40 Journal of the North Atlantic Special Volume 3 40 Introduction Stable carbon and nitrogen isotopic ratios (measured as δ13C and δ15N values) of remnant collagen in ancient human bone can provide direct information on the relative amounts of marine and terrestrial protein consumed by the human, and on the trophic position in the food chain from which the protein was obtained. This method has been widely applied in archaeology, and the details of the underlying measurement and interpretive processes have been extensively reviewed (cf. Ambrose 1993, Kelly 2000, Schoeninger and Moore 1992) and need no repetition here. We simply remind the reader that the protein of the terrestrial and marine biospheres has characteristic carbon and nitrogen isotopic signatures that can be traced through a food chain. The carbon isotopic signature changes little with increase in trophic level, while the nitrogen signature depends strongly on trophic level. The δ13C value of human collagen thus provides information on the relative amounts of protein obtained from the terrestrial and marine reservoirs, and the δ15N value provides confirmation and additional information on the trophic level of the food consumed. For this method to be reliably applied to reconstruct human diet, it is necessary to test the underlying assumptions by conducting a detailed analysis of the food from which the people under study obtained their sustenance. As will be seen in what follows, significant regional differences do exist, and so one cannot simply apply results from studies in other regions. For example, the animal data obtained by Coltrain et al. (2004) for the eastern Canadian Arctic are significantly different from those obtained here for Greenland. In this paper, the results from the first extensive carbon and nitrogen isotopic study of the most important Greenlandic wild prey species are presented and evaluated with respect to the extent that we can use the isotopic method to obtain independent information on ancient Greenlandic diet. We make no attempt here to compare and integrate this data into the Arctic literature as a whole. The Greenlandic Food Chains of Interest At first glance, isotopic measurements on bone protein should be particularly well suited to marine/ terrestrial dietary studies in Greenland, as many of the possible complicating factors are absent. A comprehensive overview of Greenlandic ecology, including human adaptation, is given in Born and Böcher’s (2001) “The Ecology of Greenland”. The summary descriptions to follow rely heavily on that text, on McGovern (1985), and on the faunal lists for excavated Neo-Eskimo sites as described in Gulløv 2012 [this volume]. The climate of southern Greenland is classed as Low Arctic, with a small Sub-Arctic area at the extreme southwestern tip; this latter portion includes the Norse Eastern Settlement. The remainder of the island is High Arctic. Neither C4 nor CAM plants are likely to be present, as these are adaptations to hot and dry climates. Plant husbandry need not be considered, as the Greenlandic climate was simply too cold; it was not a part of Neo-Eskimo culture and it was not possible for the Greenland Norse, except perhaps on a very limited basis. There were no wild plants that can have contributed directly in any significant amount to either human energy or protein needs (Arneborg et al. 2012 [this volume]). The food available to both Norse and Neo- Eskimo alike was primarily protein and fat from the terrestrial and marine ecosystems. In such circumstances, human intake of protein will greatly exceed minimum daily requirements, as the protein will also be used as a major source of energy. Human bone Stable Carbon and Nitrogen Isotopic Measurements of the Wild Animals Hunted by the Norse and the Neo-Eskimo People of Greenland D. Erle Nelson1, Jeppe Møhl2, Jan Heinemeier3, and Jette Arneborg4,5,* Abstract: Isotopic measurements of the terrestrial and marine wild animal species of greatest importance to Greenlandic Norse and Neo-Eskimo people were obtained to provide a solid basis for undertaking isotopic dietary analyses of these two human groups. The samples studied were animal bones from archaeological excavations of Norse and Neo-Eskimo middens. As expected, the values for the terrestrial and marine species were found to have characteristic isotopic composition, but there is sufficient variation within each group to require detailed consideration in interpreting isotopic information on the humans. Special Volume 3:40–50 Greenland Isotope Project: Diet in Norse Greenland AD 1000–AD 1450 Journal of the North Atlantic 1Simon Fraser University, Department of Archaeology. Burnaby, BC, Canada. 2Natural History Museum, Zoological Museum, University of Copenhagen, Copenhagen, Denmark. 3AMS 14C Dating Centre, Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark. 4National Museum of Denmark. Middle Ages and Renaissance, Research and Exhibitions, Frederiksholms Kanal 12, DK-1220 Copenhagen. Denmark. 