AbstractDetecting C4 plants consumption is central to investigating animal ecology, agriculture, dietary transitions, and socio-environmental adaptations, and can be done using carbon isotope analysis. The conventional δ¹³C threshold used to identify C4 plant intake does not consider substantial ecological variability across Europe. By analyzing over 4,000 δ13C values from archaeological C3 and C4 grains, we present a European-wide C3 grain δ13C baseline and establish adjusted δ13C threshold estimations for C4 consumption from the site to the ecozone scale using multicomponent environmental models and ecozone cluster analysis. We show that a fixed threshold lead to under- or overestimation of C4 plant consumption, particularly in northern/humid and southern/arid regions, where the threshold needs to be revised downwards or upwards by up to 2‰. This refined framework offers a more accurate baseline for interpreting human and animal diet and enhances our understanding of the spread, adoption and consumption of C4 crops across Europe.
Similar content being viewed by others
Contracting eastern African C4 grasslands during the extinction of Paranthropus boisei
Article
Open access
30 March 2021
Isotopic evidence of high reliance on plant food among Later Stone Age hunter-gatherers at Taforalt, Morocco
Article
Open access
29 April 2024
The Pleistocene high-elevation environments between 2.02 and 0.6 Ma at Melka Kunture (Upper Awash Valley, Ethiopia) based upon stable isotope analysis
Article
Open access
19 March 2024
IntroductionEstimating the proportion of C3 versus C4 plants in human and animal diet is a key part of bioarchaeological, palaeontological and ecological research. Scholars worldwide have been investigating the spread of millet across Eurasia because this highly nutritious and drought-resistant C4 plant can address various questions about past societies1,2. This includes complex social structures, the adoption of new subsistence strategies, the adaptation to challenging climatic and environmental settings, as well as mobility and health status3,4,5,6,7. In ecology and palaeontology, identifying C3 and C4 diets provides insight into (past) habitats, niche partitioning, and animal behaviours8,9,10,11,12.Stable carbon (C) isotope analyses—paired with nitrogen (N) isotope analyses when investigating collagen—represent the most efficient and preferred method to identify C4 plant consumption from bioarchaeological and palaeontological skeletal remains. Due to different photosynthetic pathways between C3 and C4 plants, the ratio of 13C/12C isotopes (expressed as δ13C in ‰) is significantly more negative in C3 plants (−35 to −23‰) compared to C4 plants (−14 to −10‰)13,14,15. Plant δ13C is transferred to the consumer’s body tissues along the food chain, following well-described and quantified fractionation processes16,17, which leads to enriched δ13C values in the tissues of C4 plant consumers compared to C3 plant consumers. In general, bone or dentine collagen δ13C values above −18‰ and enamel or bone apatite δ13C values above −10‰ indicate a mixed diet of C3 and C4 plants13,18,19. In contrast, δ13C values above −12‰ and −4‰, respectively, represent a pure C4 diet13,18,19.However, specific environmental and climatic settings influence the plant’s δ13C value20. For instance, C3 plant δ13C values are more depleted under oceanic or Mediterranean climates, forest soil, dense canopy, elevated humidity or increased CO2 concentration20,21. Conversely, continental climate, aridity, salinity (including sea-spray effect), elevated temperature or high altitude tend to enrich plant δ¹³C values, which are the basis of the terrestrial food chain20,21. The geographical location of the investigated site is also determinant. On average, skeletal tissues can be 1 to 2‰ lower in high latitudes compared to low latitudes in Europe21,22,23. Although C4 plants react differently to environmental and climatic factors14,24, their δ13C values can vary across latitude as well25,26. This implies that the threshold value for C4 plant consumption needs to be adapted depending on the geographical location and environmental settings to avoid misinterpretations of body tissue isotopic composition. This paper fills this research gap to avoid an over- or underestimation of C4 plants dietary intakes of premodern animals and human communities across Europe.The main research questions addressed here are: (i) In which regions of Europe is it required to apply an environmentally adjusted δ13C threshold value for identifying C3 versus C4 plant consumption? (ii) What is the magnitude of this adjustment? (iii) Can we identify specific biogeographic parameters related to this isotope variability in Europe? This study draws on over 400027 published δ13C values from charred archaeological C3 and C4 grains derived from Isotope-Ratio Mass Spectrometry (IRMS)28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101. We present an innovative and, to the best of our knowledge, unprecedented ecozone-based model framework that integrates multivariate environmental data to facilitate the identification of C3 versus C4 plant diets in bioarchaeological, palaeontological and ecological research.ResultsEcozone cluster modelUsing multicomponent environmental datasets from topographical and climatic variables, we applied k-means cluster analysis to determine zones of similar environmental conditions across Europe. The model provides 20 spatial clusters including one cluster with unclassified values where not all conditions were equally met (Figs. 1 and S1, Table 1). Due to numeric quantization of k observations, we labelled the clusters based on the percentiles of the respective data ranges using expressional combinations of temperature, climatic moisture index (CMI), and topography. Some of them are geographically restricted to specific regions, such as the cold and humid ecozones 4 and 5 in North-Eastern Europe, the mild and very humid ecozone 8 at the Atlantic coast, or the hot and arid ecozones in the south of Europe, northern Africa and the Near East (e.g., clusters 10, 16 and 18). Other ecozones are spatially more scattered, for example European high mountain ranges (e.g., clusters 11 and 15). The results can be compared to the biome-based ecoregions from the literature102 despite the reduced number of modelled ecozones.Fig. 1: Site distribution over the European Ecozone clusters.20 clusters based on temperature (TMP), moisture availability (CWB), and topography (DEM) were defined using k-means cluster analysis (with k = 20), including unclassified NA values (i.e., inland water). See the methods and material section for a description of the open source TMP, CWB and DEM data and their provenience. The sites are distributed over 15 clusters. See Table 1 for the ecozone descriptions and numbering and Fig. S1 for the ecozones displayed without sites. Figure by Michael Kempf, created using the open source R and QGIS software.Full size imageTable 1 K-means cluster ecozone cluster summary table including descriptionFull size tableThe sites from which the investigated grains originate cover 15 out of 20 modelled ecozones (Fig. 1, Supplementary data 1), representing most of the European geographical and ecological diversity. However, their distribution is biased by past and modern human activity, archaeological sampling and analytical strategies. Not all ecozones are equally represented in the sample and the ecozones 3, 12, 14, 15 and 18 were excluded from the analyses due to small sample sizes.Isotope diversity among grain speciesDespite the isotopic diversity between the main C3 grain species included in this study (Fig. S2A, B, Supplementary data 2), the two dominant crops of this sample, i.e. the Triticum (wheat, n = 1923) and Hordeum (barley, n = 1843) species, are largely overlapping in the northern and southern parts of Europe (Fig. S2B, Supplementary data 2). Differences from the Central/Western European samples are mostly caused by the wide geographical area represented by this region. In the UK, however, the two crops show notably distinct δ13C values (Fig. S2B, Supplementary data 2), which possibly reflects species-specific differences (e.g., the different timing of the vegetation period, which is earlier for barley, while wheat is more impacted by the summer conditions)28,103 or the environmental diversity of the fields used to grow the different crops29. Because human and animal diet is never based on one single crop, the C3 grain sample was kept as one entity for the rest of the analyses. In contrast, the C3 and C4 grains show distinct δ13C values (Fig. S3C, Supplementary data 2), as expected from their different photosynthetic pathways13,14,15. They are thus considered separately in the rest of the study.Temporal isotope variabilityAmong the entire C3 grains dataset, there is hardly any evolution of δ13C over time (linear model [lm] R² = 0.017, p = 1.75e-17; Pearson’s r value = 0.13, p = <2.2e-16; Fig. S3, Supplementary data 2). When distinguishing between the geographical subsets UK (lm R² = 0.044, p = 0.000276; Pearson’s r value = −0.21, p = 0.000276), Southern (lm R² = 0.03, p = 3.43e-17; Pearson’s r value = 0.17, p = <2.2e-16), Central/Western (lm R² = 0.215, p = 1.80e-32; Pearson’s r value = 0.46, p = <2.2e-16) and Northern Europe (including Denmark: lm R² = 0.078, p = 4.13e-18; Pearson’s r value = 0.28, p = <2.2e-16; and excluding Denmark: lm R² = 0.059, p = 1.55e-09; Pearson’s r value = 0.34, p = <2.2e-16), the positive correlation between C3 grain δ13C values and the grain mean date is particularly weak (Fig. S4A–E, Supplementary data 2). In particular, the slightly stronger relationship observed for Central/Western Europe (Fig. S4B) is biased by the youngest samples from Central France, which exhibit particularly enriched δ13C values (Fig. S4C). At the ecozone level, a weak to moderate and significant increase in C3 grains δ13C values can be observed for ecozone 1 (lm R² = 0.359, p = 5.39e-15; Pearson’s r value = 0.60, p = 5.39e-15) and ecozone 17 only (lm R² = 0.223, p = 0.00157; Pearson’s r value = 0.47, p = 1.57e-03) (Figs. 1 and S5, Supplementary data 2). The C4 grains dataset has a small sample size and each geographical area is represented by short chronologies, which does not enable any proper diachronic analysis (Fig. S6A). The slight decrease in C4 δ13C values over time might thus be only considered statistically significant in Southern Europe despite the chronological gap of nearly a thousand years between the oldest and youngest cluster (Fig. S6B, Supplementary data 2). This implies that the C3 and the C4 grains datasets were not subdivided into different chronological phases for the subsequent analyses.Geographical isotope variabilitySplitting the C3 grain dataset into geographical subsets (UK, Northern, Southern, and Central/Western Europe) shows that the median δ13C value of C3 grains from Northern Europe is approximately 1‰ lower than that from Southern Europe (Fig. 2A, Table 2). This confirms the previous observations made on different types of samples such as faunal remains22,23 and modern plants21. Yet it has to be