. Washington: American Geophysical Union (Geophysical Monograph 78); 1993: 1–36.
- 33. Heaton THE. Spatial, Species, and Temporal Variations in the 13C/12C Ratios of C3 Plants: Implications for Palaeodiet Studies. Journal of Archaeological Science. 1999;26: 637–649.
- 34. Tieszen LL. Natural variations in the carbon isotope values of plants: implications for archaeology, ecology, and paleoecology. Journal of Archaeological Science. 1991;18: 227–248.
- 35. Van der Merwe NJ, Medina E. The canopy effect, carbon isotope ratios and foodwebs in Amazonia. Journal of Archaeological Science. 1991;18: 249–259.
- 36. Bender MM. Variations in the 13C/12C ratios of plants in relation to the pathway of photosynthetic carbon dioxide fixation. Phytochemistry. 1971;10: 1239–1245.
- 37. O’Leary MH. Carbon isotopes in photosynthesis. Bioscience. 1988;38: 328–335.
- 38. Cerling TE, Harris JM. Carbon isotope fractionation between diet and bioapatite in ungulate mammals and implications for ecological and paleoecological studies. Oecologia. 1999;120: 347–363. pmid:28308012
- 39. Passey BH, Robinson TF, Ayliffe LK, Cerling TE, Sponheimer M, Dearing MD, et al. Carbon isotope fractionation between diet, breath CO2, and bioapatite in different mammals. Journal of Archaeological Science. 2005;32: 1459–1470.
- 40.
Towers JR. An isotopic investigation into calving seasonality, diet and dairying in British prehistoric cattle. Reconstructing animal husbandry at a sub-annual resolution using multi-isotope analysis and intra-tooth sampling. PhD Thesis, The University of Bradford. 2013.
- 41. Makarewicz CA. Sequential δ13C and δ18O analyses of early Holocene bovid tooth enamel: Resolving vertical transhumance in Neolithic domesticated sheep and goats. Palaeogeography, Palaeoclimatology, Palaeoecology. 2017;485: 16–29.
- 42.
Makarewicz CA. Stable isotopes in pastoralist archaeology as indicators of diet, mobility, and animal husbandry practices. In: Ventresca Miller AR, Makarewicz CA, editors. Isotopic Investigations of Pastoralism in Prehistory. Abingdon: Routledge; 2018. pp. 141–158.
- 43. Towers T, Mainland I, Montgomery J, Bond J. Calving seasonality at Pool, Orkney during the first millennium AD: an investigation using intratooth isotope ratio analysis of cattle molar enamel. Environmental Archaeology. 2017;22(1): 40–55.
- 44.
Winter-Schuh C. Investigating seasonal changes of cattle diet in terrestrial C 3 biomes through the isotopic analysis of serially sampled tooth enamel. In: Ventresca Miller AR, Makarewicz CA, editors. Isotopic Investigations of Pastoralism in Prehistory. Abingdon: Routledge; 2018. pp. 29–39.
- 45. Winter-Schuh C, Schmütz K, Eberle N, Makarewicz CA. Restricted pasturing of domesticated cattle at a Late Neolithic settlement in Central Germany. Journal of Archaeological Science: Reports. 2018;22: 285–297.
- 46. Henton E, Corriston JM, Martin L, Oches EA. Seasonal aggregation and ritual slaughter: isotopic and dental microwear evidence for cattle herder mobility in the Arabian Neolithic. Journal of Anthropological Archaeology. 2014;33: 119–131.
- 47. Balasse M, Boury L, Ughetto-Monfrin J, Tresset A. Stable isotope insights (δ18O, δ13C) into cattle and sheep husbandry at Bercy (Paris, France, 4th millennium BC): birth seasonality and winter leaf foddering. Environmental Archaeology. 2012;17(1): 29–44.
- 48. Balasse M, Obein G, Ughetto-Monfrin J, Mainland I. Investigating seasonality and season of birth in past herds: a reference set of sheep enamel stable oxygen isotope ratios. Archaeometry. 2012;54: 349–68.
- 49. Towers J, Gledhill A, Bond J, Montgomery J. An investigation of cattle birth seasonality using δ13C and δ18O profiles within first molar enamel. Archaeometry. 2014;56(Suppl. 1): 208–236.
- 50. Kamjan S, Gillis RE, Çakırlar C, Raemakers DCM. Specialized cattle farming in the Neolithic Rhine-Meuse Delta: Results from zooarchaeological and stable isotope (δ18O, δ13C, δ15N) analyses. PLoS ONE. 2020;15(10): e0240464. pmid:33085689
- 51. Coplen TB. Normalization of oxygen and hydrogen isotope data. Chemical Geology (Isotope Geoscience Section). 1988;72: 293–297.
- 52.
Van Renswoude J. Sporen, structuren en fasering. In: Van Renswoude J, Habermehl DS, editors. Opgravingen te Houten-Castellum. Bewoning langs een restgeul in de IJzertijd, Romeinse tijd en Vroege Middeleeuwen. Amsterdam: VUhbs (Zuidnederlandse Archeologische Rapporten 65); 2017. pp. 89–141.