5Institute of Geography, School of GeoSciences, University of Edinburgh, Scotland UK. *Corresponding author - Jette.arneborg@natmus.dk. 2012 2012 D.E. Nelson, J. Møhl , J. Heinemeier and J. Arneborg 41 collagen will thus directly reflect the carbon and nitrogen isotopic values of the protein consumed. The terrestrial biosphere in Greenland is also relatively simple when compared to those of temperate regions at lower latitudes. There are only two native mammals in the Greenlandic terrestrial food chain which could form a significant portion of a long-term human diet: the musk ox (Ovibos moschatus) and the caribou (Rangifer tarandus). In the time period of interest here, both had limited ranges in Greenland. The musk oxen were to be found only in the far northern and northeastern areas of the island, and were not available to either the Norse or to the southern Neo-Eskimo (Vibe 1981). From a technical viewpoint, these wild bovids form an interesting isotopic counterpart to their near relatives, the Norse cattle and sheep in the Southwest. Caribou were to be found throughout much of Greenland, but in western Greenland, they occurred in significant numbers only on the south-central coast. There they were extensively hunted by both Norse in the Western Settlement and by the Neo- Eskimo, as is seen in the frequent occurrence of caribou bone in Norse and Neo-Eskimo sites. To the south, there were only small, scattered herds, and caribou bone is infrequent there in the middens of either group (cf. McGovern 1985, Meldgaard 1986). The Arctic hare (Lepus arcticus) played a very minor role in the diet of both Neo-Eskimo and Norse. The only other terrestrial species that could have contributed to diet in any significant sense are the ptarmigan (Lagopus mutus) and one or two species of geese. All are present in midden collections, but none in significant amounts (McGovern 1985, Meldgaard 1986). Unlike the terrestrial reservoir, the marine dietary reservoir may not be simple, as the oceans and ice surrounding Greenland are complex physical systems (Buch 2001) that support complex ecosystems (Andersen 2001). For both Norse and Neo-Eskimo, it was the marine mammals, and in particular the seals, which were the wild animals of greatest importance. The primary prey species were different for these two groups. For the Neo-Eskimo, the ringed seal (Phoca hispida) was of fundamental importance in all locales. For the Norse in the Eastern Settlement, two migratory seal species, the harp (Phoca groenlandica) and the hooded (Cystophora cristata), were those targeted and consumed. As hooded seals tend not to migrate so far north as the Western Settlement, the seals taken by the Norse in that locale were the harp and the stationary harbor seal (Phoca vitulina). Both Norse and Neo-Eskimo also hunted the bearded seal (Erignathus barbatus) and the walrus (Odobenus rosmarus), but it is not clear whether these formed a substantial part of their diet or whether the animals were of greater importance for their hides and for the walrus ivory. Whales were also taken or scavenged, but it is not clear whether the Norse routinely hunted whales. There is not a great amount of whalebone in the Norse middens (McGovern 1985), but a single individual would have contributed a great amount of meat over a short time period. While the Neo-Eskimo hunted white whale (Delphinaptera leuca) and narwhal (Monodon monoceros) these species do not appear to have been a large part of their diet (Savelle 1994). The overall contribution of whale to Greenlandic diet remains uncertain, but it is clear that it was not nearly as fundamental to either Neo-Eskimo or Norse as were the seals. Marine birds, such as murres, guillemots, auks and eider ducks did provide important contributions to the Neo-Eskimo diet in some regions, but were not a significant part of the Norse diet. As strange as it may seem, fish bone is only rarely found in Neo- Eskimo sites and almost not at all in Norse sites. A recent faunal analysis (Enghoff 2003) of the Norse site Farm beneath the Sand illustrates this. There, some 166 fish bone fragments provide direct evidence for Norse fishing. In comparison, there were nearly 24,000 fragments of mammalian bone in the same collection. The amount of protein provided by fish remains unclear. Although the isotopic signatures of terrestrial and marine food chains are in general similar the world over, it is most useful to isotopically characterize the local food chains before studies on the human consumers are undertaken. To what extent are the basic requirements met? Do the isotopic values for the animals in the two reservoirs form homogeneous distributions with well-defined and distinctive averages, such that isotopic end-points can be defined for their consumers? In principle, it would be best to measure samples of the dietary protein itself, e.g., the meat of the various animals which made up the diets of the humans in question. That ideal is impossible, as such material is not preserved in the archaeological record and as there could be isotopic differences between modern animal tissue and that consumed in the past. In this study as in most, we use surrogate measurements obtained from the bone collagen of the animals actually eaten, as these bones can be found in the middens of the ancient consumers. For practical reasons, not all possible food species have been included in this first isotopic study. With one exception, those chosen for measurement are the marine and terrestrial mammals that constituted the primary prey species, especially those