stressed that the standard deviation (1 SD) is quite large for both regions (±1.35 and ±1.05, respectively), implying some overlap. Despite the high latitude, C3 grains from the UK exhibit among the highest δ13C values across time (Figs. 2A and S4), which can be related to the oceanic climate22,23. In Denmark, the low δ13C values of the oldest half of the sample (c. 3700–3000 BCE) shift to particularly enriched δ13C values for the most recent half of the sample (c. 1000 BCE–1000 CE) (Fig. S7, Supplementary data 2). This might reflect changes in agricultural practices and soil management following its decrease in quality starting from the Neolithic period30,31.Fig. 2: Geographical isotopic variability in C3 grains.A C3 grains δ13C values in Europe. B C3 grain δ13C variability compared to latitudinal bins within Europe. The middle line of the box represents the median value, the box is delimited by the quartiles Q1 on the left and Q3 on the right and contains the middle half of the sample, the horizontal lines completed by the outlier dots represent the extent of the data. The mean, median, mean absolute deviation (MAD) and standard deviation (1 SD) values for each region and each latitudinal bin are listed in Table 2. The results of the one-way ANOVA tests related to (A) and (B) and of the Pearson’s correlations related to the C3 grain δ13C versus latitude for the whole dataset and for specific subsets are reported in Supplementary data 2. Figure by Margaux L. C. Depaermentier, created using the open source R software.Full size imageTable 2 Statistical summary for the C3 and C4 grain δ13C values over the latitude bins, regions, modern countries and ecozonesFull size tableUsing the whole dataset, we observe a significant decrease in C3 grain δ13C values with increasing latitude (Fig. 2B, Supplementary data 2). From the median values calculated for each latitudinal bin (Table 2, Fig. 2B), the C3 grains δ13C values from sites above 50° latitude are on average 0.54 to 1.72‰ lower than those of grains from sites at latitudes below 50° (Fig. 2B, Table 2). This confirms the mean offset of around 1–2‰ between Southern and Northern Europe. In comparison, there is a mean variation of 0.46‰ among the median δ13C values of the latitudinal bins above 50° and approximately 0.33‰ among the median δ13C values of the bins below 50°. The difference between southern and northern sites is therefore substantial, yet related to an increasing degree of variability towards the north. When excluding the UK and/or Denmark from this dataset due to the overall elevated values in these regions despite their northern latitude, the decrease in δ13C values with increasing latitude is accordingly even stronger and more significant (Pearson’s r value: −0.25 for the whole sample, −0.27 excluding UK and −0.30 excluding UK and Denmark, with a p-value < 2.2e-16 in each case; see Supplementary data 2).At the modern country level, the C3 grains from Lithuania (median −25.18 ± 1.16‰, n = 153) are on average nearly 2.5‰ lower than those from Jordan (median −22.86 ± 0.74‰, n = 46) (Fig. 3A, Table 2) which exceeds the previously defined offset of 1–2‰ between Southern and Northern Europe21,22,23. On the contrary, the Jordan sample is on average 0.80‰ more enriched than those from Greece (median −23.50 ± 0.82‰, n = 383) or Italy (median −23.50 ± 0.97‰, n = 497) despite their shared southern location. Consequently, the North–South-dichotomy is not enough to characterize the different isotopic composition of grains from diverse parts of Europe and does not account for micro-regional environmental diversity. Moreover, the northernmost countries show standard deviations (SD) of 1.29‰ on average (1.12 to 1.67‰ in total), which is sensibly more than in most of the southern (from 0.40 to 1.43‰; mean: 0.92‰) and of the central/western countries (from 0.39 to 1.51‰, mean: 0.90‰) (see the ANOVA test in Supplementary data 2).Fig. 3: Differences in archaeological charred C3 and C4 grain δ13C values in Europe.A C3 grain δ13C values over the modern countries. B C4 grain δ13C values over modern countries. C The comparison between latitude and C4 grain δ13C values shows a weak but significant negative correlation. Boxplots are defined in Fig. 2. The mean, median, MAD and SD values for each modern country are listed in Table 2. The results of the related one-way ANOVA tests and Pearson’s correlations are available from Supplementary data 2. The red labels show non-representative sample sizes (n < 10). Figure by Margaux L. C. Depaermentier, created using the open source R software.Full size imageSimilarly, the C4 grain δ13C values from Lithuania (median: −10.83 ± 0.48‰, n = 20) are on average nearly 1‰ lower than those from Greece (median: −10.19 ± 0.15‰, n = 12), France (median: −10.15‰, n = 16) or Poland (median: −10.19 ± 0.15‰, n = 9) (Fig. 3B, C; Table 2). The sample sizes from Spain (n = 3, median: −10.69 ± 0.15‰) and from the Czech Republic (n = 1) are too low to be considered representative. This trend confirms previous studies from China with depleted C4 grain δ13C values recorded at higher latitudes25,26. The pattern is further supported by the larger C4 grain δ13C dataset resulting from Alpha Magnetic Spectrometer (AMS) in Europe104, showing generally lower δ13C values in Northern compared to Southern or Central Europe (Fig. S8). However, AMS stable isotopic data lack precision due to differences in calibration compared to IRMS and provide particularly wide and unusual δ13C ranges for C4 grains105. Therefore, these data cannot be used to extend the IRMS dataset in this study. Differences in local plant genotypes are considered more likely triggers for the isotopic variability than climate and water availability24. Both C3 and C4 grains exhibit lower δ13C values in some northern regions relative to the rest of Europe, leading to regional variation in the δ¹³C threshold for identifying C4 consumption.Ecological isotopic variabilityThe geographical isotopic variability is related to environmental factors captured in the ecozone model. C3 grain samples from Lithuania (n = 153), Estonia (n = 11), Finland (n = 44), and from parts of Denmark (n = 86) fall into the subhumid temperate lowlands of North-Eastern Europe represented by ecozone 5 (n = 294). Together with ecozone 17 (n = 42)—represented by grains from Norway only—these samples exhibit the lowest median δ13C values (−25.01 ± 1.25‰ and −25.08 ± 1.20‰, respectively) for charred C3 grains in Europe (Fig. 4, Table 2). The high humidity, low temperature and low to moderately elevated topography of these ecozones can explain the depleted δ13C values20. Ecozone 20 (n = 717), represented by balanced temperate plains scattered over Europe and including samples mostly from Denmark, northern Germany and England, exhibit a much higher median δ13C value (−23.00 ± 1.16‰) and its range hardly overlaps with the other northern samples. Beyond the influence of agricultural practices mentioned above30,31, this can be explained by higher temperatures and moderate humidity characterizing this ecozone. In contrast, the 1007 samples from ecozone 17 and the 140 samples from ecozone 1 show the lowest median δ13C values among the sites below 50° latitude (−23.47 ± 0.82‰ and −23.35 ± 1.41‰, respectively). This reflects the mild and moist conditions of these transition zones at a mid-range altitude. In Southern Europe, the ecozones 7 and 13, representing warm highlands in the Mediterranean area, show the highest median δ13C values (−22.93 ± 1.35‰ and −22.56 ± 0.87‰, respectively), deriving from the drier and warmer climatic conditions. The ecozones 2, 8 and 19 are scattered over wide areas of Europe and are not related to extreme temperatures. Their isotopic ratios show intermediate values (Table 2). Ecozones 3, 12, 14, 15 and 18 cannot be included in this isotope investigation due to their small sample size.Fig. 4: C3 grains δ13C values over the European ecozones clusters.The ecozone numbers refer to the numbering in Table 1. Boxplots are defined in Fig. 2. The boxplots are ordered from top to bottom according to decreasing latitude and to increasing temperature within the ecozone. The red labels underline non-representative sample sizes (n < 10). The mean, median, MAD and SD values for each ecozone can be found in Table 2. The results of the one-way ANOVA test are available from Supplementary data 2. Figure by Margaux L. C. Depaermentier, created using the open source R software.Full size imageDiscussionBuilding on the substantial geographical and ecological variation in isotope values within C3 (and to a lesser extent C4) plants, it is essential to revise the commonly used δ13C threshold for identifying C4 consumption, such as −18.0‰ for mammal collagen (for example, ref. 7). At each investigated site, the C3 and C4 grain δ13C values from this dataset (Fig. 5A) were used to create theoretical collagen δ13C values for a diet based exclusively on these crops (Fig. 5B)—which is no realistic diet for humans or animals and was only used for a first theoretical model. This resulted in site-specific estimations for an overall C3 grain-based diet with 10% to 20% C4 grain inputs (Fig. 5B and Supplementary data 1). Our model shows that at several sites from the Baltic and Nordic countries, human or animal collagen δ13C values below −19.0‰ (with a mean SD of 0.59‰)—and up to −19.68 ± 0.94‰ at Bėlis lake, Lithuania (n = 10), for example—already reflect a low C4 input within a primarily C3-grain-based diet. In the Mediterranean, the same C4 input would result in collagen δ13C values above −17.0 ± 0.77 or −16.0 ± 0.57‰ (Fig. 5B and Supplementary data 1).Fig. 5: Point-based approach for baseline C3 grain δ13C values and estimated threshold values for C4 diet identification in mammal collagen at sites with n ≥ 10 grains.A Median C3 grain δ¹³C values (left) and related SD (right). B Median estimated threshold δ¹³C values for mammal collagen (left) and related SD (right) based on a theoretical 100%-grain-based-diet. The mean, median, SD and MAD values for each site are listed in Supplementary data 1. The same maps including even site with n < 10 are in Fig. S9. Figure by Margaux L. C. Depaermentier, created using the open source R software.Full size imageIt is generally accepted that at least 20% of dietary protein from an alternative source (such as C4 compared to C3 crops) is required to be detected in collagen106. However, with a model based on theoretical grain-based diets, a 20% C4 input produces excessively elevated δ13C estimates for mammal collagen (Supplementary data 1). This highlights the limits of a model based on non-realistic grain-based diets, as despite the important role of grains in human diet, both humans and animals have more varied diets in reality—the latter even consuming mostly other parts of the plant and only rarely grains. And with collagen δ13C values reflecting the protein component of the diet16, plant intake contributes less to collagen isotopic composition than animal-derived proteins, which may alter this threshold value in omnivorous diets. Moreover, in regions where the diet includes significant proportions