INTRODUCTION
In archaeology, radiocarbon (14С) dating is one of the most accurate and commonly used methods for both determining the age of individual organic samples and building reliable chronologies for archaeological cultures and historical events. Reservoir effects may become a factor introducing error into individual 14С dates and entire chronological reconstructions. Specialists often sample human or faunal bones for 14С dating because these are frequently of particular interest given the research questions; however, if a part of their diet comes from a reservoir with lower 14С than the atmospheric level (such as oceanic or in some cases inland fresh water), the sample may be affected by a reservoir offset (for a detailed description of the chemistry and causes of reservoir effects see Deevey et al. Reference Deevey, Gross, Hutchinson and Kraybill1954; Deevey and Stuiver Reference Deevey and Stuiver1964; Keaveney and Reimer Reference Keaveney and Reimer2012; Wood et al. Reference Wood, Higham, Buzilhova, Suvorov, Heinemeier and Olsen2013; Fernandes et al. Reference Fernandes, Rinne, Nadeau and Grootes2016 and others).
As has been pointed out previously (e.g. Svyatko et al. Reference Svyatko, Mertz and Reimer2015, 2016), research on the extent of the freshwater reservoir effect (FRE) is rather scarce compared to the studies of marine reservoir effects, although such studies are rapidly increasing in number. Published FRE (modern and archaeological) studies in Siberia and the Eurasian Steppe have focused on the north Peri-Caspian Sea region and the lower course of the Don River (Shishlina et al. Reference Shishlina, van der Plicht, Hedges, Zazovskaya, Sevastyanov and Chichagova2007, Reference Shishlina, Zazovskaya, van der Plicht, Hedges, Sevastyanov and Chichagova2009, 2010, Reference Shishlina, Zazovskaya, van der Plicht and Sevastyanov2012, Reference Shishlina, Sevastyanov, Zazovskaya and van der Plicht2014; Motuzaite-Matuzeviciute et al. Reference Motuzaite-Matuzeviciute, Lillie and Telizhenko2015; van der Plicht et al. Reference van der Plicht, Shishlina and Zazovskaya2016), the middle and lower Dnieper basin (Lillie et al. Reference Lillie, Budd, Potekhina and Hedges2009), the north Caucasus (Higham et al. Reference Higham, Warren, Belinskij, Härke and Wood2010), the Upper Lena River and Baikal Lake regions (Nomokonova et al. Reference Nomokonova, Losey, Goriunova and Weber2013; Schulting et al. Reference Schulting, Ramsey, Bazaliiskii, Goriunova and Weber2014, Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015), northeast Kazakhstan (Svyatko et al. Reference Svyatko, Mertz and Reimer2015), the Serteyka River in the Smolensk Oblast (Kulkova et al. Reference Kulkova, Mazurkevich, Dolbunova, Regert, Mazuy, Nesterov and Sinai2015), the site of Minino on the shores of Kubenskoye Lake in Vologda Oblast (Wood et al. Reference Wood, Higham, Buzilhova, Suvorov, Heinemeier and Olsen2013), and the Minusinsk basin of southern Siberia (Svyatko et al. 2016). Several studies have also assessed the extent of the reservoir offsets within the Caspian (Olsson Reference Olsson1980; Arslanov and Tertychnaya Reference Arslanov and Tertychnaya1983; Karpytchev Reference Karpytchev1993; Kuzmin et al. Reference Kuzmin, Nevesskaya, Krivonogov and Burr2007; Leroy et al. Reference Leroy, Marret, Gibert, Chalie, Reyss and Arpe2007) and Aral Seas (Kuzmin et al. Reference Kuzmin, Nevesskaya, Krivonogov and Burr2007), both technically being lakes because they not connected to the ocean.
The aim of this study is to further access the variation in modern FREs across various regions of the Eurasian Steppe, especially from water sources associated with archaeological sites, through 14С dating of modern fish samples, and combining the results with the existing data from previous research. As such, the study is particularly topical for archaeologists working in the inland areas of the Eurasian Steppe, providing an indication of the extent of possible 14С offsets. The results will allow us to re-evaluate the potential and limitations of the 14С dating of bone material, most notably human remains, as well as samples containing fish and/or shell remains (e.g. pottery made with inclusions of fish bones and shells, etc.) or food crusts on pottery.