taken by the Norse. The exception was a small sample of thick-billed murre (Uria lomvia) bones taken from a Neo-Eskimo site in which this species was so abundantly represented that it numerically overwhelmed all other species (Gulløv 1997), and 42 Journal of the North Atlantic Special Volume 3 so it must have been locally very important to the Neo-Eskimo. All measurements were taken on bone samples obtained from previous excavations of Norse sites and of Neo-Eskimo middens from about the same period and slightly later, e.g., within the time range A.D. ≈1000 to A.D. ≈1600. While there were extensive collections of such materials available to us, this process did sometimes place limits on sample choice. The Samples Table 1 gives the specific and common names of the Norse and Neo-Eskimo sites from which bone samples were taken from archival faunal collections at the Zoological Museum of the University of Copenhagen. For both technical and practical reasons, the numbers of samples measured for each species roughly reflects its importance to human diet. The heavily exploited animals needed to be well characterized and there were numerous such bones in the collections, while it was sometimes difficult to find appropriate samples of the peripheral species. Figure 1 locates the sites on a map. Table 2 gives a list of the species and numbers of bone samples (n) examined for each. As all came from archaeological excavations, it was not always possible to be certain that each sample represented a different individual. Where circumstances permitted, the same bone element was chosen, or samples were taken from different locales and excavation levels within a site. As will be seen, the isotopic analyses themselves provide further information, as samples that have very different isotopic values are unlikely to be from the same individual. The last column in Table 2 gives a best estimate of the number of individuals (N) measured. While there is a possibility of a double measurement in a few cases, the average results obtained for each species is likely to provide a very good first estimate for that population. Measurement Methods Each bone was sampled at the Zoological Museum by using a clean, slow-speed drill to bore 3–6- mm holes into the bone after the surface had been cleaned by either scraping or milling. The drillings were collected on fresh sheets of aluminum foil, transferred to a tared, heat-cleaned glass vial, and weighed. At the outset, typical sample sizes taken were 300–500 mg, of which some 50–80 mg were used for stable isotope determinations. Later, these amounts were reduced by a factor of four or five. All samples were taken to the Archaeometry Laboratory at Simon Fraser University for preparation and measurement. Various methods for extracting the remnant collagen from bone are in use. Here, we used the SFU method (Takahashi and Nelson 2012 [Appendix 1, this volume]), which is designed to select from the sample only those molecular fragments that are a large fraction of the original protein molecule. Aliquots of the material thus isolated were then measured using a Carlo-Erba CHN analyzer coupled to a Micromass Prism mass spectrometer. Determinations of the carbon and nitrogen concentrations of the extract were thus obtained in addition to the isotopic results. These concentration measures provide a direct indication of the extract purity, as collagen has C and N concentrations of about 44% and 15% by weight Table 1. The sites from which samples were obtained. Site location Danish FM site Lat- Long- Site name NM # number* itude itude Neo-Eskimo sites Illuminersuit 75V1-II-002 75°21' 58°37' Nugarsuk 72V1-IV-017 72°49' 55°43' Annertusup Nuua 68V2-IV-078 68°35' 51°52' Qajaa 69V2-II-002 69°10' 52°40' Sermermiut 69V2-II-013 69°12' 51°08' IIlorpaat 64V1-III-029 64°05' 52°00' Dødemandsbugten 74Ø2-II-005 74°07' 20°53' Rypefjeldet 76Ø-025 76°56' 20°23' Norse sites Niaquusat V48 64V2-III-507 64°14' 50°20' Nipaatsoq V54 64V2-III-502 64°07' 50°07' Kilaarsarfik/Sandnes V51 64V2-III-511 64°15' 50°10' Farm beneath the Sand 64V2-III-555 64°07' 50°07' Igaliku/Gardar Ø47 60V2-IV-621 60°59' 45°26' VatnahverfiØ167 60V2-IV-603 60°50' 45°21' Qorlortoq Ø34 61V3-III-525 61°12' 45°33' *FM site number refers to the register of ancient monuments of the Greenland National Museum and Archives. Table 2. The species sampled. Number of Samples individuals Common name Species measured (n) (N) Terrestrial Musk ox Ovibos moschatus 6 6 Caribou Rangifer tarandus 27 23 to 27 Hare Lepus arcticus 1 1 Marine Harp seal Phoca groenlandica 18 14 to 18 Hooded seal Cystophora cristata 12 11 or 12 Ringed seal Phoca hispida 23 20 to 23 Harbor seal Phoca vitulina 9 8 or 9 Bearded seal Erignathus barbatus 5 4 or 5 Walrus Odobenus rosmarus 9 9 Baleen whale Not determined 5 5 White/narwhale Delphinapterus leucas/ 1 1 Monodon monoceros Figure 1 (opposite page). Map of Norse and Neo-Eskimo sites used in this study. 