- 53. Brown WAB, Christofferson PV, Massler M, Weiss MB. Postnatal tooth development in cattle. American Journal of Veterinary Research. 1960;21: 7–34. pmid:13805043
- 54.
Van Renswoude J, Habermehl DS, editors. Opgravingen te Houten-Castellum. Bewoning langs een restgeul in de IJzertijd, Romeinse tijd en Vroege Middeleeuwen. Amsterdam: VUhbs (Zuidnederlandse Archeologische Rapporten 65); 2017.
- 55.
Groot M, Van Haasteren M. Dierlijk bot. In: Van Renswoude J, Habermehl DS, editors. Opgravingen te Houten-Castellum. Bewoning langs een restgeul in de IJzertijd, Romeinse tijd en Vroege Middeleeuwen. Amsterdam: VUhbs (Zuidnederlandse Archeologische Rapporten 65); 2017. pp. 687–733.
- 56.
Kooistra LI. Botanische materialen. In: Van Renswoude J, Habermehl DS, editors. Opgravingen te Houten-Castellum. Bewoning langs een restgeul in de IJzertijd, Romeinse tijd en Vroege Middeleeuwen. Amsterdam: VUhbs (Zuidnederlandse Archeologische Rapporten 65); 2017. pp. 747–788.
- 57.
Van Beurden L. Vegetatie en ontwikkelingen in het rivierengebied in de Bronstijd. Zaandam: BIAX Consult (BIAXiaal 331); 2008.
- 58.
Arnoldussen S. A Living Landscape. Bronze Age settlement sites in the Dutch river area (c. 2000–800 BC). Leiden: Sidestone Press; 2008.
- 59.
Kooistra LI. Borderland Farming. Possibilities and limitations of farming in the Roman Period and Early Middle Ages between the Rhine and Meuse. Assen: Van Gorcum; 1996.
- 60.
Kooistra LI, Van Haaster H. Archeobotanie. In: Sier MM, Koot CW, editors. Kesteren-De Woerd. Bewoningssporen uit de IJzertijd en de Romeinse tijd. Amersfoort: Rijksdienst voor het Oudheidkundig Bodemonderzoek (Rapportage Archeologische Monumentenzorg 82); 2001. pp. 293–359.
- 61.
Kooistra LI. Botanische materialen. In: Van Renswoude J, Van Kerckhove J, editors. Opgravingen in Geldermalsen-Hondsgemet. Een inheemse nederzetting uit de Late IJzertijd en Romeinse tijd. Amsterdam: Archeologisch Centrum Vrije Universiteit (Zuidnederlandse Archeologische Rapporten 35); 2009. pp. 411–457.
- 62.
Lange AG. De Horden near Wijk bij Duurstede. Plant remains from a native settlement at the Roman frontier: a numerical approach. Amersfoort: Rijksdienst voor het Oudheidkundig Bodemonderzoek (Nederlandse Oudheden 13); 1990.
- 63.
Oudemans TFM, Kubiak-Martens L. Residu-analyse. In: Weterings PGH, Meijer Y, editors. Op zoek naar de weg. LR60: onderzoek naar de Romeinse limesweg in De Meern (gemeente Utrecht). Utrecht: Gemeente Utrecht (Basisrapportage Archeologie 33); 2011. pp. 91–104.
- 64.
Oudemans T. Chemisch onderzoek van residuen op vroeg-Romeins handgevormd aardewerk. In: Habermehl D, Van Kampen J, Van Renswoude J, editors. Opgravingen te Tiel-Medel-Hazenkamp en -De Reth. Twee grafvelden uit de Romeinse tijd en nederzettingssporen uit de Late IJzertijd, vroeg-Romeinse tijd en laat-Romeinse tijd. Amsterdam: VUhbs (Zuidnederlandse Archeologische Rapporten 80); 2019. pp. 701–720.
- 65.
Zandboer S, Veldman HAP, Blom E. Romeinse tijd. In: Blom E, Van der Velde HM, editors. De archeologie van Boxmeer-Sterckwijck. 4500 jaar wonen, werken en begraven langs de Maas. Amersfoort: ADC Archeo-Projecten (ADC Monografie 18 / ADC Rapport 3500); 2015. pp. 361–447.
- 66. Oudemans TFM, Eijkel GB, Boon JJ. Identifying biomolecular origins of solid organic residues preserved in Iron Age Pottery using DTMS and MVA. Journal of Archaeological Science. 2007;34: 173–193.
- 67. Towers J, Bond J, Evans J, Mainland I, Montgomery J. An isotopic investigation into the origins and husbandry of Mid-Late Bronze Age cattle from Grimes Graves, Norfolk. Journal of Archaeological Science: Reports. 2017;15: 59–72.