Variability and Sources of FRE
The “aging” of a carbon compound starts when it ceases to exchange with its surroundings, e.g. with the death of a living organism. 14С has a half-life of 5730 yr, so after ca. 50,000 yr no detectable 14C remains. As many sedimentary rocks are composed of the skeletal fragments of marine organisms that died millions of years ago, they represent a major depository of 14C-free or “old” carbon. Therefore, a major source of “old” carbon in freshwater flowing over limestone is dissolved inorganic carbon from weathering of 14C-free carbonate minerals (e.g. Sveinbjornsdottir et al. Reference Sveinbjörnsdóttir, Heinemeier and Arnorsson1995), together with inputs of old soil humus (Keith and Anderson Reference Keith and Anderson1963) or decaying organic matter that is washed into the water from the catchment (Goh Reference Goh1991). A long residence time of water in a lake or aquifer (leading to a slow CO2 exchange between the atmosphere and the water) can also lead to an older 14C age of the water (Broecker and Walton Reference Broecker and Walton1959; Hakansson Reference Hakansson1976; Geyh et al. Reference Geyh, Schotterer and Grosjean1998; Culleton Reference Culleton2006). Other contributing factors include melting glaciers and the consequent release of “old” carbon dioxide (CO2) (Hall and Henderson Reference Hall and Henderson2001; Osipov and Khlystov Reference Osipov and Khlystov2010), underwater output of methane hydrates (Prokopenko and Williams Reference Prokopenko and Williams2004) or methane from chalk beds (Trimmer et al. Reference Trimmer, Hildrew, Jackson, Pretty and Grey2009), and geothermal activity (Ascough et al. Reference Ascough, Cook, Church, Dunbar, Einarsson, McGovern, Dugmore, Perdikaris, Hastie, Friðriksson and Gestsdóttir2010; Higham et al. Reference Higham, Warren, Belinskij, Härke and Wood2010).
The extent of the FRE in a region is thought to be closely related to the geological composition of the underlying bedrock or surficial deposits. Keaveney and Reimer (Reference Keaveney and Reimer2012) demonstrated a high correlation between the FRE and carbonate alkalinity in modern lakes in Britain and Ireland. One can expect a greater FRE from areas rich in carbonaceous formations such as limestone, although presumably the particular structure, depth, and layout of a deposition may also affect the extent of the carbonate exchange with the groundwater.
As a result of underwater photosynthesis, plants and algae become enriched with “old” carbon (with low 14C content relative to the atmosphere), which then passes up the food chain to aquatic fauna (shellfish, fish, and mammals), and further to terrestrial animals that consume aquatic foods (including humans). The 14C age of such samples thus appears older. It must be considered that the FRE value within the reservoir may vary depending on the type and age of the aquatic animal analyzed, subject to its specific habitat and diet (e.g. Fernandes et al. Reference Fernandes, Dreves, Nadeau and Grootes2013). For example, because of the greater carbon exchange between the atmosphere and water, littoral fish and shellfish may be susceptible to FRE to a lesser extent than deep-water animals. There are also data on shifts of the FRE values over time as a result of changes in the hydrological structure, including geothermal conditions (e.g. Ascough et al. Reference Ascough, Cook, Church, Dunbar, Einarsson, McGovern, Dugmore, Perdikaris, Hastie, Friðriksson and Gestsdóttir2010) and depth of the reservoir (Geyh et al. Reference Geyh, Schotterer and Grosjean1998), or climatic conditions, leading to the melting of glaciers and/or permafrost and thus releasing a large amount of “old” carbon into the local reservoir (see discussion in Schulting et al. Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015; also Hågvar and Ohlson Reference Hågvar and Ohlson2013; Hågvar et al. Reference Hågvar, Ohlson and Brittain2016).
Being highly variable in space and time, modern FREs may reach much larger 14C offsets than observed in most marine surface water, often reaching thousands of 14C years (Geyh et al. Reference Geyh, Schotterer and Grosjean1998; Hall and Henderson Reference Hall and Henderson2001; Ascough et al. Reference Ascough, Cook, Church, Dunbar, Einarsson, McGovern, Dugmore, Perdikaris, Hastie, Friðriksson and Gestsdóttir2010).
MATERIALS AND METHODS
To explore the extent of modern freshwater reservoir effects in the Eurasian Steppe, 12 samples of modern fish were analyzed for 14C age and carbon (δ13C) and nitrogen (δ15N) stable isotopes (eight of which have been published in Svyatko Reference Svyatko2016 and Svyatko et al. in press). Samples were sourced in various freshwater rivers and lakes of Siberia and Kazakhstan near archaeological sites in a region from 45.5°N, 61.4°E to 52.9°N, 111.9°E (Figure 1, Table 1). The geology of the region is very complex so that it is not possible to specify the source of old carbon for each water body. Fish species utilized included pike (Esox lucius), perch (Perca sp.), roach (Rutilus rutilus), burbot (Lota lota), crucian carp (Carassius sp.), and grayling (Thymallus thymallus).