2012 D.E. Nelson, J. Møhl , J. Heinemeier and J. Arneborg 43 44 Journal of the North Atlantic Special Volume 3 respectively, and a C/N ratio of about 2.9. Samples with values deviating significantly from these values are suspect. The isotopic measurements are given, as is common practice, as δ13C (‰ PDB) and δ15N (‰ AIR). Experience has shown that these methods give a measurement precision (reproducibility of international laboratory standards) of ≤0.1‰ for carbon and ≤0.2‰ for nitrogen. These uncertainties are much smaller than the differences of interest and so have little influence on interpretation. Results and Interpretations Terrestrial animals The isotopic composition of bone collagen of northern grazers and browsers are discussed in Nelson et al. (2012c [this volume]). The δ13C values typically fall within the range -22 to -21‰ if no C4 or CAM plants are present. Such animals will typically have δ15N values in the approximate range 2–6‰, unless they are water-stressed or their diet includes nitrogen fixers. Suckling young animals can be expected to have δ15N values about 3‰ higher than those of their mothers, a value which gradually falls to that of an adult after weaning (see Nelson et al. 2012c [this volume] and references therein). These are values that can be anticipated for Greenlandic animals. As is shown below, this is usually but not always the case. Musk oxen and hare. Table 3 gives the results obtained for the musk oxen from the northeast and the single southwest coast hare. The average δ13C value and standard deviation (< δ13C > = -20.2 ± 0.2‰) for the musk oxen is as expected for terrestrial C3 plant-consuming animals, and the small variability for the six measures indicates both that the measurement method is reliable and that the individual animals consumed a very similar diet. The value for the hare (-20.6‰) is very similar, even though it is from a very different part of the island. For the musk oxen, the nitrogen average (< δ15N > = 2.8 ± 0.4‰) is also well within the range expected for northern herbivores. The hare is slightly higher at 3.8‰. In all, these results are exactly as anticipated. From them, we can predict that the Norse domestic bovids will have values similar to those of the musk oxen unless there are regional or latitudinal effects which for some reason are not reflected in these data. Caribou. As seen in Table 4, the caribou are also clearly terrestrial in their isotopic values, yet a little different from the musk oxen. The six individuals from northeastern sites have average values < δ13C > = -19.3 ± 0.2‰ and < δ15N > = 1.5 ± 0.5‰. As with the northeastern musk oxen, the variability is small. For these two co-existing herbivores, both carbon and nitrogen averages are significantly different (Students t: P < 0.0001 and = 0.0005, respectively). The isotopic values are so clearly distinct that it is possible to distinguish between these species. The west coast caribou samples are from Norse sites in the Western Settlement, with only one from an Eastern Settlement site. As this individual has isotopic values well within the range of the others, all these southwest coast samples have been combined, yielding averages of < δ13C > = - 18.2 ± 0.4‰ and < δ15N > = 2.0 ± 0.7‰. Their carbon values are different from those of the northeastern musk oxen and caribou. As a group, the nitrogen values are also slightly lower than those of the musk oxen, although some individuals have values that are similar. These anomalous carbon isotopic values have been noted when Greenlandic caribou bone and antler has been radiocarbon dated (e.g., the Danish National Museum and Aarhus University radiocarbon date lists). The cause is usually attributed to caribou consumption of lichen (cf. Coltrain et al. 2004, see also Nelson and Møhl 2003). Marine animals Overly simplistic explanations of the marine/ terrestrial isotopic differences would indicate that North Atlantic animals which exclusively use marine protein will have bone collagen δ13C values of about -12 to -13‰ (e.g., Arneborg et al. 1999) and δ15N values ranging from about 10 to 18‰, depend- Table 3. Musk oxen and hare. Sample Danish FM site Extract δ13C δ15N ID NM number number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Musk oxen Northeast Greenland #357 74Ø2-II-005 Metatarsus 45.1 14.6 3.1 -20.1 3.2 #373 76Ø1-025 Humerus, dex. 44.2 15.2 2.9 -20.1 2.7 #374 76Ø1-025 43.9 15.1 2.9 -20.6 2.6 #375 76Ø1-025 44.1 15.3 2.9 -20.3 3.3 #376 76Ø1-025 44.8 14.9 3.0 -20.3 2.6 #377 76Ø1-025 44.0 15.2 2.9 -19.9 2.3 n = 6 Average -20.2 2.8 St Dev 0.2 0.4 Hare Western Settlement #31 V48 64V2-III-507 Tibia 44.7 16.1 2.8 -20.6 3.8 2012 D.E. Nelson, J. Møhl , J. Heinemeier and J. Arneborg 45 The harp seals from northeast Greenland are part of a different population, and this fact is clearly refl ected in both carbon and nitrogen isotopes (<δ13C > = - 15.7 ± 0.4‰; < δ15N > = 11.8 ± 21‰), which are both lower than those of their western counterparts. This difference in nitrogen isotopic composition is almost as large as that observed between adjacent trophic levels in a food chain. It is evident that these two harp seal populations occupy different ecological niches. In this data-set, there is an individual from each coast (#122 and #378) which would seem to fit better within the other population, and it is interesting to speculate whether interchanges of animals between these two groups sometimes take place. The reason for the observed differences in isotopic levels could be either differences in trophic level or in the base values of the food chain; this question cannot be resolved on the basis of the