- 68. Balasse M, Tornero C, Bréhard S, Ughetto-Monfrin J, Voinea V, Bălăşescu A. Cattle and Sheep Herding at Cheia, Romania, at the Turn of the Fifth Millennium cal BC. Proceedings of the British Academy. 2014;198: 115–142.
- 69. Vaiglova P, Gardeisen A, Buckley M, Cavanagh W, Renard J, Lee-Thorp J, et al. Further insight into Neolithic agricultural management at Kouphovouno, southern Greece: expanding the isotopic approach. Archaeological and Anthropological Sciences. 2020;12: 43.
- 70. Balasse M, Bălăşescu A, Janzen A, Ughetto-Monfrini JL, Mirea P, Andreescu R. Early herding at Măgura-Boldul lui Moş Ivănuş (early sixth millennium BC, Romania): environments and seasonality from stable isotope analysis. European Journal of Archaeology. 2013;16: 221–246.
- 71. Balasse M, Bălăşescu A, Tornero C, Fremondeau D, Hovsepyan R, Gillis R, et al. Investigating the scale of herding in Chalcolithic pastoral communities settled along the Danube River in the 5th millennium BC: A case study at Borduşani-Popina and Hârşova-tell (Romania). Quaternary International. 2017;436: 29–40.
- 72. Balasse M, Ambrose SH, Smith AB, Price TD. The seasonal mobility model for prehistoric herders in the south-western Cape of South Africa assessed by isotopic analysis of sheep tooth enamel. Journal of Archaeological Science. 2002;29: 917–932.
- 73. Fricke HC, O’Neil JR. Inter- and intra-tooth variation in the oxygen isotope composition of mammalian tooth enamel phosphate: implications for palaeoclimatological and palaeobiological research. Palaeogeography, Palaeoclimatology, Palaeoecology. 1996;126: 91–99.
- 74. Fricke HC, Clyde WC, O’Neil JR. Intra-tooth variations in d18O (PO4) of mammalian tooth enamel as a record of seasonal variations in continental climate variables. Geochimica et Cosmochimica Acta. 1998;62(11): 1839–1850.
- 75.
Gat JR. The isotopes of hydrogen and oxygen in precipiation. In: Fritz P, Fontes JC, editors. Handbook of environmental isotope geochemistry. Amsterdam: Elsevier; 1980. pp. 21–42.
- 76.
Knipper C. Die räumliche Organisation der linearbandkeramischen Rinderhaltung: naturwissenschaftliche und archäologische Untersuchungen. Oxford: BAR (BAR International Series 2305; 2011).
- 77. Pederzani S, Britton K. Oxygen Isotopes in Bioarchaeology: Principles and Applications, Challenges and Opportunities. Earth-Science Reviews. 2019;188: 77–107.
- 78. Britton K, Grimes V, Dau J, Richards MP. Reconstructing faunal migrations using intra-tooth sampling and strontium and oxygen isotope analyses: a case study of modern caribou (Rangifer tarandus granti). Journal of Archaeological Science. 2009;36: 1163–1172.
- 79. Henton E. The combined use of oxygen isotopes and microwear in sheep teeth to elucidate seasonal management of domestic herds: the case study of Çatalhöyük, central Anatolia. Journal of Archaeological Science. 2012;39: 3264–3276.
- 80.
Hermes T, Pederzani S, Makarewicz CA. Ahead of the curve? Implications for isolating vertical transhumance in seasonal montane environments using sequential oxygen isotope analyses of tooth enamel. In: Ventresca Miller AR, Makarewicz CA, editors. Isotopic Investigations of Pastoralism in Prehistory. Abingdon: Routledge; 2018. pp. 57–76.
- 81.
Van Renswoude J. Sporen en structuren. In: Van Renswoude J, Van Kerckhove J, editors. Opgravingen in Geldermalsen-Hondsgemet. Een inheemse nederzetting uit de Late IJzertijd en Romeinse tijd. Amsterdam: Archeologisch Centrum Vrije Universiteit (Zuidnederlandse Archeologische Rapporten 35); 2009. pp. 85–114.
- 82. Kohn MJ, Schoeninger MJ, Valley JW. Herbivore tooth oxygen isotope compositions: effects of diet and physiology. Geochimica et Cosmochimica Acta. 1996;60: 3889–3896.
- 83. Zazzo A, Balasse M, Patterson WP. The reconstruction of mammal individual history: refining high-resolution isotope record in bovine tooth dentine. Journal of Archaeological Science. 2006;33: 1177–1187.
- 84. Balasse M, Tresset A. Environmental constraints on the reproductive activity of domestic sheep and cattle: what latitude for the herder? Anthropozoologica. 2007;42(2): 71–88.
- 85. Mainland I, Towers J, Ewens V, Davis G, Montgomery J, Batey C, et al. Toiling with teeth: An integrated dental analysis of sheep and cattle dentition in Iron Age and Viking–Late Norse Orkney. Journal of Archaeological Science: Reports. 2016;6: 837–855.