the biochemical purity of the organic matter extracted from fossil bones, and by checking ... that the links between diet and the chemical composition of bone and tooth are ... environments, which growing phase occurs during the cool season. ..... could be performed on ancient human specimens, as long as they are available ...
Anthrop. Anz.
Jg.55
2
101-120
Stuttgart, Juni 1997
Isotopic biogeochemistry as a marker of Neandertal diet H. Bocherens Laboratoire de Biogeochimie Isotopique, Universite Pierre et Marie Curie, Paris, France With 11 figures and 1 table in the text
Summary: Natural abundances in I3C and 15N of bone collagen are linked to those of the diet. This isotopic signal can thus be linked to the dietary parameters of a given individual, such as the plants at the beginning of his food web and his position in the trophic web. In order to use this approach to study the diet of ancient humans, it is crucial to be sure that the original isotopic abundances of fossil collagen are preserved. This is done by controlling the biochemical purity of the organic matter extracted from fossil bones, and by checking that the isotopic differences observed in modern environments between herbivorous and carnivorous species are indeed measured in the fossil samples. Upper Pleistocene sites with a good isotopic preservation of collagen have been recognized in temperate and arctic environments. The isotopic signatures measured in such sites highlight particularities of the "mammoth steppe" fauna, and improve our knowledge of the diet of Neandertals. Zusammenfassung: Das nattirliche Vorkommen von 13C und 15N im Knochenkollagen steht in enger Beziehung zu jener in der Nahrung des Konsumenten. Die Kohlenstoff- und Stickstoffisotopie ist daher zur Rekonstruktion individuellen Nahrungsverhaltens geeignet, z. B. in bezug auf die pflanzlichen Nahrungskomponenten und die generelle Position des Konsumenten im Nahrungsnetz. Urn diesen Forschungsansatz flir die ErschlieBung prahistorischer men schlicher Ernahrungsweise nutzen zu kbnnen, muB unbedingt sichergestellt sein, daB die native Isotopie im gealterten Knochenkollagen konserviert ist. Hierzu ist eine sorgfaltige Kontrolle beztiglich der biochemischen Reinheit des extrahierbaren Proteins notwendig, ebenso wie der Nachweis, daB die an rezenten Organismen beobachtbaren Unterschiede in der Isotopie zwischen herbivoren und carnivoren Spezies auch in fossilen Tierknochenfunden erhalten und meBbar sind. Mehrere Fundstellen des oberen Pleistozans erftillen diese Bedingungen ftir die Kollagenkonservierung, sowohl an gemaBigten als auch arktischen Standorten. Die Isotopensignaturen dieser Funde geben Auskunft tiber die speziellen Charakteristika der Fauna der "Mammutsteppe" , und sind gleichermaBen geeignet, das Wissen tiber die Ernahrungsstrategie der Neandertaler zu erweitern.
Paper presented at the 2nd Congress of the Gesellschaft flir Anthropologie, Berlin 3.-6.10.1996; Sitzung "Perspektiven in der Anthropologie: Molekulare Aspekte".
Introduction The determination of the diet of ancient humans is a crucial problem in palaeoanthropoloy. Traditional approaches deal mainly with bone remnants, plant fragments and artefacts in order to reconstruct the alimentation of ancient humans. However, archaeological record is usually biased against perishable items, such as
0003-5548/97/0055-0101 $ 5.00 © 1997 E. Schweizerbart'sche Verlagsbuchhandlung, D-70176 Stuttgart
102
H. Bocherens
plant material. Thus, bone is over-represented, and moreover, the remnants recovered during an excavation are not representative of the consumed species during the whole year, since most of the sites are seasonally occupied by huntergatherers (Tauber 1981). Thus, the dietary remnants are never preserved in proportions reflecting exactly their proportions in the diet of ancient humans (Ambrose 1993). The bones and teeth of an animal do record in their chemical composition some information about the material they are formed on (in other words, the dietary input) and about the conditions under which they are formed (i. e. the physiological and environmental conditions). Thus, the chemical composition of fossil bones and teeth might provide valuable information about ancient ways of life, provided that the links between diet and the chemical composition of bone and tooth are sufficiently known on modern material, and provided that diagenesis did not interfere with the biogenic record. In this paper, the focus will be on the isotopic composition of carbon and nitrogen in an ancient biomolecule, collagen, and about the information that it can provide on Neandertal's diet. Emphasis will be put on the conditions in which this approach can be used safely. A first part will briefly describe the behaviour of carbon and nitrogen stable isotopes in the collagen of modern vertebrates. A second part will present results obtained on Eurasian Pleistocene sites, setting the limits of applications, and providing examples of applications and perspectives.