present data. Hooded seals. The hooded seal results (Table 6) represent a sample obtained entirely from Norse sites in the Eastern Settlement, as hooded seal bones only rarely occur in either Western Settlement sites or in Neo-Eskimo sites on either the west or east coasts. The west coast hooded seal population is ing on the animals' position in the marine food chain. However, the results reported below indicate that this generalization does not hold true for the marine mammals studied here. Harp seals. Table 5 gives the results for the harp seal samples taken from Eastern Settlement and Western Settlement sites and from northeastern Neo- Eskimo sites. None of the δ13C values are as positive as -13‰. As one would expect, the average values for the samples from the Eastern- and Western Settlement are similar, since the animals were taken from the same population of migrating animals. One of the Eastern Settlement samples (#122) is a little lighter than those from the Western Settlement. Two others (Ø34–75 and Ø34–77) have both carbon and nitrogen values identical within measurement accuracy, and may represent a single individual. More detailed statistical comparisons between these two groups are not warranted with the present data, nor required for the present purpose. Both groups of samples can be combined to give average values for southwest coast harp seals of < δ13C > = -14.3 ± 0.5‰ and < δ15N > = 14.5 ± 0.7‰. Table 4. Caribou. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Northeast Greenland #366 74Ø2-II-005 Radius, sin. 44.0 15.7 2.8 -19.2 1.3 #367 74Ø2-II-005 44.6 15.7 2.8 -19.1 1.2 #368 74Ø2-II-005 Humerus, dex. 44.6 15.5 2.9 -19.2 2.1 #369 74Ø2-II-005 43.8 15.4 2.8 -19.1 1.8 #370 74Ø2-II-005 43.8 15.6 2.8 -19.6 1.0 #371 74Ø2-II-005 44.1 14.8 3.0 -19.5 2.0 n = 6 Average -19.3 1.5 St Dev 0.2 0.5 Western Settlement #42 V54 64V2-III-502 Radius, sin. 44.5 16.1 2.8 -17.9 2.1 #34 V48 64V2-III-507 Costa 44.6 16.1 2.8 -18.4 2.7 #35 V48 64V2-III-507 Tibia 43.4 15.3 2.8 -17.9 1.9 #45 V48 64V2-III-507 Radius, dex. 43.4 15.3 2.8 -17.7 2.4 #47 V48 64V2-III-507 Humerus,sin. 41.6 14.2 2.9 -18.1 3.1 #48 V48 64V2-III-507 Humerus, dex. 43.9 15.7 2.8 -18.2 1.6 #49 Farm beneath the Sand 64V2-III-555 Femur, dex. 44.5 15.5 2.9 -18.5 2.6 #50 Farm beneath the Sand 64V2-III-555 Metatarsus, sin. 44.8 15.9 2.8 -18.5 1.8 #51 Farm beneath the Sand 64V2-III-555 Phalange 43.9 15.6 2.8 -17.9 3.0 #52 Farm beneath the Sand 64V2-III-555 Humerus, sin. 44.3 15.9 2.8 -18.7 1.3 #55 Farm beneath the Sand 64V2-III-555 Tibia, sin. 44.0 16.0 2.8 -18.3 2.5 #106 Farm beneath the Sand 64V2-III-555 Astragulus, dex. 45.4 16.7 2.7 -18.6 1.0 #107 Farm beneath the Sand 64V2-III-555 Humerus, sin. 45.6 16.2 2.8 -18.8 1.4 #398 V51 64V2-III-511 " " 44.5 16.7 2.7 -17.4 0.9 #401 V51 64V2-III-511 Scapula, sin. 45.3 17.0 2.7 -17.3 1.9 #406 V51 64V2-III-511 Ulna, sin. 45.0 17.1 2.6 -18.6 1.5 #411 V51 64V2-III-511 Tibia 44.8 16.8 2.7 -18.4 1.3 #412 V51 64V2-III-511 Pelvis, dex. 44.6 16.5 2.7 -18.3 2.8 #413 V51 64V2-III-511 Ulna, sin. 45.9 16.9 2.7 -18.6 1.1 #416 V51 64V2-III-511 Scapula, sin. 45.0 16.9 2.7 -18.1 2.7 Eastern Settlement Ø34-64 Ø34 61V3-III-525 Scapula 45.4 15.8 2.9 -18.0 2.5 n = 17–21 Average -18.2 2.0 St Dev 0.4 0.7 46 Journal of the North Atlantic Special Volume 3 harp and hooded seals. It is shelf-stationary, it tends to stay clear of the ice, and its food is primarily higher trophic-level fish (Vibe 1981). This different adaptation is clearly evident in both the carbon and nitrogen isotopic results (Table 7), with average values of < δ13C > = -12.6 ± 0.3‰ and < δ15N > =17.0 ± 0.9‰. Both are 2–3‰ more positive than those of the harp and hooded seals. Ringed seals. This coastal seal was of fundamental importance to the Neo-Eskimo, but few were taken by the Norse (Table 8). The four samples from the Norse sites have C and N isotopic values little different from those from the west coast Neo-Eskimo sites (Students t: P = 0.36 and 0.31), and so have been inmigratory, with a pattern similar to that of the west coast harp, but not extended so far to the north (Vibe 1981). The average values are < δ13C > = -13.6 ± 0.5‰ and < δ15N > = 15.8 ± 1.0‰. These results are similar to, but still significantly different from those of the harp seals from the same area. The nitrogen result in particular suggests food choice at a slightly higher trophic level than that of the harp seals. No measures of hooded seal from the eastern population were made, as the east coast Neo-Eskimo apparently made very little use of this animal, perhaps because the extensive sea-ice in the area limits access. Harbor seals. The harbor seal occupies a different locale and ecological niche than the migratory Table 6. Hooded seals. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Eastern Settlement #88 Ø167 60V2-IV-603 Cranium, bulla 43.4 15.4 2.8 -13.7 15.5 #89 Ø167 60V2-IV-603 " " 41.7 14.6 2.9 -14.2 13.9 #90 Ø167 60V2-IV-603 " " 43.5 14.8 2.9 -14.8 14.5 #121 Ø167 60V2-IV-603 Mandible, dex. 47.7 16.5 2.9 -13.1 16.6 #123 Ø167 60V2-IV-603 " , dex. 46.8 15.9 2.9 -13.5 16.3 #260 Ø47 60V2-IV-621 " , dex. 42.9 14.4 3.0 -13.3 15.3 #296 Ø47 60V2-IV-621 Mandible, sin. 43.0 15.3 2.8 -13.5 15.4 #297 Ø47 60V2-IV-621 " , dex. 42.8 15.3 2.8 -13.1 17.0 #298 Ø47 60V2-IV-621 " , dex. 43.7 15.8 2.8 -13.4 16.9 #299 Ø47 60V2-IV-621 " , dex. 43.9 14.6 3.0 -13.7 16.4 #300 Ø47 60V2-IV-621 " , dex. 44.3 14.6 3.0 -13.4 16.5 #301 Ø47 60V2-IV-621 " , dex. 44.2 14.9 3.0 -13.2 15.4 n = 11 or 12 Average -13.6 15.8 St Dev 0.5 1.0 Table 5. Harp seals. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Northeast Greenland #363 74Ø2-II-005 Humerus, dex. 44.0 15.4 2.9 -16.1 9.9 #364 74Ø2-II-005 " , dex. 44.3 15.4 2.9 -15.9 11.2 #365 74Ø2-II-005 " , dex. 44.3 15.6 2.8 -16.0 11.5 #372 74Ø2-II-005 " , sin. 44.3 14.8 3.0 -15.3 11.0 #378 76Ø-025 Pars petrosa, sin. 44.4 14.5 3.1 -15.3 15.5 n = 5 Average -15.7 11.8 St Dev 0.4 2.1 Western Settlement #68 64V2-III-555 Femur 43.4 15.7 2.8 -13.9 13.7 #69 64V2-III-555 Humerus, dex. 43.5 15.7 2.8 -14.2 14.5 #70 64V2-III-555 Femur, dex. 42.9 15.5 2.8 -14.3 13.9 #71 64V2-III-555 Ulna, dex. 43.1 15.6 2.8 -14.3 15.5 #263 64V2-III-555 Pars petrosa 41.8 12.8 3.3 -14.6 13.8 #72 V48 64V2-III-507 Mandible, sin. 43.0 15.5 2.8 -13.8 15.1 #74 V48 64V2-III-507 Ulna 41.1 14.6 2.8 -14.1 15.1 #75 V48 64V2-III-507 Cranium, bulla 42.8 14.0 3.1 -14.5 15.6 #76 V48 64V2-III-507 Mandible, dex. 41.2 14.5 2.8 -13.4 15.3 n = 6–9 Average -14.1 14.7 St Dev 0.4 0.8 Eastern Settlement #122 Ø167 60V2-IV-603 Mandible, sin. 47.2 15.3 3.1 -15.3 14.7 Ø34-10 Ø34 61V3-III-525 " , dex. 47.9 17.2 2.8 -14.0 14.3 Ø34-75 Ø34 61V3-III-525 " , sin. 43.4 15.4 2.8 -14.8 13.7 Ø34-77 Ø34 61V3-III-525 " , dex. 45.9 16.2 2.8 -14.8 13.6 n = 3 or 4 Average -14.7 14.1 St Dev 0.6 0.5 2012 D.E. Nelson, J. Møhl , J. Heinemeier and J. Arneborg 47 cluded in the same group, giving overall average values of < δ13C > = -14.1 ± 0.7‰ and < δ15N > = 16.6 ± 1.4‰ for the west coast ringed seals. It is noteworthy that two of the specimens with high nitrogen values (#423 and #425) are juveniles with δ15N-values that must reflect their mothers milk. The other individuals with nitrogen values far above the average may also have been juveniles. Young ringed seals are easier to hunt than the more experienced adults. As with the eastern harp and hooded seals, the eastern ringed seals have significantly more negative average isotopic values (<δ13C > = -15.2 ± 0.3‰ and < δ15N > = 14.3 ± 1.3‰), although there is some overlap in the values for individuals (Students t: P < 0.0001 and = 0.012). Bearded seals. As discussed, the solitary bearded seal was not a large part of the diet of either Neo- Eskimo or Norse (Table 9). While the average carbon value (< δ13C > = -12.6 ± 0.3‰) is identical to that for the harbor seals, the much lower nitrogen values (< δ15N > = 13.5 ± 1.9‰) reflect their different trophic positions in the food chain. The bearded seal tends to be a benthic feeder with many lower level organisms in its diet (Vibe 1981). Table 7. Harbor seals. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Western Settlement #73 V48 64V2-III-507 Cranium 42.5 13.7 3.1 -13.0 16.9 #140 V48 64V2-III-507 Mandible 43.7 15.2 2.9 -12.6 18.0 #141 V48 64V2-III-507 Pars petrosa 42.6 14.4 3.0 -12.2 17.8 #142 V48 64V2-III-507 " " 41.8 14.2 2.9 -13.0 17.4 #143 V48 64V2-III-507 " " 43.1 15.0 2.9 -12.6 17.4 #144 V48 64V2-III-507 Mandible 43.6 14.9 2.9 -12.2 17.0 #145 V48 64V2-III-507 " 43.0 14.9 2.9 -12.1 17.3 #261 Farm beneath the Sand 64V2-III-555 Mandible, sin. 43.1 14.1 3.1 -12.7 15.5 #262 Farm beneath the Sand 64V2-III-555 Parspetrosa, sin. 43.6 14.1 3.1 -12.8 15.8 n = 8 or 9 Average -12.6 17.0 St Dev 0.3 0.9 Table 8. Ringed seals. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Neo-Eskimo sites, Eastern Greenland #358 74Ø2-II-005 Humerus, dex. 44.3 14.9 3.0 -15.1 15.6 #359 74Ø2-II-005 " , dex. 44.0 14.8 3.0 -14.9 15.5 #360 74Ø2-II-005 " , dex. 43.9 14.8 3.0 -15.4 14.5 #361 74Ø2-II-005 " , dex. 43.1 15.6 2.8 -15.2 12.8 #362 74Ø2-II-005 " , dex. 44.0 15.8 2.8 -15.5 13.1 n = 5 Average -15.2 14.3 St Dev 0.3 1.3 Neo-Eskimo sites, Western Greenland #417 72V1-IV-017 Humerus, sin. 44.6 16.1 2.8 -13.8 15.7 #418 72V1-IV-017 " , dex. 44.6 15.7 2.8 -13.8 15.6 #419 72V1-IV-017 " , dex. 44.9 15.8 2.8 -14.1 15.5 #420 72V1-IV-017 " , sin. 45.0 15.7 2.9 -14.3 18.1 #421 72V1-IV-017 " , dex. 44.8 16.0 2.8 -13.2 16.6 #422 72V1-IV-017 " , dex. 45.0 15.9 2.8 -14.3 15.4 #423 64V1-III-029 Bulla, sin. 44.8 15.2 2.9 -15.3 19.7 #424 64V1-III-029 " , sin. 44.8 15.1 3.0 -14.9 14.6 #425 64V1-III-029 " , dex. 46.8 15.0 3.1 -14.8 18.0 #426 69V2-II-002 Humerus, sin. 47.4 14.4 3.3 -15.2 15.8 #427 69V2-II-013 " , dex. 45.5 16.3 2.8 -13.4 18.7 #428 68V2-IV-078 Humerus, sin. 44.9 15.8 2.8 -13.7 15.4 #429 75V1-II-002 " , sin. 44.9 16.0 2.8 -13.9 17.6 #430 75V1-II-002 " , sin. 45.1 15.5 2.9 -13.2 17.5 Western Settlement #77 V48 64V2-III-507 Bulla 43.6 13.7 3.2 -13.8 15.3 #264 Farm beneath the Sand 64V2-III-555 Pars petrosa, dex. 44.3 15.0 3.0 -14.6 16.1 Eastern Settlement #293 Ø47 60V2-IV-621 " " , dex. 43.3 13.5 3.2 -13.5 17.0 #294 Ø47 60V2-IV-621 " " , sin. 43.2 14.1 3.1 -13.2 16.2 n = 15–18 Average -14.1 16.6 St Dev 0.7 1.4 48 Journal of the North Atlantic Special Volume 3 niche of the walrus, which feeds almost exclusively on clams (Vibe 1981). Whales. Only a few whales were measured, as only a few samples were available from the midden collections. In most cases, it was not possible to identify the species, but an attempt was made to distinguish between the large baleen whales and the smaller