of mushrooms107, forest-derived foods influenced by the canopy effect108,109, and/or freshwater fish110,111,112, consumers are exposed to more depleted δ13C values. These foods have a much stronger influence on collagen δ13C values than any enrichment from C4 plants. And since such conditions are highly plausible in the Baltics and in Scandinavia, the threshold δ13C value for C4 consumption might be even lower than those suggested by the model with 10% C4 input (Figs. 5A, B and 6). On the contrary, potential marine food consumption17 or sea-spray effects in coastal areas113 need to be considered to avoid any over-estimation of C4 plant intake.Fig. 6: Ecozone-based interpolation of the C3 grain δ13C baseline and of the estimated threshold δ13C values for detecting C4 consumers.A Median δ13C (left) and SD values (right) of charred C3 grains interpolated at the ecozone level. B Median δ13C (left) and SD values (right) of the estimated threshold ranges for C4 consumption for each ecozone. Ecozones 3, 12, 14, 15 and 18 are left grey due to their too small sample sizes. Ecozones 4, 9, 10, 11 and 15 are left grey due to the absence of data. The ecozone’s mean, median, SD and MAD values for C3 grains and for C4 consumption are listed in Table 2 and Tab. S1, respectively. Figure by Michael Kempf, created using the open source R and QGIS software.Full size imageA model involving all other food resources would go beyond the scope of this paper. But to accurately estimate the actual local threshold values for C4 consumption, it is essential to use δ13C and δ15N values from as many local and contemporaneous food resources as possible, including crops (for example32). Wild plants would further represent a better baseline for herbivore’s diets. Yet these are much more seldom in archaeological remains and their isotopic ratios are hardly represented in bioarchaeological studies so far. A model based on modern plants21 or on tree δ13C values114 can thus further serve as comparison δ13C baseline for herbivore’s diets. The combination of various isotope systems113,115,116 and/or the application of mixing models117 are further powerful approaches to disentangle the diverse dietary sources.Considerable isotopic variability is observed within each site, region and ecozone (Figs. 2–4 and 5A, Supplementary data 1–2), indicating that a range rather than a sharp threshold value is more appropriate (Figs. 5B and 6B). This isotopic variability is not only related to the gradual and complex variation of environmental settings across Europe, but is also intrinsic to the grains, as grain δ13C may vary for up to 0.5‰ within one ear despite same species and same growing conditions33. Differing growing conditions, in particular various watering practices, can further impact local grain δ13C values by up to 1.7‰24,28. Wild plants δ13C values may be good comparison references to disentangle anthropogenic and natural differences in water regimes21,118,119. Another variability might be induced by diverging analytical uncertainty between datasets, as the data result from different laboratories and protocols and were obtained with different calibrations—which are relevant information to make sure that the datasets are comparable120. Unfortunately, information on calibration, precision and accuracy of the published isotope data were mostly missing, thus it was not possible to assess the impact on the presented results. Yet when available, the error value for replicated samples was very low and still within the range of analytical errors.An evolution of the grain δ13C values might be expected over time as well, especially when considering the various climatic phases that implied a great variation in moisture and temperature throughout Europe over the investigated time frame. The presented regional differences in grain δ13C values have therefore varied between the climatic phases (Fig. S10), yet the isotopic variations within each region over the various climatic phases is overall significantly weak (Supplementary data 2). A larger dataset for the most recent periods might reveal more effective trends. In this study, the impact of the chronological depth and imbalance covered by this dataset can be considered weak (Figs. S2–S5 and S10).The isotopic difference reported in this study between regions and ecozone therefore remains strong enough to highlight environmentally-driven differences (Figs. 2–7, Table 2) even from grains that were possibly undergoing various cultivation practices. Our approach thus demonstrates that the threshold at −18.0‰ for consumer’s collagen is not universal and is mostly valid in the ecozone 2 (−17.51 ± 1.28‰), whereas the northern ecozone 5 and the widespread ecozone 17 require a threshold δ13C value closer to −19.0‰, i.e., −18.59 ± 1.12 and −18.67 ± 1.08‰, respectively. In high-temperature Mediterranean lowlands (ecozone 16), European plains (ecozone 19), and Atlantic regions (ecozone 20), the threshold shifts to −17.0‰ (i.e., −16.88 ± 0.90‰, −17.15 ± 0.74‰, and −16.71 ± 1.05‰, respectively; Figs. 6, 7, Tab. S1). In the arid southern regions (ecozone 13 [−16.31 ± 0.78‰], and partially ecozones 7 and 16), it approaches −16.0‰ (Fig. 7, Tab. S1). Yet in each case, the SD ranges from 0.74 to 1.26‰, stressing again the isotopic variability within ecozones.Fig. 7: Theoretical δ13C threshold ranges for C4 consumption across the European ecozone clusters.The ecozone numbers refer to the numbering in Table 1. Boxplots are defined in Fig. 1. The boxplots are ordered from top to bottom according to decreasing latitude and to increasing temperature within the ecozone. The red labels underline non-representative sample sizes (n < 10). The mean, median, MAD and SD values for each ecozone are listed in Tab. S1. The purple line represents the revised δ13C threshold value for C4 consumption in mammal collagen (i.e., −18.0‰). Figure by Margaux L. C. Depaermentier, created using the open source R software.Full size imageTo conclude, this paper offers both a European-wide δ13C baseline from archaeological charred C3 grains and a threshold-value-model for environmentally adjusted identification of C4 consumption from the site to the ecozone level across Europe (Fig. 7, Tab. S1). The grain baseline offers the advantage to consider a fundamental dietary resource (in particular for humans) as reference data and can be completed by local foods resources at the site level for more holistic and accurate interpretations. A comparison to wild proxies further enhances the results for animal diets. The threshold estimations are particularly suitable in bioarchaeological, ecological or palaeontological studies for which local plants or food resources are unavailable for calculating an isotopic baseline. However, it requires to account for the great degree of isotopic variability within each geographical entity as underlined by the SD values—a variability slightly increasing with latitude. In this context, the point-based approach (Fig. 5) provides more accurate yet geographically discrete data, while the interpolation-based approach (Fig. 6) offers ecologically sensitive estimates over large areas at the ecozone level, both related to some degree of uncertainty or variability. Future datasets could be used to test (and if necessary adjust) the interpolated values in areas of currently low site density. This innovative and context-sensitive ecozone clustering model based on temperature, humidity, and elevation thus enables more accurate interpretations of both animal ecologies and anthropogenic social and agricultural dynamics across Europe by avoiding over- or underestimation of C4 consumption.Methods and materialIsotopic datasetThe material used in this study consists of published δ13C values from charred grains derived from archaeological context, compiled into one single dataset27. The data was collected from 75 publications until September 202528,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101, also using the open access online repositories IsoArch121, MAIA122, Isotòpia123, IsoMedIta124 and CIMA125. In total, this represents 4,210 δ13C values of C3 and C4 grains derived from 260 sites dated between 8000 BCE and 1800 CE. The represented C3 plants are oat (Avena species, n = 58), rye (Secale species, n = 325), barley (Hordeum species, n = 1843), and wheat (Triticum species, n = 1923). Broomcorn millet (Panicum miliaceum, n = 57) and foxtail millet (Setaria italica, n = 4) represent the C4 crops from this dataset. To facilitate visualization and pattern-recognition, the C3 and C4 grain datasets were considered separately for the various analyses due to their distinct δ13C values and due to the particularly small C4 grain sample size. C3 and C4 grain δ13C values were then combined to address the question of identifying the introduction of C4 crops in C3 plants-based diets. Despite the fact that anthropogenic agricultural practices such as irrigation can impact grain δ13C values24,28, altering the natural and ecological signal, crops remain an important proxy for human diet regardless of agricultural practices, as their isotopic composition would be transferred to human tissues all the same. The ecological differences are considered important enough across Europe to be detectable from grain isotopic composition despite anthropogenic alterations. Wild plants would have represented a better proxy for animal diet, however, this proxy is lacking from archaeological contexts—or could be derived from tree δ13C values for the most recent periods114.Geographically, the research area spans modern Europe and the Mediterranean countries of the Near East, i.e., between 30 and 63° (N) latitude and between −8 and 45° (E) longitude. Yet the data is not evenly distributed and Denmark is over-represented in terms of sites, while Greece is over-represented in terms of number of samples. There are considerable gaps in several regions of Europe (see Fig. 1 and the isotopic dataset27). Chronologically, this dataset covers most archaeological and historical periods and ranges from 8000 BCE to 1800 CE. In this context, it is important to stress that the oldest samples (8000–6000 BCE) exclusively originate from Greece, whereas samples from Northern Europe are predominantly younger than 1000 BCE—except for some larger site samples dated between 4000 BCE and 2000 BCE. Modern data created in the framework of experimental archaeology were not included in the dataset because of the controlled conditions in which they were produced and because of the different present-day atmospheric composition compared to pre-industrial periods126.Only data obtained from Isotope-Ratio Mass Spectrometry (IRMS) were selected for analyses. Despite the fact that these values are not comparable to IRMS isotopic values due to different calibrations105, a European-wide dataset obtained from Accelerator Mass Spectrometry (AMS) in the context of radiocarbon measurements104 was also used as a comparison dataset to verify whether the same trend is visible among both datasets. Importantly, grain δ13C values are sometimes (i.e., in 12 out of 64 publications used in this study) published