Methodology Carbon and nitrogen stable isotopes Both carbon and nitrogen present two stable isotopes, with slightly different atomic weights (Table 1). The isotopes of a given chemical element present similar chemical characteristics, but the small differences in the weight of their respective cores lead to small differences in their behaviour during chemical reactions. Indeed, during a chemical reaction involving a mixture of molecules containing one or the other isotope of a given chemical element, the reaction will be biased in favor of one isotopic type relative to the other, leading to an isotopic fraction aTable 1. Natural variations of the amounts of carbon and nitrogen stable isotopes (Values from Letolle et al. 1991). Chemical element
Limits of observed isotopic abundances on Earth (%)
Atomic weight
Carbon
12 13 Standard: Marine carbonate PDB I3CII2C Nitrogen
14 15 Standard: atmospheric nitrogen 15N/14N
98.99 1.15 = 11 237.20 ± 2.9 ppm
99.651 0.378 = 3676.50 ± 8.1 ppm
98.86 1.01
99.622 0.349
Isotopic biogeochemistry as a marker of Neandertal diet
103
tion. Thus, the product of the chemical reaction will present a slightly different amount of a given isotope when compared to the starting product. Mass spectrometry allows the measurement of isotopic enrichments or depletions of very small amplitude, thanks to a comparison with internationally defined standards. The following isotopic notation "delta" (0) is used to express the measured isotopic amounts: OEX = (Rsample/R standard -1) . 1000 (%0), where X stands for C or N, E stands for 13 or 15, respectively, and R stands for the isotopic ratios l3C/12C and lsN/14N , respectively. A positive value of 0 corresponds to the case of a sample enriched in the heavy isotope relative to the standard, whereas in the case of a negative value of 0, the sample is depleted in the heavy isotope relative to the standard.
Carbon isotopes in food webs The ol3C values of the whole body of an animal are slightly higher (+ 0 to 1 %0) than those of its diet (DeNiro & Epstein 1978). In a given ecosystem, the 013C values of the animals are thus linked to those of the plants at the basis of the trophic webs to which they belong. An important variability is observed in plant oJ3C values, due to three factors: the ol3C values of the sources of inorganic carbon, the metabolic pathways used for the fixation of inorganic carbon and the environmental conditions. In the terrestrial environment, plant ol3C values clearly exhibit a bimodal distribution (Deines 1980). These two modes correspond to the two major photosynthetic pathways used by terrestrial plants, i. e. the "C3 " photosynthesis and the "C4 " photosynthesis. The average 013C value of Cr piants is -27.1 ± 2.0 %0 whereas the average Ol3C value of C4 -plants is -13.1 ± 1.2 %0 (O'Leary 1981). C 3 -plants are all the trees and all the plants in temperate and cold environments, which growing phase occurs during the cool season. Their Ol3C values range from -34 to - 20%0. C 4 -plants are herbaceous plants in tropical dry environments, where the growing season is the hot season (examples are maize, sugar cane and sorgho). Their 013C values are close to -13 %0. A few plants (mostly succulent plants such as cactuses) exhibit a third photosynthetic pathway, the socalled CAM plants (CAM means Crassulacean Acid Metabolism). The ol3C values of the CAM plants cover the whole range of values of C 3- and C4-plants, since they use alternatively both photosynthetic pathways depending on the environmental conditions (O'Leary 1981). The source of inorganic carbon for terrestrial plants is the atmospheric CO 2 , presenting an average 013C value of -7.8 %0, with variations in time and space. This value decreases since the middle of the nineteenth century due to the input of CO 2 from industrial release and deforestation, which 013C value is close to -25 %0. The ol3C value of atmospheric CO 2 thus decreased from around -6.5 %0 in 1850 to -6.7 %0 in 1956, and to -7.8 %0 in 1989 (Marino & McElroy 1991). Under a close canopy, for instance in the underbush of a forest the CO 2 generated by the respiration and the decay of organic matter leads to a local depletion of the atmospheric CO 2 013C values, and thus the underbush plants are significantly 13Cdepleted, with 013C values ranging from -28 to -35 %0 (Van der Merwe & Medina 1991). Marine plants use mostly dissolved bicarbonate with 013C values close to 0%0 and present 013C values around -20 %0.