Data on FROs for modern organisms and other material in the Eurasian Steppe from present and previous research. Numbers in parentheses correspond to those in Table 1. 1. Kharga Lake; 2. Kyzylkoi River; 3. Shat River (Sartyksu); 4. Nura River; 5. Syr-Darya River; 6. Karasuk Bay; 7. Yenisei River; 8. Edarma River; 9. Chuya River; 10. Katun River; 11. Lena River (Schulting et al., Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015); 12. Deed-Khulsun Lake (van der Plicht et al. Reference van der Plicht, Shishlina and Zazovskaya2016); 13. Volga River (van der Plicht et al. Reference van der Plicht, Shishlina and Zazovskaya2016), note that exact location is not available; 14. Tsimlyansk city (van der Plicht et al. Reference van der Plicht, Shishlina and Zazovskaya2016); 15. Serteya II site (Kulkova et al., Reference Kulkova, Mazurkevich, Dolbunova, Regert, Mazuy, Nesterov and Sinai2015); 16. Podkumok River (Higham et al. Reference Higham, Warren, Belinskij, Härke and Wood2010); 17. Caspian Sea, various locations (Olsson Reference Olsson1980; Arslanov and Tertychnaya Reference Arslanov and Tertychnaya1983; Kuzmin et al. Reference Kuzmin, Nevesskaya, Krivonogov and Burr2007); 18. Aral Sea, various locations (Kuzmin et al. Reference Kuzmin, Nevesskaya, Krivonogov and Burr2007).
Results of 14C dating and stable isotope analysis of modern samples (with known collection year) from the Eurasian Steppe.
1Svyatko (2016); 2Schulting et al. (Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015); 3Shishlina (Reference Shishlina2010); 4van der Plicht et al. (2016); 5Kulkova et al. (2015); 6Higham et al. (Reference Higham, Warren, Belinskij, Härke and Wood2010); 7Olsson (1980); 8Arslanov and Tertychnaya (1983; as cited in Karpytchev [Reference Karpytchev1993]); 9Kuzmin et al. (2007).
a Note two aliquots of sample UBA-28386 were analyzed once with the lipid removal process (UBA-28386-2) and once without.
b F14C atm for the year 2012 was used (Levin et al. Reference Levin, Kromer and Hammer2013).
c F14C and FRO were calculated from published BP ages using CALIBomb calibration program (http://calib.org/CALIBomb/).
d Values are taken from the source.
The following water bodies have been analyzed (with sample-collection locations in parentheses):
-
∙ Kharga Lake (Eravninskiy Reg., Buryatia, Russia; 52°52′5′′N, 111°50′47′′E) is one of the largest lakes of the Eravno-Harginskaya Lake System (the total area of the latter is ca. 380 km2). The system is fed by rainwater, snowmelt, and groundwater;
-
∙ Nura River (the sample was taken near Tegiszhol cemetery and Temirkash settlement, Kazakhstan; 50°05′45.66′′N, 72°45′53.14′′E) is the largest river of the Nura-Sarysuk basin at 978 km in length. It rises in the Kyzyktas Mountains and flows within the Kazakh Uplands into Tengiz Lake; the total area of the river basin is 58,100 km². Over 90% of the annual runoff of the river happens during the spring floods, while in the summer the upper streams of the river dry out;
-
∙ Kyzylkoi River (near Korzhar site, Kazakhstan; 49°19′06.04′′N, 73°30′55.62′′E) is a ca. 40-km-long tributary of the Sherubai-Nura River (tributary of the Nura River), fed mostly by snowmelt and groundwater;
-
∙ Shat River (also known as Sartyksu, Kazakhstan; 50°38′20.51′′N, 74°14′59.55′′E) is a tributary of the Bala-Shiderty River (flowing into Shiderty River); the total length of Shat and Bala-Shiderty is ca. 80 km. The river is fed mostly by snowmelt and groundwater;
-
∙ Syr-Darya River (near Juvara site, Kazakhstan; 45°30′58.0′′N, 61°28′04.4′′E) is 2212 km long, originating in the Tian Shan Mountains in Kyrgyzstan and eastern Uzbekistan and flows west and northwest through Uzbekistan and southern Kazakhstan to the remains of the Aral Sea;
-
∙ Yenisei River (near Tepsei Mountain, Minusinsk Basin, Russia; 53°56′48.0′′N, 91°37′14.2′′E), at 748 km long, this is the largest river system flowing to the Arctic Ocean; it originates in Mongolia and flows north to the Yenisei Gulf in the Kara Sea through a large part of central Siberia. The average depth is 14 m and the maximum depth is 24 m;