in form of corrected values to consider the charring effect on the carbon isotope composition of archaeological C3127,128 and C4 grains129,130. But following the approach by Gron et al.30, we are considering here only uncorrected values in order to enhance the comparability between datasets. This is considered to have no significant impact on this study’s results, as the charring effect is not systematic33 particularly low on grain δ13C values (i.e., 0.06 to 0.18‰33,127,128,129,130 on C3 and C4 grains for a heat up to 300 °C) and remains below analytical errors for isotope analyses131.Statistical analysesAll statistical analyses were performed using R software132 and the results and relevant values are summarised in Supplementary data 2. To determine the relationship between grain δ13C values and chronology or geographical location, we applied both Pearson’s correlation tests and linear models—the latter using the lme4 package133. The Pearson correlation coefficient (Pearson’s r value) indicated the strength and direction of a linear relationship between two tested variables, the confidence interval gives the uncertainty range for the true correlation. While in the linear model, the R-squared values show the percentage of the dataset affected by the relationship and the p-values show the significance of the results. To determine the geographical scale at which significant changes in grain values occur, the dataset was divided into various bins. At the largest scale, the main regions of Europe, split into Northern Europe, Southern Europe, Central/Western Europe and the UK. This is not only convenient but also follows expectations based on previous research21,22,23. At the smallest geographical scale, the dataset was binned according to the borders of modern countries. Boxplots were used for data visualization and one-way ANOVA tests were performed to test the difference in grain δ13C values between the investigated clusters. The results and relevant values for the ANOVA tests are summarised in Supplementary data 2). Because our dataset shows a particularly important isotopic variability, the results for each considered bin or entity/group are presented in the text using the median value (since the mean value is more sensitive to extreme outliers) and the related one standard deviation (1 SD). The tables are showing the mean, median, median absolute deviation (MAD) and 1 SD for each category/bin. In order to integrate an ecological dimension to the investigation of grain δ13C variability, the dataset was also binned into newly determined ecozones, as presented in the section below.Environmental clusterFor the cluster analysis, we used these R-packages: terra134, sf 135,136, gtools137, dplyr138 and ggplot2139, ggspatial140, gridExtra141 for plotting, as described in the R-code provided in the open access repository to this paper142. We used an unsupervised k-means clustering approach based on three spatial predictors to differentiate environmental ecozones across Europe. Components include elevation (DEM), mean temperature (T), and the Climatic Moisture Index (CMI). A DEM derived from the USGS (United States Geological Survey, Global Multi-resolution Terrain Elevation Data 2010, https://earthexplorer.usgs.gov/; last accessed 19th of June 2025). Monthly resolved climate variables for the period 1980-2018 were downloaded from CHELSA143. The CMI represents a standardized water availability index, calculated as the ratio of precipitation (P) to potential evapotranspiration (PET) with CMI = P − PET. We used CHELSA v2.1 monthly CMI data based on the Penman-Monteith equation for PET and downscaled from ERA5 reanalysis. The grids were reprojected to a meter-based projection (EPSG:3857) and cropped to the extent of the geographic European landmass. The dataset (n = 468) were aggregated to a 1000 m resolution using bilinear interpolation prior to clustering. Monthly layers were averaged to create a multiannual mean (1980–2018). To allow comparison between variables with different scales and units, all raster layers were standardized using z-score normalization with mean and standard deviation (sd) of the raster (z = cell value − mean_raster / sd_raster). A regular grid of 1 km spacing was generated across the study area and the centroid of each grid cell was calculated for regular point sampling. To reduce edge effects and avoid NA values near the coast, centroids were restricted to land areas using a simplified buffer around the European landmass boundaries. At each centroid location, the values were extracted from the normalized rasters.
K-Means clusteringK-means clustering was performed using the extracted values, with the determined number of clusters k = 20. This dimensionality was chosen to balance regional ecological resolution with model interpretability, including NA values (replaced by −99999 during the cluster analysis to protect correct geographical raster reassignment). Each centroid was assigned a cluster label based on the combined environmental profile of elevation, temperature, and moisture availability. Cluster labels were spatially joined to centroid coordinates and rasterized back into a continuous spatial layer using the DEM grid as a template. The resulting map classifies observational sites into discrete clustered ecozones (Fig. 1).To characterize each of the 20 ecozone clusters based on the data variability, we summarized the environmental properties of each cluster using the mean and standard deviation of the three input variables: Elevation (ELEV), T, and CMI. All values were normalized using z-scores prior to clustering, enabling direct comparison across variables. The clusters were then qualitatively interpreted based on their relative environmental signatures. For example, clusters with low temperatures, moderately dry conditions and high elevations were categorized as Cool|Moderately Dry|Alpine zones. Clusters with high temperatures and low moisture availability in low areas were described as Very Hot|Very Dry|Moderately Low. Intermediate clusters were labeled based on transitional or temperate climate conditions (Table 1).Determining consumer’s δ
13C valuesTo determine the theoretically expected δ13C values of consumer tissues from a mixed C3-C4 diet, we first associated each C3 grain δ13C value to a measured or assumed C4 grain δ13C value. This means that for each site at which δ13C values of both C3 and C4 grains were available, each C3 grain δ13C value got associated with the mean C4 grain δ13C value of the site. However, most of the sites included in this study provided no C4 grain. In this case, a theoretical C4 grain δ13C value was determined for the site based on the observed values from this dataset. In most regions of Europe, the C4 grain δ13C value is thus expected to be −10‰32,72,87. Yet we observed that the C4 grain δ13C values in Lithuania—and by extension presumably in the northernmost latitudes of Europe—were rather around −11‰68. Similarly, C4 grain δ13C values from the western Mediterranean area seem to be closer to −10.5‰35,47. These regional values were thus used as theoretical C4 grain δ13C values associated with the measured C3 grain δ13C values at each site lacking C4 grains (see summary in Supplementary data 1).In a second step, 5‰ was added to each measured or theoretical grain δ13C value to mimic the fractionation offset that applies between the diet and the consumer’s collagenous tissues after consumption16,17. We applied this to both C3 and C4 grains, resulting in theoretical end-members collagen δ13C values for 100% C3 and 100% C4 based diet, respectively. In a third step, we used these end-members values for each grain to create theoretical collagen δ13C values for a C3-grain-based diet including either 10% or 20% C4 grains. These three first steps were done at the grain level to minimize the loss of resolution and information when working with mean values. In a fourth step, we eventually calculated a mean δ13C value for these two types of diet at the site level (Supplementary data 1). It is fundamental to note that these fictitious diets based on 100% grains are not existing in nature and only represent a theoretical model using grains only.Because a 100% grain-based diet does not exist in nature, the model using 10% of C4 input was considered the most reliable basis for estimating the related collagen δ13C values with low C4 input in a normal mixed diet. These results are presented at the site-level to account for local variability in threshold δ13C values for C4 consumption (Fig. 5). The ecozone cluster model was used to create a map of interpolated threshold δ13C values for C4 consumption and hence suggest environmentally adjusted threshold δ13C values for C4 consumption (Fig. 6C). The European background maps used to create Figs. 5 and 6 are vector map data from https://www.naturalearthdata.com/, implemented using the package rnaturalearth in R-Software132,144.Materials & correspondenceThe corresponding authors are MLCD and MK. The isotopic dataset used in this study is available from the open access repository: Depaermentier, M. L. C. (2025). Isotopic Dataset to: Depaermentier, MLC, Kempf, M, Motuzaitė Matuzevičiūtė, G. “Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe” [Data set]. In Communications Earth & Environment. Zenodo. https://doi.org/10.5281/zenodo.17571650 [ref. 27 in this paper]. The data to reproduce the ecozone clusters is available from this open access repository: Kempf, M. (2025): Related files to: Depaermentier, MLC; Kempf, M; Motuzaitė Matuzevičiūtė, G: Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe (2025) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.15695070 [ref. 147 in this paper]. Climate variables used in this article are freely available from Karger et al. (2017): https://chelsa-climate.org/ (last accessed 19th of June 2025) [ref. 143 in this paper]. The Digital Elevation Model (DEM) can be downloaded from the USGS earthexplorer server: https://earthexplorer.usgs.gov/, last accessed 19th of June 2025.Reporting summaryFurther information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The compiled isotopic dataset used in this study is available from the open access repository: Depaermentier, M. L. C. (2025). Isotopic Dataset to: Depaermentier, MLC, Kempf, M, Motuzaitė Matuzevičiūtė, G. “Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe” [Data set]. In Communications Earth & Environment. Zenodo. https://doi.org/10.5281/zenodo.17571650 [ref. 27 in this paper]. The data to reproduce the ecozone clusters is available from this open access repository: Kempf, M. (2025): Related files to: Depaermentier, MLC; Kempf, M; Motuzaitė Matuzevičiūtė, G: Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe (2025) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.15695070 [ref. 147 in this paper]. Climate variables used in this article are freely available from Karger et al. (2017)143: https://chelsa-climate.org/ (last accessed 19th of June 2025). The Digital Elevation Model (DEM) can be downloaded from the USGS earthexplorer server: https://earthexplorer.usgs.gov/, last accessed 19th of June 2025.