carnivorous whales as is shown in Table 11. These distinctions are clearly reflected in the data obtained, in that the baleen whales have nitrogen values indicating a lower trophic level position, while the single white or narwhal specimen has carbon and nitrogen values which are both well within the range of values for the high-trophic-level harbor seals. Thick-billed murre. As birds were not of great importance to the Greenlandic Norse, only one species, the thick-billed murre was included in this Walrus. Seemingly, neither Norse nor Neo- Eskimo included much walrus in their diet, the importance of the ivory to the Norse being emphasized by the nature of the samples measured here. Seven of the nine samples in Table 10 are premaxillary portions of the skulls of animals, which had been killed in areas north of the settlements. The tusk-bearing snouts were then chopped off and transported back to the Eastern Settlement so that the precious ivory could be carefully removed later. The average carbon result (< δ13C > = -12.7 ± 0.2‰) for these animals is not significantly different from those for the harbour and bearded seals, while the nitrogen average (< δ15N > = 11.7 ± 0.9‰) is about one trophic level lower than that of the bearded seal and two lower than that of the harbor seal. This result clearly reflects the specialized benthic Table 9. Bearded seals. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Western Settlement #147 V54 64V2-III-502 Pelvis 43.8 14.8 3.0 -12.3 12.9 #148 V48 64V2-III-507 - 43.4 15.0 2.9 -12.6 16.6 Eastern Settlement #302 Ø47 60V2-IV-621 Humerus, dex. 42.9 15.1 2.8 -12.7 13.7 #303 Ø47 60V2-IV-621 Femur, sin. 42.9 15.9 2.7 -13.0 12.3 #304 Ø47 60V2-IV-621 Humerus, dex. 42.5 15.7 2.7 -12.5 11.9 n = 4 or 5 Average -12.6 13.5 St Dev 0.3 1.9 Table 10. Walrus. Danish Extract δ13C δ15N Sample ID NM number FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Eastern Settlement #310 Ø47 60V2-IV-621 Premaxillary 45.1 14.2 3.2 -12.9 10.5 #311 Ø47 60V2-IV-621 " 44.6 15.9 2.8 -12.7 10.9 #312 Ø47 60V2-IV-621 " 43.4 15.4 2.8 -12.5 11.5 #313 Ø47 60V2-IV-621 " 43.0 15.9 2.7 -12.8 12.4 #314 Ø47 60V2-IV-621 " 43.8 14.2 3.1 -12.5 12.3 #315 Ø47 60V2-IV-621 " 44.5 15.0 3.0 -12.6 10.9 #316 Ø47 60V2-IV-621 " 44.7 14.8 3.0 -13.1 12.2 Ø34-1 Ø34 61V3-III-525 Costa 43.4 16.0 2.7 -13.0 11.2 Western Settlement #79 V48 64V2-III-507 Tooth 44.0 15.4 2.9 -12.6 13.3 n = 9 Average -12.7 11.7 St Dev 0.2 0.9 Table 11. Whales. Sample Bone Extract δ13C δ15N ID Danish NM number FM site number element C % N % C/N (‰ v PDB) (‰ v AIR) Western Settlement Whale (baleen?) #78 V48 64V2-III-507 - 42.5 15.3 2.8 -13.7 11.0 B. mysticetus? #36 V48 64V2-III-507 Vertebra 42.1 13.8 3.1 -14.9 15.2 Whale (baleen?) #317 Farm beneath the Sand 64V2-III-555 " 44.0 15.2 2.9 -13.8 10.9 " " #318 Farm beneath the Sand 64V2-III-555 " 42.9 15.5 2.8 -14.4 13.1 #319 Farm beneath the Sand 64V2-III-555 " 45.0 15.7 2.9 -15.2 12.8 Eastern Settlement White/narwhale Ø34-56 Ø34 61V3-III-525 Costa 45.3 15.8 2.9 -12.6 16.1 2012 D.E. Nelson, J. Møhl , J. Heinemeier and J. Arneborg 49 terrestrial and marine animals are approximately as expected, it is immediately evident that neither group as a whole forms a narrow, random distribution about a mean value. In both reservoirs, there is a range of isotopic values to be considered. Of greater importance, there is patterning evident between different species within the same ecosystem and between the same species in different locales; e.g., the northeastern caribou and musk oxen are isotopically distinguishable at the level of the individual, and there are clear differences between east and west coast populations of caribou, harp seal, and ringed seal. Arctic marine food webs are known to be isotopically complex (cf. Hobson and Welch 1992), but it is beyond the scope of this study to attempt to measurement series (Table 12). These birds are completely marine-adapted (Boertmann 2001), as is evident from the nitrogen isotopic results (< δ15N > = 14.6 ± 1.3‰). (It is curious that two individuals seem to be at a lower trophic level than the others.) In contrast, the average δ13C (< δ13C > = -16.1 ± 0.6‰) is the lightest observed for any marine species measured here and only 2‰ different from that for the caribou from the same region. Conclusions This wealth of data is summarized in a plot of the average values and standard deviations for all species measured (Fig. 2). While the results for both Table 12. Thick-billed murres. Extract δ13C δ15N Sample ID FM site number Bone element C % N % C/N (‰ v PDB) (‰ v AIR) Neo-Eskimo site, W. Greenland #433 64V1-III-029 Humerus, sin. 46.1 15.7 2.9 -16.1 12.5 #434 64V1-III-029 " " 45.4 15.3 3.0 -15.0 15.3 #435 64V1-III-029 " " 43.6 15.7 2.8 -16.4 15.8 #436 64V1-III-029 " " 44.4 15.6 2.8 -16.2 15.0 #437 64V1-III-029 " " 44.1 14.7 3.0 -16.7 15.5 #438 64V1-III-029 " " 44.5 15.3 2.9 -16.0 13.3 n = 6 Average -16.1 14.6 St Dev 0.6 1.3 Figure 2. Summary plot of average values. The uncertainties are one standard deviation. Full species names are given in Table 2. 