-
∙ Karasuk Bay (Minusinsk Basin, Russia; 54°40′31.4′′N, 90°49′47.0′′E) is 0.5 km wide and 8 km long; it was formed in the lower stream of the Kharasuk River, a tributary of the Yenisei River, in the late 1960s as a result of the construction of the Krasnoyarsk hydroelectric station;
-
∙ Edarma River (Russia; 58°45′46.6′′N, 102°34′47.9′′E) is a 153-km-long tributary of the Angara River;
-
∙ Katun River (Uimon Basin, Russia; 50°11′35.0′′N, 85°57′06.8′′E) is 688 km long and originates in the Katun glaciers on the southern slope of Belukha Mountain; its drainage basin covers 60,900 km2; and
-
∙ Chuya River (Kurai Basin, Russia; 50°12′47.3′′N, 87°53′22.7′′E) is a 320-km-long tributary of the Katun River; its drainage basin covers 11,200 km2. Left tributaries of the Chuya River originate from the glaciers of the North-Chuya Mountain Ridge.
Sample pretreatment and analysis were performed in the 14CHRONO Centre for Climate, the Environment and Chronology (Queen’s University, Belfast). Collagen extraction was based on the ultrafiltration method (Brown et al. Reference Brown, Nelson, Vogel and Southon1988; Bronk Ramsey et al. Reference Bronk Ramsey, Higham, Bowles and Hedges2004), which included the following steps:
-
1. Bone demineralization in 2% HCl, followed by MilliQ® ultrapure water wash;
-
2. Gelatinization in pH=2 HCl at 58°C for 16 hr;
-
3. Filtration, using ceramic filter holders, glass filter flasks, and 1.2-µm glass microfiber filters;
-
4. Ultrafiltration using Vivaspin® 15S ultrafilters with MWCO 30 kDa; 3000–3500 rpm for 30 min; and
-
5. Freeze-drying; the dried collagen was stored in a desiccator.
Bone collagen stable carbon and nitrogen isotopes were measured in duplicate on a Thermo Delta V isotope ratio mass spectrometer (IRMS) coupled to a Thermo Flash 1112 elemental analyzer (EA) peripheral. The measurement uncertainty (1σ) of δ13C and δ15N based on 6–10 replicates of 7 archaeological bone collagen samples was 0.22‰ and 0.15‰, respectively. The reference standards used were IA-R041 L-alanine, IAEA-N-2 ammonium sulphate, IA-R001 wheat flour, IAEA-CH-6 sucrose, and nicotinamide. Results are reported using the delta convention relative to international standards: VPDB for δ13C and AIR for δ15N (Hoefs Reference Hoefs2009). No lipid removal was applied to the majority of fish samples as the primary aim of their analysis was to assess the freshwater reservoir effect in the area and the retention of lipids in the sample does not affect its 14C age or δ15N ratios (e.g. Guiry et al. Reference Guiry, Szpak and Richards2016). Only one sample, UBA-28386-2, was subjected to the lipid removal process following Bligh and Dyer (1959) as well as being analyzed without lipid removal.
For the accelerator mass spectrometry (AMS) 14C measurements, prepared bone collagen samples were sealed under vacuum in quartz tubes with an excess of CuO and combusted at 850°C. The CO2 was converted to graphite on an iron catalyst using a zinc reduction method (Slota et al. Reference Slota, Jull, Linick and Toolin1987). The pressed graphite “target” was then measured on a 0.5 MV National Electrostatics Compact AMS. The sample 14C/12C ratio was background corrected and normalized to the HOXII standard (SRM 4990C; National Institute of Standards and Technology). The 14C/12C ratio corrected for isotopic fractionation using the AMS-measured δ13C, is equivalent to fraction modern (F14C; Reimer et al. Reference Reimer, Brown and Reimer2004). The 14C age and 1σ were calculated from F14C using the Libby half-life (5568 yr) following the conventions of Stuiver and Polach (Reference Stuiver and Polach1977). The 14C ages were calibrated using the Calib 7.0 program (Stuiver et al. Reference Stuiver, Reimer and Reimer2013) and the IntCal13 calibration curve (Reimer et al. Reference Reimer, Bard, Bayliss, Beck, Blackwell, Bronk Ramsey, Buck, Cheng, Edwards, Friedrich, Grootes, Guilderson, Haflidason, Hajdas, Hatté, Heaton, Hoffmann, Hogg, Hughen, Kaiser, Kromer, Manning, Niu, Reimer, Richards, Scott, Southon, Staff, Turney and van der Plicht2013) where appropriate.