Code availability
The code to reproduce the ecozone clusters is available from this open access repository: Kempf, M. (2025): Related files to: Depaermentier, MLC; Kempf, M; Motuzaitė Matuzevičiūtė, G: Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe (2025) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.15695070 [ref. 147 in this paper].
ReferencesVentresca-Miller, A. R. et al. Adaptability of millets and landscapes: ancient cultivation in North-Central Asia. Agronomy 13, 2848 (2023).Article
CAS
Google Scholar
Motuzaitė Matuzevičiūtė, G. Broomcorn millet: from the past to the future. AFF 2, 177–198 (2025).Article
Google Scholar
Hakenbeck, S. E., Evans, J., Chapman, H. & Fothi, E. Practising pastoralism in an agricultural environment: an isotopic analysis of the impact of the Hunnic incursions on Pannonian populations. PloS One 12, e0173079 (2017).Article
Google Scholar
Kaupová, S. et al. Dukes, elites, and commoners: dietary reconstruction of the early medieval population of Bohemia (9th–11th Century AD, Czech Republic). Archaeol. Anthropol. Sci. 11, 1887–1909 (2019).Article
Google Scholar
Martin, L. et al. The place of millet in food globalization during Late Prehistory as evidenced by new bioarchaeological data from the Caucasus. Sci. Rep. 11, 13124 (2021).Article
CAS
Google Scholar
Sneha, M. L. & Arjun, R. Medicinal knowledge In South India (during neolithic to early historic period): an analysis of staple plant dietary nutrition. CMDR J. Soc. Res. 1, 35–48 (2024).
Google Scholar
Lightfoot, E., Liu, X. & Jones, M. K. Why move starchy cereals? A review of the isotopic evidence for prehistoric millet consumption across Eurasia. World Archaeol. 45, 574–623 (2013).Article
Google Scholar
Drucker, D. G. The isotopic ecology of the mammoth steppe. Annu. Rev. Earth Planet. Sci. 50, 395–418 (2022).Article
CAS
Google Scholar
Drucker, D. G. et al. Ecology of large ungulates in the northeastern Iberian Peninsula during the Upper Palaeolithic through stable isotopes and tooth wear analysis. Quat. Environ. Hum. 2, 100011 (2024).
Google Scholar
Terry, R. C., Guerre, M. E. & Taylor, D. S. How specialized is a diet specialist? Niche flexibility and local persistence through time of the Chisel-toothed kangaroo rat. Funct. Ecol. 31, 1921–1932 (2017).Article
Google Scholar
Saarinen, J., Mantzouka, D. & Sakala, J. Aridity, cooling, open vegetation, and the evolution of plants and animals during the cenozoic. In Nature through Time, edited by E. Martinetto, E. Tschopp & R. A. Gastaldo, pp. 83–107 (Springer International Publishing, Cham, 2020).Prasifka, J. & Heinz, K. The use of C3 and C4 plants to study natural enemy movement and ecology, and its application to pest management. Int. J. Pest Manag. 50, 177–181 (2004).Article
Google Scholar
Cerling, T. E., Wang, Y. & Quade, J. Expansion of C4 ecosystems as an indicator of global ecological change in the late Miocene. Nature 361, 344–345 (1993).Article
Google Scholar
Farquhar, G. D. On the nature of carbon isotope discrimination in C4 species. Funct. Plant Biol. 10, 205 (1983).CAS
Google Scholar
O’Leary, M. H. Carbon isotope fractionation in plants. Phytochemistry 20, 553–567 (1981).Article
Google Scholar
Ambrose, S. H. Isotopic Analysis of Paleodiets: Methodological and Interpretative Considerations. In Investigations of ancient human tissue. Chemical analyses in anthropology, edited by M. K. Sandford, pp. 59–130 (Gordon and Breach, Philadelphia, 1993).Lee-Thorp, J. A. On isotopes and old bones. Archaeometry 50, 925–950 (2008).Article
CAS
Google Scholar
Kellner, C. M. & Schoeninger, M. J. A simple carbon isotope model for reconstructing prehistoric human diet. Am. J. Phys. Anthropol. 133, 1112–1127 (2007).Article
Google Scholar
Froehle, A. W., Kellner, C. M. & Schoeninger, M. J. Multivariate carbon and nitrogen stable isotope model for the reconstruction of prehistoric human diet. Am. J. Phys. Anthropol. 147, 352–369 (2012).Article
CAS
Google Scholar
van Klinken, G. J., Richards, M. P. & Hedges, R. E. M. An Overview of Causes for Stable Isotopic Variations in Past European Human Populations. Environmental, Ecophysiological, and Cultural Effects. In Biogeochemical approaches to paleodietary analysis. Advances in archaeological and museum science, edited by S. H. Ambrose & M. A. Katzenberg, pp. 39–63 (New York, London, 2002).Cooper, C. G., Cooper, M. D., Richards, M. P. & Schmitt, J. Geographic and seasonal variation in δ13C values of C3 plant arabidopsis: Archaeological implications. J. Archaeol. Sci. 149, 105709 (2023).Article
CAS
Google Scholar
van Klinken, G. J., van der Plicht, J. & Hedges, R. E. M. Bone 13C/12C ratios reflect (palaeo) climatic variations. Geophys. Res. Lett. 21, 445–448 (1994).Article
Google Scholar
Hedges, R. E., Stevens, R. E. & Richards, M. Bone as a stable isotope archive for local climatic information. Quat. Sci. Rev. 23, 959–965 (2004).Article
Google Scholar
Lightfoot, E. et al. Carbon and nitrogen isotopic variability in foxtail millet (Setaria italica) with watering regime. Rapid Commun. Mass Spectrom. 34, e8615 (2020).Article
CAS
Google Scholar
An, C.-B. et al. Variability of the stable carbon isotope ratio in modern and archaeological millets: evidence from northern China. J. Archaeol. Sci. 53, 316–322 (2015).Article
Google Scholar
Dong, Y. et al. The potential of stable carbon and nitrogen isotope analysis of foxtail and broomcorn millets for investigating ancient farming systems. Front. plant Sci. 13, 1018312 (2022).Article
CAS
Google Scholar
Depaermentier, M. L. C. Isotopic dataset to: Depaermentier, MLC, Kempf, M, Motuzaitė Matuzevičiūtė, G. “Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe” (2025).Araus, J. L. et al. Identification of Ancient Irrigation Practices based on the Carbon Isotope Discrimination of Plant Seeds: a Case Study from the South-East Iberian Peninsula. J. Archaeol. Sci. 24, 729–740 (1997).Article
Google Scholar
Lightfoot, E. & Stevens, R. E. Stable isotope investigations of charred barley (Hordeum vulgare) and wheat (Triticum spelta) grains from Danebury Hillfort: implications for palaeodietary reconstructions. J. Archaeol. Sci. 39, 656–662 (2012).Article
CAS
Google Scholar
Gron, K. J. et al. Archaeological cereals as an isotope record of long-term soil health and anthropogenic amendment in southern Scandinavia. Quat. Sci. Rev. 253, 106762 (2021).Article
Google Scholar
Hald, M. M. et al. Farming during turbulent times: agriculture, food crops, and manuring practices in bronze age to viking age Denmark. J. Archaeol. Sci. Rep. 58, 104736 (2024).
Google Scholar
Nitsch, E. et al. A bottom-up view of food surplus: using stable carbon and nitrogen isotope analysis to investigate agricultural strategies and diet at Bronze Age Archontiko and Thessaloniki Toumba, northern Greece. World Archaeol. 49, 105–137 (2017).Article
Google Scholar
Heaton, T. H., Jones, G., Halstead, P. & Tsipropoulos, T. Variations in the 13C/12C ratios of modern wheat grain, and implications for interpreting data from Bronze Age Assiros Toumba, Greece. J. Archaeol. Sci. 36, 2224–2233 (2009).Article
Google Scholar
Aguilera, M., Zech-Matterne, V., Lepetz, S. & Balasse, M. Crop fertility conditions in north-eastern Gaul during the la tène and roman periods: a combined stable isotope analysis of archaeobotanical and archaeozoological remains. Environ. Archaeol. 23, 323–337 (2018).Article
Google Scholar
Alagich, R., Gardeisen, A., Alonso, N., Rovira, N. & Bogaard, A. Using stable isotopes and functional weed ecology to explore social differences in early urban contexts: the case of Lattara in mediterranean France. J. Archaeol. Sci. 93, 135–149 (2018).Article
Google Scholar
Antanaitis, I. & Ogrinc, N. Chemical analysis of bone: stable isotope evidence of the diet of Neolithic and Bronze Age people In Lithuania. Istorija XLV, 3–12 (2000).