50 Journal of the North Atlantic Special Volume 3 compare these data with those of others in the hope of explaining this patterning. Our purpose was to explore the implications for human dietary studies in Greenland, and those are evident. If all humans under consideration randomly consumed food from all species in both marine and terrestrial reservoirs, then it would be possible to use these data as the basis for generic dietary end-points that could be applied to human groups in all Greenlandic locales. That is not the case, as we know from archaeological and palaeoecological research that past people in Greenland had differing access to, and differing preferences for, these animals. Depending on the specific foods utilized, the carbon isotopic difference between terrestrial and marine reservoirs can vary between about 7–8‰ to as little as 2‰. While the nitrogen data does provide most useful information on the terrestrial/marine ratio, it contains the additional variable of trophic level. As a consequence, it would appear that each circumstance must be evaluated separately, and appropriate dietary endpoints calculated for each locale and for each human group under consideration. These calculations must therefore incorporate information on the habits and the past distributions of the prey animals, as well as information provided by the archaeological faunal studies. Those species that were not present or not hunted can be removed from consideration, and those that were of great importance can be emphasized. Some examples illustrate this point. Neither the musk ox nor the ringed seal need be considered in studies of the diet of the Norse, and the caribou provided at most a very small part of the protein of the Eastern Settlement Norse. For the southwestern Neo-Eskimo, the ringed seal was the species of primary importance, and the only terrestrial mammal of dietary consequence was the caribou. It follows that isotopic analyses of human bone cannot provide a detailed estimate of diet in Greenland that is entirely independent of other knowledge. This conclusion has a further important consequence. Precise radiocarbon dating of human bone requires an accurate estimate of the relative amounts of marine and terrestrial carbon in the bone collagen so that the age obtained can be corrected for the oceanic reservoir effect. In Greenland, obtaining this information will require careful consideration of the isotopic circumstances in each area under study. Literature Cited Ambrose, S.H. 1993 Isotopic analysis of paleodiets: Methodological and interpretive considerations. Pp. 59–130, In M.K. Sandford (Ed.). Investigations of Ancient Human Tissue: Chemical Analyses in Archaeology. Gordon and Breach, PA, USA. Andersen, O.N. 2001. The Marine Ecosystem. Pp. 123– 135, In E.W. Born and J. Böcher (Eds.). The Ecology of Greenland. Atuakkiorfik, Nuuk, Greenland. Arneborg, J., J. Heinemeier, N. Lynnerup, H.L. Nielsen, N. Rud, and Á.E Sveinbjörndóttir 1999. Change of diet of the Greenland Vikings determined from stable carbon isotope analysis and 14C dating of their bones. Radiocarbon 41(5):157–168. Arneborg, J., N. Lynnerup, J. Heinemeier, J. Møhl, N. Rud, and Á.E. Sveinbjörnsdóttir. 2012a [this volume]. Norse Greenland dietary economy ca. AD 980–ca. AD 145: Introduction. Journal of the North Atlantic Special Volume 3: 1–39. Boertmann, D. 2001. Sea birds. Pp. 170–184, In E.W. Born and J. Böcher (Eds.). The Ecology of Greenland. Atuakkiorfik, Nuuk, Greenland. Born, E.W. and J. Böcher (Eds.). 2001. The Ecology of Greenland. Atuakkiorfik, Nuuk, Greenland. 429 pp. Buch, E. 2001. The ocean environment. Pp. 111–122, In E.W. Born and J. Böcher (Eds.). The Ecology of Greenland. Atuakkiorfik, Nuuk, Greenland. Coltrain, J.B., M.G. Hayes, and D.H. O’Rourke 2004. Sealing, whaling, and caribou: The skeletal isotope chemistry of Eastern Arctic foragers. Journal of Archaeological Science 31:39–57. Enghoff, I.B., 2003. Hunting, fishing, and animals husbandry at The Farm beneath the Sand, Western Greenland. Meddelelser om Grønland, Man and Society 28. 104 pp. Gulløv, H.C. 1997. From Middle Ages to colonial times: Archaeological and ethnohistorical studies of the Neo- Eskimo culture of South West Greenland 1300–1800 AD. Meddelelser om Grønland, Man and Society 23. 501 pp. Hobson, K.A., and H.E. Welch 1992. Determination of trophic relationships within a high-Arctic marine food web using δ13C and δ15N analysis. Marine Ecology Progress Series 84:9–18. Kelly, J.F. 2000. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology. Canadian Journal of Zoology 78:1–27. McGovern, T.H. 1985. Contribution to Paleoeconomy of Norse Greenland. Acta Archaeologica 54:73–122. Meldgaard, M. 1986. The Greenland caribou: Zoogeography, taxonomy, and population dynamics. Meddelelser om Grønland, Bioscience 20. 87 pp. Nelson, D.E., and J. Møhl 2003. Radiocarbon dating caribou antler and bone: Are they different? Arctic 56(3):262–265. Savelle, J.M. 1994. Prehistoric exploitation of white whales (Delphinapterus leucas) and narwhals (Monodon monoceros) in the eastern Canadian Arctic. Meddelelser om Grønland, Bioscience 39:101–117. Schoeninger, M.J., and K. Moore 1992. Bone stable isotope studies in archaeology. Journal of World Prehistory 6(2):247–296. Takahashi, C.M., and D.E. Nelson 2011. Appendix 1. SFU Archaeometry Laboratory Methods. Journal of the North Atlantic Special Volume 3:134–135. Vibe, C. 1981. Pattedyr (Mammalia). Pp. 363–459, In F. Salomensen (Ed.). Grønlands Fauna. Nordisk Forlag, Copenhagen, Denmark.