Freshwater reservoir offsets (FRO) were calculated as a difference in the 14C ages between the fish and atmosphere. Atmospheric 14C age and 14C age for the modern fish samples (conventionally given as >modern) were calculated using the following equation: 14C age=–8033×ln(F14C). F14Catm for various years were taken as an average of the monthly 14Catm measurements from Levin et al. (Reference Levin, Kromer and Hammer2013). For samples collected in 2012 and later, average F14Catm for the year 2012 was used. 14C age uncertainties were calculated using the following formula: σ14C=–8033×ln(F14C+σF14C – (–8033×ln(F14C)) for each sample. FRO uncertainty was calculated using σFRO=
$$\sqrt {\sigma a^{2} {\plus}\sigma b^{2} } $$
, where σa and σb are 14C age uncertainties for fish samples and atmosphere. For the pre-1950 sample the atmospheric value was taken from IntCal13 (Reimer et al. Reference Reimer, Bard, Bayliss, Beck, Blackwell, Bronk Ramsey, Buck, Cheng, Edwards, Friedrich, Grootes, Guilderson, Haflidason, Hajdas, Hatté, Heaton, Hoffmann, Hogg, Hughen, Kaiser, Kromer, Manning, Niu, Reimer, Richards, Scott, Southon, Staff, Turney and van der Plicht2013).
RESULTS AND DISCUSSION
The results, along with those from previous research, are presented in Table 1 and Figures 1 and 2 and are available online at http://chrono.qub.ac.uk/FRE/. In most samples analyzed, the collagen content varied between 7.4 and 20.4%, and the atomic C:N (C:Nat.) ratio varied between 3.0 and 3.9 (although we cannot exclude the potential admixture of lipids in the collagen samples).
Stable isotope results of modern freshwater fish (n=15, including three samples from previous research) and plants (n=1) analyzed. Data on fish and plant matter from previous research were taken from Higham et al. (Reference Higham, Warren, Belinskij, Härke and Wood2010), Schulting et al. (Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015), and Shishlina (Reference Shishlina2010). As the Salmo trutta sample was taken from flesh, 4‰ was added to its δ13C value to make it more comparable to bone collagen measurements.
The resulting data, combined with the existing published results, allow us to make a number of observations:
Freshwater Reservoir Effects
-
1. FREs are present in the majority of the areas analyzed, apart from Kharga Lake (FRO=15±35 14C yr) and Kyzylkoi River (FRO=2±63 14C yr)—these two reservoirs need further analysis of modern aquatic fauna to confirm the absence of any offset. The apparent lack of the FRO in these waterbodies is possibly due to specifics of local geology (such as the absence of old 14C-free bedrock exposures). For other reservoirs, the particular sources of old carbon are not clear but might include meltwater from permafrost such as in the case of the Katun and Chuya Rivers.
-
2. The FROs detected in present study are highly variable, ranging between 15±35 and 2±63 14C yr (i.e., no offset) to 1097±40 14C yr in various locations. Even higher offsets have been observed in previous research for the Caspian lowlands (1477±52 and 1037±52 14C yr; Shishlina Reference Shishlina2010; van der Plicht et al. Reference van der Plicht, Shishlina and Zazovskaya2016), Upper Lena River basin (Lake Baikal region, 1981±30 14C yr; Schulting et al. Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015), and Podkumok River (Karachay-Cherkess Republic, Russia, 3819±39 14C yr; Higham et al. Reference Higham, Warren, Belinskij, Härke and Wood2010).
-
3. The FROs also vary within single reservoirs between different species and different sizes/ages of fish. As a particular example, in the Minusinsk basin of southern Siberia, FRO values vary both between different reservoirs (with the tributary Karasuk having a higher offset compared to that of the main Yenisei River) and between species and fish of different age within the reservoir depending on their diet (the two pikes from the Karasuk Bay having the highest offsets, and the larger pike being ca. 120 14C yr “older” than the smaller one).