Google Scholar
Araus, J. L. & Buxó, R. Changes in carbon isotope discrimination in grain cereals from the north-western mediterranean basin during the past seven millenia. Funct. Plant Biol. 20, 117 (1993).CAS
Google Scholar
Araus, J. L. et al. Changes in carbon isotope discrimination in grain cereals from different regions of the western Mediterranean Basin during the past seven millennia. Palaeoenvironmental evidence of a differential change in aridity during the late Holocene. Glob. Change Biol. 3, 107–118 (1997).Article
Google Scholar
Araus, J. L. et al. Isotope and morphometrical evidence reveals the technological package associated with agriculture adoption in western Europe. PNAS 121, e2401065121 (2024).Article
CAS
Google Scholar
Ben Makhad, S. et al. Crop manuring on the Beauce plateau (France) during the second iron age. J. Archaeol. Sci.: Rep. 43, 103463 (2022).
Google Scholar
Bernardini, S. et al. New multi-proxy isotopic data on the copper age of eastern Liguria. Riv. di Sci. Preistoriche LXXIII S3, 1037–1043 (2023).
Google Scholar
Bogaard, A. et al. From traditional farming in morocco to early urban agroecology in northern mesopotamia: combining present-day arable weed surveys and crop isotope analysis to reconstruct past agrosystems in (semi-)arid regions. Environ. Archaeol. 23, 303–322 (2018).Article
Google Scholar
Bogaard, A. et al. Crop manuring and intensive land management by Europe’s first farmers. Proc. Natl. Acad. Sci. USA 110, 12589–12594 (2013).Article
CAS
Google Scholar
Cortese, F. et al. Isotopic reconstruction of the subsistence strategy for a Central Italian Bronze Age community (Pastena cave, 2 nd millennium BCE) (2022).DiBenedetto, K. E. Investigating Land Use by the Inhabitants of Western Cyprus During the Early Neolithic (2018).Eklund, M. Changing Agriculture. Stable isotope analysis of charred cereals from Iron Age Öland. (Master thesis, Stockholm University, Stockholm, 2019).Fernández-Crespo, T., Ordoño, J., Bogaard, A., Llanos, A. & Schulting, R. A snapshot of subsistence in Iron Age Iberia: the case of La Hoya village. J. Archaeol. Sci.: Rep. 28, 102037 (2019).
Google Scholar
Fiorentino, G. et al. Third millennium B.C. climate change in Syria highlighted by Carbon stable isotope analysis of 14C-AMS dated plant remains from Ebla. Palaeogeogr. Palaeoclimatol. Palaeoecol. 266, 51–58 (2008).Article
Google Scholar
Fiorentino, G., Caracuta, V., Casiello, G., Longobardi, F. & Sacco, A. Studying ancient crop provenance: implications from δ(13)C and δ(15)N values of charred barley in a Middle Bronze Age silo at Ebla(NW Syria). Rapid Commun. Mass Spectrom. 26, 327–335 (2012).Article
CAS
Google Scholar
García-Collado, M. I. et al. First direct evidence of agrarian practices in the alava plateau (northern Iiberia) during the middle ages through carbon and nitrogen stable isotope analyses of charred seeds. Environ. Archaeol. 1–11 (2022).Gavériaux, F. et al. L’alimentation des premières sociétés agropastorales du Sud de la France: premières données isotopiques sur des graines et fruits carbonisés néolithiques et essais de modélisation. Comptes Rendus. Palevol. (2021).Gavériaux, F., Motta, L., Bailey, P., Brilli, M. & Sadori, L. Crop husbandry at gabii during the iron age and archaic period: the archaeobotanical and stable isotope evidence. Environ. Archaeol. 29, 370–383 (2024).Article
Google Scholar
Gillis, R. E. et al. Stable isotopic insights into crop cultivation, animal husbandry, and land use at the Linearbandkeramik site of Vráble-Veľké Lehemby (Slovakia). Archaeol. Anthropol. Sci. 12; https://doi.org/10.1007/s12520-020-01210-2 (2020).Gron, K. J. et al. Nitrogen isotope evidence for manuring of early Neolithic Funnel Beaker Culture cereals from Stensborg, Sweden. J. Archaeol. Sci.: Rep. 14, 575–579 (2017).
Google Scholar
Halvorsen, L. S., Mørkved, P. T. & Hjelle, K. L. Were prehistoric cereal fields in western Norway manured? Evidence from stable isotope values (δ15N) of charred modern and fossil cereals. Veget. Hist. Archaeobot. 32, 583–596 (2023).Article
Google Scholar
Fraser, R. A., Bogaard, A., Schäfer, M., Arbogast, R. & Heaton, T. H. E. Integrating botanical, faunal and human stable carbon and nitrogen isotope values to reconstruct land use and palaeodiet at LBK Vaihingen an der Enz, Baden-Württemberg. World Archaeol. 45, 492–517 (2013).Article
Google Scholar
Isaakidou, V. et al. Changing land use and political economy at neolithic and bronze Age Knossos, Crete: stable carbon (δ 13 C) and nitrogen (δ 15 N) isotope analysis of charred crop grains and faunal bone collagen. Proc. Prehist. Soc. 88, 155–191 (2022).Article
Google Scholar
Kanstrup, M. When δ15N values reveal manuring practice. Empirical evidence from fieldwork, charring experiments and archaeobotanical remains (Aarhus Universitet, Institut for Agroøkologi, Aarhus, 2012).Karakaya, D. Botanical Aspects of the Environment and Economy at Tell Tayinat from the Bronze to Iron Ages (ca. 2.200–600 BCE), In south-central Turkey (Doctoral Dissertation, Universität Tübingen (Germany), Doctoral Dissertation, Universität Tübingen (Germany, 2020).Karaliūtė, R., Motuzaitė Matuzevičiūtė, G., Styring, A. & Stroud, E. 2600 Years of farming in Eastern Lithuania: soil management according to ancient barley isotopic values. Quaternary Environments and Humans (in preparation).Knipper, C. et al. What is on the menu in a Celtic town? Iron age diet reconstructed at Basel-Gasfabrik, Switzerland. Archaeol. Anthropol. Sci. 9, 1307–1326 (2017).Article
Google Scholar
Knipper, C. et al. Reconstructing Bronze Age diets and farming strategies at the early Bronze Age sites of La Bastida and Gatas (southeast Iberia) using stable isotope analysis. PloS One 15, e0229398 (2020).Article
CAS
Google Scholar
Lodwick, L. Cultivating villa economies: archaeobotanical and isotopic evidence for iron age to roman agricultural practices on the chalk downlands of Southern Britain. Eur. j. archaeol. 26, 445–466 (2023).Article
Google Scholar
Lodwick, L., Campbell, G., Crosby, V. & Müldner, G. Isotopic evidence for changes in cereal production strategies in iron age and Roman Britain. Environ. Archaeol. 26, 13–28 (2021).Article
Google Scholar
Maltas, T., Şahoğlu, V., Erkanal, H. & Tuncel, R. From horticulture to agriculture: New data on farming practices in Late Chalcolithic western Anatolia. J. Archaeol. Sci.: Rep. 43, 103482 (2022).
Google Scholar
Martínez Sánchez, R. M. et al. Archaeology, chronology, and age-diet insights of two late fourth millennium cal BC pit graves from central southern Iberia (Córdoba, Spain). Int. J. Osteoarchaeol. 30, 245–255 (2020).Article
Google Scholar
Messager, E. et al. Archaeobotanical and isotopic evidence of Early Bronze Age farming activities and diet in the mountainous environment of the South Caucasus: a pilot study of Chobareti site (Samtskhe–Javakheti region). J. Archaeol. Sci. 53, 214–226 (2015).Article
CAS
Google Scholar
Minkevičius, K. et al. New insights into the subsistence economy of the Late Bronze Age (1100–400 cal BC) communities in the southeastern Baltic. Archaeol. Balt. 30, 58–79 (2023).Article
Google Scholar
Mnich, B. et al. Terrestrial diet in prehistoric human groups from southern Poland based on human, faunal and botanical stable isotope evidence. J. Archaeol. Sci.: Rep. 32, 102382 (2020).
Google Scholar
Mora-González, A. et al. The isotopic footprint of irrigation in the western Mediterranean basin during the Bronze Age: the settlement of Terlinques, southeast Iberian Peninsula. Veget Hist. Archaeobot. 25, 459–468 (2016).Article
Google Scholar
Mora-González, A., Teira-Brión, A., Granados-Torres, A., Contreras-Cortés, F. & Delgado-Huertas, A. Agricultural production in the 1st millennium BCE in Northwest Iberia: results of carbon isotope analysis. Archaeol. Anthropol. Sci. 11, 2897–2909 (2019).Article
Google Scholar
Mueller-Bieniek, A. et al. Spatial and temporal patterns in Neolithic and Bronze Age agriculture in Poland based on the stable carbon and nitrogen isotopic composition of cereal grains. J. Archaeol. Sci.: Rep. 27, 101993 (2019).