104
H. Bocherens
Nitrogen isotopes in food webs Nitrogen is mostly present in the proteins of organisms. Contrarily to carbon, the 815N values of animals are significantly higher than those of their diet (DeNiro & Epstein 1981). Thus 815N values increase at each trophic level, in every ecosystem (Minagawa & Wada 1984, Schoeninger & DeNiro 1984). The increase of the 81 5N values is around 4 %0 between two successive trophic levels. Thus, the observed difference between the 815N values of collagen from herbivorous and carnivorous mammals in the same ecosystem range from 3 %0 (Schwarcz 1991) to 5.7 %0 (Ambrose & DeNiro 1986). The most frequently measured 815N values are 2 to 7 %0 for terrestrial herbivores, 7 to 12 %0 for terrestrial carnivores, and 12 to 20 %0 for marine vertebrates. The influence of local ecological conditions on the 8 J5N values precludes the determination of exact ranges of variations according to an animal trophic level on a global scale. Indeed, parameters such as aridity and soil acidity lead to important local variations of the 8 J5N values of animals and humans. The collagen 81 5N values of a given herbivorous species are significantly higher for individuals living in an arid area relative to the individuals inhabiting a wetter region, independantly to the 815N values of the local plants (Ambrose & DeNiro 1986, Heaton et al. 1986, Sealy et al. 1987). This effect is probably due to a water and/or dietary stress, leading to recycling of nitrogen within the herbivore body, and thus producing a 15N-enrichment, which is more pronounced with increasing stress intensity (Ambrose 1991). Such an effect can interfere with the determination of the trophic level, but it may also provide additional information concerning the environmental conditions. Moreover, local conditions can provoke significant variations in plant 81 5N values, and thus a shift of the corresponding trophic web. For instance, the plants of the Dourdan forest (south of Paris, France) exhibit very negative 815N values, probably because of the soil acidity, and these very low 815N values are reflected in the 8 J5N values of the forest roe deer which show collagen 815N values as low as -3 %0 (Rodiere et al. 1996).
Recognition of major food resources in Eurasia Due to the isotopic particularities described above, it is possible to characterize different food resources according to their carbon and nitrogen isotopic signatures. Reviews of major food ressources have been performed on a worldwide scale (DeNiro 1987, Ambrose 1993). In the prospect of the determination of Neandertal diet, a more limited area, restricted to temperate and arctic Eurasia, has been selected to present the different food sources distinguishable using their DC and 15N amounts (Fig. 1). In this figure, ecosystems including C 4 -plants have not been considered since these plants are restricted to tropical regions in a natural state. This figure is however schema tical since local parameters can lead to significant deviations from the presented framework .
Stable isotopes in collagen Within a given organism, the different biochemical fractions exhibit significant differences in their isotopic composition. For instance, carbohydrates show roughly
105
Isotopic biogeochemistry as a marker of Neandertal diet
25
20
'0'
15
~ Z ~ (,Q
10
5
0
-30
-35
-25
-20
-15
-5
-10
o
Fig. 1. Recognition of major food resources in Eurasia using carbon and nitrogen stable isotopic abundances.
CarrriVOreSI~----~-+_5~%~~--~----~----------~-----~ body collagen
C3-Plants
~
_ _IIIIIi_r--_ _ _--r-_ _ _21
-30
-20
-10
o
~13
U
C pDB %0
Fig. 2. Carbon isotopic abundances in collagen in C3-plants trophic webs.
CanrivoresII-.--------.--------.----~~~------
(+ suckling carnivores I)
Canrivores I
-~~-~ -~
---y------
(+suckling herbivores)
Herbivores
-.--------~===--.----r--~
Plants
o
5
10
15
~15N U AIR
%0
Fig. 3. Nitrogen isotopic abundances in collagen in trophic webs. 22
herbivores omnivores carnivores 0 walrus @ whales fish-eaters 0
¢
20
•
polar
18
-