Many of the analyzed Eurasian Steppe reservoirs clearly contain depleted carbon; however, its particular sources remain unclear and are beyond the scope of the current study.
Stable Isotope Results
As no lipid removal procedure was applied to the majority of samples, their δ13C values potentially reflect combined collagen and lipid carbon isotopic signal. Lipids are known to be significantly lower in δ13C—recent research has demonstrated up to 5‰ increase in δ13C of fish-bone samples subjected to lipid removal (Guiry et al. Reference Guiry, Szpak and Richards2016). It must also be acknowledged that for modern aquatic samples, the isotopic values can be affected by present day agricultural and industrial activities in the area.
-
1. A variety of both δ13C and δ15N values can be observed in modern freshwater fish from the Eurasian Steppe. The majority of samples (with the exception of four samples from Kharga Lake and Edarma River in western Siberia, Deed-Khulsun Lake in Caspian lowlands and Syr-Darya River in Kazakhstan) apparently reflect C3 ecology of local reservoirs (though again, it should be noted here that the potential presence of lipids may have biased the resulting δ13C towards lower values). The higher δ13C values of the four samples above could reflect modern agricultural and industrial practices in the areas, and, as only single samples have been analyzed for these reservoirs, further isotopic measurements of additional specimens is needed to verify this. However, notably, a number of modern fish from the western Siberia regions, including Lake Baikal, have also previously demonstrated elevated δ13C levels (e.g. Katzenberg and Weber Reference Katzenberg and Weber1999; Weber et al. Reference Weber, White, Bazaliiskii, Goriunova, Savel’ev and Katzenberg2011 and others). Nitrogen isotope ratios of the fish appear related primarily to the diet of different species, with predatory species such as pike, perch, salmon, and burbot (Figure 2, filled symbols) demonstrating the highest δ15N. Surprisingly, the highest δ15N can be seen in a crucian carp from the Nura River (Kazakhstan). Whether this reflects the diet of the fish, the river’s natural state, or is the effect of modern agricultural or industrial activities, cannot be said.
-
2. No general relationship has been observed between FROs and either δ13C or δ15N values of the samples (in both cases p>0.05). To explore any possible local trends in this regard, FROs in individual regions would need to be explored in greater detail involving larger number of modern aquatic samples.
SUMMARY AND CONCLUSION
This research represents the first broad-scale study of FRE across the Eurasian Steppe region, specifically focused on modern samples. The aim of this work was not only to demonstrate the widespread variability of FREs in the territory of Russia and Kazakhstan, but also to draw the attention of specialists working in the area to the necessity of taking into account this important factor in 14C dating human and some faunal remains from archaeological sites.
In calculating the average FRO for particular regions or archaeological cultures, several factors need to be considered. Firstly, as FROs are extremely diverse geographically, an average offset should be calculated for a specific area (or even a particular reservoir), rather than for entire cultures/populations as the distribution of a population can span multiple regions with various geological characteristics. Secondly, for terrestrial animals (including humans), consuming aquatic foods, the degree of the offset for each individual will directly depend on the proportion of aquatic food in the diet, and further stable isotope analysis is essential to assess the latter (e.g. see Schulting et al. Reference Schulting, Ramsey, Bazaliiskii, Goriunova and Weber2014). For example, the difference between the paired human-herbivore bone samples from the Mesolithic to the Early Bronze Age in the upper Lena River basin ranges from 255 to 1010 14C yr (Schulting et al. Reference Schulting, Bronk Ramsey, Bazaliiskii and Weber2015). For the aquatic samples themselves (fish and shellfish), the extent of the FRO, as mentioned above, is directly linked to the diet and habitat of the animal and can vary significantly between individuals. Thus, at the moment there is a need for further research to demonstrate the appropriateness of the average FRO correction for different populations.
Acknowledgments
The study was supported by the Leverhulme Trust grant RPG-2014-08. We would like to thank our colleagues for their advice and help with acquiring samples, namely N V Tsydenova (Institute for Mongolian, Buddhist and Tibetan Studies, SB RAS), V V Varfolomeev (Karaganda State University n.a. E A Buketov), D Voyakin (scientific-research organization “Archaeological Expertise,” Almaty), V M Novoseltseva (Irkutsk Palaeoecology and Archaeology Laboratory, Institute of Archaeology and Ethnography, SB RAS), A Polyakov (Institute for the History of the Material Culture, Russian Academy of Sciences, Saint-Petersburg), and V I Soenov (Research Centre for the History and Culture of the Turkic Peoples, Gorno-Altaisk State University).