Google Scholar
Niekamp, A. N. Crop growing conditions and agricultural practices in bronze age greece: a stable isotope analysis of archaeobotanical remains from Tsoungiza (Master’s thesis, University of Cincinnati, 2016).Nitsch, E. K., Jones, G., Sarpaki, A., Hald, M. M. & Bogaard, A. Farming practice and land management at Knossos, Crete: new insights from δ13C and δ15N analysis of Neolithic and Bronze Age crop remains. In Country in the City: Agricultural functions of protohistoric urban settlements (Aegean and Western Mediterranean), edited by D. Garcia, R. Orgeolet, M. Pomadère & J. Zurbach (Archaeopress, Oxford, 2019), pp. 159–173.O’connell, T. C. et al. Living and dying at the Portus Romae. Antiquity 93, 719–734 (2019).Article
Google Scholar
Pate, F. D., Henneberg, R. J. & Henneberg, M. Stable carbon and nitrogen isotope evidence for dietary variability at ancient Pompeii, Italy; https://doi.org/10.5281/zenodo.35526 (2015).Pilaar Birch, S. E. et al. Herd management and subsistence practices as inferred from isotopic analysis of animals and plants at Bronze Age Politiko-Troullia, Cyprus. PloS One 17, e0275757 (2022).Article
CAS
Google Scholar
Piličiauskas, G. et al. The earliest evidence for crop cultivation during the Early Bronze Age in the southeastern Baltic. J. Archaeol. Sci.: Rep. 36, 102881 (2021).
Google Scholar
Riehl, S., Bryson, R. & Pustovoytov, K. Changing growing conditions for crops during the Near Eastern Bronze Age (3000–1200 BC): the stable carbon isotope evidence. J. Archaeol. Sci. 35, 1011–1022 (2008).Article
Google Scholar
Speciale, C. et al. The case study of Case Bastione: first analyses of 3rd millennium cal BC paleoenvironmental and subsistence systems in central Sicily. J. Archaeol. Sci.: Rep. 31, 102332 (2020).
Google Scholar
Styring, A. K. et al. Urban form and scale shaped the agroecology of early ‘cities’ in northern Mesopotamia, the Aegean and Central Europe. J. Agrar. Change 22, 831–854 (2022).Article
Google Scholar
Vaiglova, P. et al. An integrated stable isotope study of plants and animals from Kouphovouno, southern Greece: a new look at Neolithic farming. J. Archaeol. Sci. 42, 201–215 (2014).Article
CAS
Google Scholar
Vaiglova, P. et al. Of cattle and feasts: Multi-isotope investigation of animal husbandry and communal feasting at Neolithic Makriyalos, northern Greece. PloS One 13, e0194474 (2018).Article
Google Scholar
Vaiglova, P. et al. Exploring diversity in neolithic agropastoral management in mainland greece using stable isotope analysis. Environ. Archaeol. 28, 62–85 (2021).Article
Google Scholar
Vanhanen, S. & Ilves, K. Flax use, weeds and manuring in Viking Age Åland: archaeobotanical and stable isotope analysis. Veget. Hist. Archaeobot. 34, 501–517 (2025).Article
Google Scholar
Varalli, A. et al. Bronze Age innovations and impact on human diet: a multi-isotopic and multi-proxy study of western Switzerland. PloS One 16, e0245726 (2021).Article
CAS
Google Scholar
Varalli, A. et al. Insights into the frontier zone of Upper Seine Valley (France) during the Bronze Age through subsistence strategies and dietary patterns. Archaeol. Anthropol. Sci. 15; https://doi.org/10.1007/s12520-023-01721-8 (2023).Wallace, M. P. et al. Stable carbon isotope evidence for neolithic and bronze age crop water management in the Eastern Mediterranean and Southwest Asia. PloS One 10, e0127085 (2015).Article
Google Scholar
Lempiäinen-Avci, M. et al. New insight into medieval cultivation at the village of Mankby in Espoo, Finland – comparing stable isotopes of carbon δ¹³C and nitrogen δ¹⁵N of Secale and Hordeum from Mankby to 14th century grain materials from Estonia. In Shattered and Scattered Pasts. Festschrift for Professor Georg Haggrén, edited by T. Heinonen, et al., pp. 68–85 (Waasa Graphics, Vaasa, 2025).Ferrio, J. P., Alonso, N., Voltas, J. & Araus, J. L. Grain weight changes over time in ancient cereal crops: Potential roles of climate and genetic improvement. J. Cereal Sci. 44, 323–332 (2006).Article
Google Scholar
Della Penna, V. Tradizione e modernità delle pratiche agricole nei Monti dauni: storia e archeologia dei sistemi agroalimentari subappenninici (Unpublished dissertation, Università degli Studi di Foggia, 2022).Dreslerová, D. et al. Maintaining soil productivity as the key factor in European prehistoric and Medieval farming. J. Archaeol. Sci.: Rep. 35, 102633 (2021).
Google Scholar
Drtikolová Kaupová, S. et al. Stav izotopových výzkumů stravy, rezidenční mobility a zemědělského hospodaření populace Velké Moravy (9.–10. století). Arch. Rozhl. 74, 203–240 (2022).Article
Google Scholar
Hamerow, H. et al. An integrated bioarchaeological approach to the medieval ‘agricultural revolution’: a case study from Stafford, England, c.ad 800–1200. Eur. j. archaeol. 23, 585–609 (2020).Article
Google Scholar
Herrscher, E. et al. Dietary practices, cultural and social identity in the Early Bronze Age southern Caucasus. Paléorient, 151–174; (2021).Látková, M., Skála, R. & Drtikolová Kaupová, S. Bioarchaeological characteristics of the wheat (triticum aestivum) consumed at different parts of the early medieval settlement agglomeration of Mikulčice-Kopčany (9th–10th Century AD, Czech Republic). Environ. Archaeol. 30, 267–279 (2025).Article
Google Scholar
Reed, K. & Wallace, M. To pretreat, or not to pretreat, that is the question. The value of pretreatment protocols in the stable carbon and nitrogen isotope analysis of archaeobotanical cereal grains from Croatia and Serbia. Sci. Technol. Archaeol. Res. 10; https://doi.org/10.1080/20548923.2024.2410092 (2024).Russell, N., Cook, G. T., Ascough, P., Barrett, J. H. & Dugmore, A. Species specific marine radiocarbon reservoir effect: a comparison of ΔR values between Patella vulgata (limpet) shell carbonate and Gadus morhua (Atlantic cod) bone collagen. J. Archaeol. Sci. 38, 1008–1015 (2011).Article
Google Scholar
Schlütz, F. et al. Stable isotope analyses (δ15N, δ34S, δ13C) locate early rye cultivation in northern Europe within diverse manuring practices. Philos. Trans. R. Soc. Lond. Ser. B, Biol. Sci. 380, 20240195 (2025).
Google Scholar
Treasure, E. R. The frontier of Islam: an archaeobotanical study of agriculture in the Iberian Peninsula (c.700 – 1500 CE) (Unpublished dissertation, Durham University, 2020).Vaiglova, P. Neolithic agricultural management in the Eastern Mediterranean: new insight from a multi-isotope approach (Doctoral dissertation, University of Oxford, 2016).Olson, D. M. et al. Terrestrial ecoregions of the world: a new map of life on earth. BioScience 51, 933 (2001).Article
Google Scholar
Larsson, M., Bergman, J. & Olsson, P. A. Soil, fertilizer and plant density: Exploring the influence of environmental factors to stable nitrogen and carbon isotope composition in cereal grain. J. Archaeol. Sci. 163, 105935 (2024).Article
CAS
Google Scholar
Filipović, D. et al. New AMS 14C dates track the arrival and spread of broomcorn millet cultivation and agricultural change in prehistoric Europe. Sci. Rep. 10, 13698 (2020).Article
Google Scholar
Vaiglova, P., Lazar, N. A., Stroud, E. A., Loftus, E. & Makarewicz, C. A. Best practices for selecting samples, analyzing data, and publishing results in isotope archaeology. Quat. Int.; https://doi.org/10.1016/j.quaint.2022.02.027 (2023).Hedges, R. E. M. Isotopes and red herrings: comments on Milner et al. and Lidén et al. Antiquity 78, 34–37 (2004).Article
Google Scholar
O’Regan, H. J., Lamb, A. L. & Wilkinson, D. M. The missing mushrooms: Searching for fungi in ancient human dietary analysis. J. Archaeol. Sci. 75, 139–143 (2016).Article
Google Scholar
Drucker, D. G., Bridault, A., Hobson, K. A., Szuma, E. & Bocherens, H. Can carbon-13 in large herbivores reflect the canopy effect in temperate and boreal ecosystems? Evidence from modern and ancient ungulates. Palaeogeogr. Palaeoclimatol. Palaeoecol. 266, 69–82 (2008).Article
Google Scholar
Bonafini, M., Pellegrini, M., Ditchfield, P. & Pollard, A. M. Investigation of the ‘canopy effect’ in the isotope ecology of temperate woodlands. J. Archaeol. Sci. 40, 3926–3935 (2013).Article
Google Scholar
Häberle, S. et al. Carbon and nitrogen isotopic ratios in archaeological and modern Swiss fish as possible markers for diachronic anthropogenic activity in freshwater ecosystems. J. Archaeol. Sci.: Rep. 10, 411–423 (2016).
Google Scholar
Guiry, E. Complexities of stable carbon and nitrogen isotope biogeochemistry in ancient freshwater ecosystems: implications for the study of past subsistence and environmental change. Front. Ecol. Evol. 7; https://doi.org/10.3389/fevo.2019.00313 (2019).Robson, H. K. et al. Carbon and nitrogen stable isotope values in freshwater, brackish and marine fish bone collagen from Mesolithic and Neolithic sites in central and northern Europe. Environ. Archaeol. 21, 105–118 (2016).Article
Google Scholar
Göhring, A., Hölzl, S., Mayr, C. & Strauss, H. Identification and quantification of the sea spray effect on isotopic systems in α-cellulose (δ13C, δ18O), total sulfur (δ34S), and 87Sr/86Sr of European beach grass (Ammophila arenaria, L.) in a greenhouse experiment. Sci. Total Environ. 856, 158840 (2023).Article
Google Scholar
Büntgen, U. Scrutinizing tree-ring parameters for Holocene climate reconstructions. WIREs Clim. Change 13; https://doi.org/10.1002/wcc.778 (2022).Montgomery, J. et al. Strategic and sporadic marine consumption at the onset of the Neolithic: increasing temporal resolution in the isotope evidence. Antiquity 87, 1060–1072 (2013).Article
Google Scholar
Göhring, A., Hölzl, S., Mayr, C. & Strauss, H. Multi-isotope fingerprints of recent environmental samples from the Baltic coast and their implications for bioarchaeological studies. Sci. Total Environ. 874, 162513 (2023).Article
Google Scholar
Hopkins, J. B. & Ferguson, J. M. Correction: estimating the diets of animals using stable isotopes and a comprehensive bayesian mixing model. PloS One 7; https://doi.org/10.1371/annotation/d222580b-4f36-4403-bb1f-cfd449a5ed74 (2012).Ferrio, J. P., Araus, J. L., Buxó, R., Voltas, J. & Bort, J. Water management practices and climate in ancient agriculture: inferences from the stable isotope composition of archaeobotanical remains. Veget. Hist. Archaeobot. 14, 510–517 (2005).Article
Google Scholar
Jones, P. J., O’Connell, T. C., Jones, M. K., Singh, R. N. & Petrie, C. A. Crop water status from plant stable carbon isotope values: a test case for monsoonal climates. Holocene 31, 993–1004 (2021).Article
Google Scholar
Szpak, P., Metcalfe, J. Z. & Macdonald, R. A. Best practices for calibrating and reporting stable isotope measurements in archaeology. J. Archaeol. Sci.: Rep. 13, 609–616 (2017).
Google Scholar
Salesse, K. et al. IsoArcH.eu: an open-access and collaborative isotope database for bioarchaeological samples from the Graeco-Roman world and its margins. J. Archaeol. Sci.: Rep. 19, 1050–1055 (2018).
Google Scholar
Farese, M. MAIA: mediterranean archive of isotopic dAta. Pandora https://doi.org/10.48493/55v1-xg54 (2023).Article
Google Scholar
Formichella, G., Soncin, S. & Cocozza, C. Isotòpia: a stable isotope database for classical antiquity. Pandora v.1 19.09.2023; https://doi.org/10.48493/m0m0-b436 (2023).Mantile, N., Fernandes, R., Lubritto, C. & Cocozza, C. IsoMedIta: a stable isotope database for Medieval Italy. Res. Data J. Humanit. Soc. Sci. 8, 1–13 (2023).Article
Google Scholar
Cocozza, C., Cirelli, E., Groß, M., Teegen, W.-R. & Fernandes, R. Presenting the compendium isotoporum medii aevi, a multi-isotope database for Medieval Europe. Sci. Data 9, 354 (2022).Article
Google Scholar
Graven, H., Keeling, R. F. & Rogelj, J. Changes to carbon isotopes in atmospheric CO2 over the industrial era and into the future. Glob. Biogeochem. Cycles 34, e2019GB006170 (2020).Article
CAS
Google Scholar
Nitsch, E. K., Charles, M. & Bogaard, A. Calculating a statistically robust δ 13 C and δ 15 N offset for charred cereal and pulse seeds. Sci. Technol. Archaeol. Res. 1, 1–8 (2015).
Google Scholar
Stroud, E., Charles, M., Bogaard, A. & Hamerow, H. Turning up the heat: Assessing the impact of charring regime on the morphology and stable isotopic values of cereal grains. J. Archaeol. Sci. 153, 105754 (2023).Article
Google Scholar
Teira-Brión, A., Stroud, E., Charles, M. & Bogaard, A. The effects of charring on morphology and stable carbon and nitrogen isotope values of common and foxtail millet grains. Front. Environ. Archaeol. 3; https://doi.org/10.3389/fearc.2024.1473593 (2024).Varalli, A., D’Agostini, F., Madella, M., Fiorentino, G. & Lancelotti, C. Charring effects on stable carbon and nitrogen isotope values on C4 plants: Inferences for archaeological investigations. J. Archaeol. Sci. 156, 105821 (2023).Article
CAS
Google Scholar
Styring, A. K. et al. Recommendations for stable isotope analysis of charred archaeological crop remains. Front. Environ. Archaeol. 3; https://doi.org/10.3389/fearc.2024.1470375 (2024).R. Core Team. R: A language and environment for statistical (R Foundation for Statistical Computing, Vienna, 2021).Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. J. Stat. Soft. 67; https://doi.org/10.18637/jss.v067.i01 (2015).Hijmans, R. J. _terra: Spatial Data Analysis_ (2024).Pebesma, E. & Bivand, R. Spatial Data Science (Chapman and Hall/CRC, New York, 2023).Pebesma, E. Simple Features for R: standardized support for spatial vector data. R. J. 10, 439 (2018).Article
Google Scholar
Warnes, G. et al. _gtools: Various R Programming Tools_ (2023).Wickham, H., François, R., Henry, L., Müller, K. & Vaughan, D. _dplyr: A Grammar of Data Manipulation_ (2023).Wickham, H. Ggplot2. Elegant graphics for data analysis (Springer Science+Business Media, LLC, New York, NY, 2016).Dunnington, D. _ggspatial: Spatial Data Framework for ggplot2_ (2023).Auguie, B. _gridExtra: Miscellaneous Functions for “Grid” Graphics (2017).Kempf, M. Environmental data to: Depaermentier, MLC; Kempf, M; Motuzaitė Matuzevičiūtė, G: environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe (2025).Karger, D. N. et al. Climatologies at high resolution for the earth’s land surface areas. Sci. Data 4, 170122 (2017).Article
Google Scholar
South, A., Michael, S. & Massicotte, P. rnaturalearthdata: world vector map data from natural earth used in ‘rnaturalearth’ (2017).Download referencesAcknowledgementsM.L.C.D. and G.M.M. were funded by the European Union with a Consolidator Grant awarded to Giedrė Motuzaitė Matuzevičiūtė (ERC-CoG, MILWAYS, 101087964). Views and opinions expressed are those of the authors only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. MK’s research is funded by the Swiss National Science Foundation (SNSF/SNF): Project EXOCHAINS − Exploring Holocene Climate Change and Human Innovations across Eurasia (SNSF grant number: TMPFP2_217358).Author informationAuthor notesThese authors contributed equally: Margaux L. C. Depaermentier, Michael Kempf.Authors and AffiliationsFaculty of History, Vilnius University, Vilnius, LithuaniaMargaux L. C. Depaermentier & Giedrė Motuzaitė MatuzevičiūtėDepartment of Environmental Sciences, University of Basel, Basel, SwitzerlandMichael KempfAuthorsMargaux L. C. DepaermentierView author publicationsSearch author on:PubMed Google ScholarMichael KempfView author publicationsSearch author on:PubMed Google ScholarGiedrė Motuzaitė MatuzevičiūtėView author publicationsSearch author on:PubMed Google ScholarContributionsConceptualization: M.L.C.D., M.K. and G.M.M. Isotope data collection and formal analyses: M.L.C.D. Environmental and cluster analyses: M.K. Writing: M.L.C.D and M.K. Editing: M.L.C.D., M.K., G.M.M. Visualisation: M.K. and M.L.C.D. Revision: M.L.C.D., M.K.Corresponding authorsCorrespondence to
Margaux L. C. Depaermentier or Michael Kempf.Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Communications Earth & Environment thanks Ashley McCall and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Nicola Colombo and Aliénor Lavergne. [A peer review file is available].
Additional informationPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Supplementary informationTransparent Peer Review fileSupplementary InformationDescription of Additional Supplementary FilesSupplementary data 1Supplementary data 2Reporting summaryRights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Reprints and permissionsAbout this articleCite this articleDepaermentier, M.L.C., Kempf, M. & Motuzaitė Matuzevičiūtė, G. Environmentally adjusted δ13C thresholds for accurate detection of C4 plant consumption in Europe.
Commun Earth Environ 6, 1021 (2025). https://doi.org/10.1038/s43247-025-03031-4Download citationReceived: 13 July 2025Accepted: 12 November 2025Published: 18 December 2025Version of record: 18 December 2025DOI: https://doi.org/10.1038/s43247-025-03031-4Share this articleAnyone you share the following link with will be able to read this content:Get shareable linkSorry, a shareable link is not currently available for this article.Copy shareable link to clipboard
Provided by the Springer Nature SharedIt content-sharing initiative More
