T. Douglas Price
Abstract
Human mobility at Uppåkra
A preliminary report on isotopic proveniencing
Isotopic proveniencing of human remains from Uppåkra has been initiated and a preliminary analysis is reported here. The principle is straightforward. Tooth enamel forms during early childhood from nutrients, ele-ments, and isotopes in food. Some of the isotopic ratios in enamel are geographically variable. Individuals who move during their lifetime will have enamel ratios different from their place of burial and can be identified as non-local. Three isotopic ratios have been measured in tooth enamel from Uppåkra: strontium, oxygen, and carbon. Strontium and oxygen carry provenience information; carbon isotopes provide information on child-hood diet. A total of 10 human samples have been isotopically measured from Uppåkra, along with 10 cattle and 6 pig teeth. The animal teeth have been used to determine the local signal for the site for comparison to the human remains. Strontium and oxygen isotope ratios suggest that four of the ten individuals in the human sample are non-local. The results of this preliminary study indicate that there is significant variation in isotope ratios at Uppåkra, that non-local individuals can be identified, and that further investigation is warranted.
T. Douglas Price, Section for Prehistoric Archaeology, Aarhus University, Denmark, dougprice@me.com
164 T. DOUGLAS PRICE
past and tracing first generations of migrants.
The principle is straightforward. The strontium isotope ratio of 87Sr/86Sr varies among different kinds of rocks. Because the 87Sr forms through a radiogenic process as a product of decay from rubidium-87 over time, older rocks with more rubidium have a higher 87Sr/86Sr ratio, while younger rocks with less rubidium are at the opposite end of the range with low ratios (e.g., Montgomery et al. 2006). Sediments reflect the ratio of their parent material. The amount of 87S in nature varies but is roughly 7% of total strontium and 86Sr is 10% (87Sr/86Sr ⋍ 0.7). This ratio normally varies from about 0.700 in young rocks with low Rb to >0.730 in high-Rb rocks that are billions of years old.
Strontium moves into humans from rocks and sediment through the food chain (Sillen
& Kavanagh 1982; Price 1989; Price et al.
2001). Most measurements of human enamel fall in the range of 0.703 to 0.723. This ratio in enamel then reflects the underlying geology of the area where one was born when the tooth enamel formed. If an individual moved to a new location in a different geologic context, or was buried in a new place, the enamel iso-topes will differ from those of the new location, allowing the designation of that individual as a non-local.
There are several published summaries of the method (e.g., Bentley 2006; Montgom-ery 2010; Slovak & Paytan 2011). Analytical methods are described in detail in a number of publications (e.g., Price et al. 1994; Sjögren et al. 2009; Slovak & Paytan 2011; Frei and Price 2012). Numerous examples of the application of strontium isotope ratios to archaeological questions have been published (e.g., Benson et al. 2003; Montgomery et al. 2003; Wright 2005; Price & Gestsdóttir 2006; Knudson et al.
2008; Hedman et al. 2009; Price et al. 2011).
Strontium Isotopic Baselines in Scania
An essential issue in strontium isotope analysis involves determination of the local strontium isotope signal for the area in which a burial is found. In fact, levels of strontium isotopes in human tissue may vary from the actual geo-logical background for a number of reasons (e.g., Sillen et al. 1998; Price et al. 2002; Mau-rer et al. 2012). Factors include differential weathering of minerals in rock, atmospheric dust, the deposition of aeolian, alluvial, or glacial sediments on top of bedrock geology.
Complex geological areas may have several different sources of 87Sr/86Sr contributing to human diets. Coastal populations are impact-ed by several phenomena. Marine foods, for example, have a constant strontium isotope ratio of 0.7092. The same ratio, 0.7092, may also be introduced by salt spray and rainfall in coastal areas. For these reasons, it is necessary to measure bioavailable levels of 87Sr/86Sr to ascertain local strontium isotope ratios (e.g., Price et al. 2002; Frei & Price 2012).
Bioavailable 87Sr/86Sr is the range of val-ues actually available in the food chain. The local bioavailable isotopic signal of the place of burial can be determined in several ways.
in human bone from the individuals whose teeth are analyzed, from the bones of other humans or archaeological fauna at the site, or from modern fauna, water, soil extracts or vegetation in the vicinity (Maurer et al. 2012).
Geology of Scania
The geology of the Swedish province of Scan-ia has three major components. ScanScan-ia forms part of the boundary between the two major classes of the bedrock of Europe (Fig. 1). To the north and east lies an ancient craton, the Fennoscandian Shield in Sweden, Norway, and
165
HUMAN MOBILITY AT UPPÅKRA
Finland. To the south and west in the rest of Europe there is a mobile belt of crustal blocks. The border zone between these two regions is marked by the Torn-quist Line, which extends 2000 km from the North Sea to the Black Sea.
In Scania, this ca. 50 km wide zone runs through the province on a NW-SE line from Helsingborg to Ystad and forms a typical horst and graben landscape formed primarily during the Mesozoic (Fig. 1) (Graversen 2009). To the north and east of the Tornquist Line lie the pro-terozoic (2,500 to 542 million years ago) granites and gneisses of the Fennoscandi-an Shield. These very old rocks generally have very high 87Sr/86Sr values typical of a cratonic landscape. To the south and west of the Tornquist Line are the phanerozoic (the last 542 million years of earth’s history) sediments, largely of Mesozoic age. This area contains the only Jurassic sedimentary deposits in Sweden.
The bedrock in this area also includes substantial marine deposits of Late Mesozoic Senonian and Early Cenozoic Danien age. Strontium isotope ratios in bedrock in this zone should largely fol-low the known curve for seawater 87Sr/86Sr over time with values ranging between ca. 0.707 and 0.7085 (Vezier 1989).
There is another important source of strontium isotopes on the landscape of Scania. This area was glaciated repeatedly during the Pleisto-cene. Both Denmark and southwestern Skåne are located in a zone that is characterized by soft sedimentary substratum and thick glacial deposits. The region was subjected to
mul-tiple glaciations by ice lobes coming from the north, northeast, east, southeast, and south.
As noted in Fig. 1 the boundary between the most recent tills from the Northeastern and the Baltic lobes can be identified across Scania (Lagerlund 1987). These glacial tills covering the surface of much of Scania should have
87Sr/86Sr values different from the underlying bedrock depending on the origins of the sed-imentary load in the glacial lobes.
Bioavailable Strontium Isotopes in Scania
In addition to geological sources, the sea influ-ences strontium isotope ratios in southern Swe-den. The value of 0.7092, a constant 87Sr/86Sr for seawater today, may also be introduced by sea spray and rainfall in coastal areas. Measure-ment of marine aerosols in southern Sweden showed transport of across the entire region, Fig. 1. Bedrock map of Scania, Sweden, with
the location of the site of Uppåkra. The dotted line marks the boundary between the Northe-astern till (north) and the Baltic Till (south).
166 T. DOUGLAS PRICE
a distance of some 300 km, with a decrease in concentration from west to east (Gustafsson
& Franzén 2000). Other factors such as differ-ential weathering of minerals in bedrock may cause variation in 87Sr/86Sr values. For these reasons, it is necessary to measure bioavaila-ble levels of 87Sr/86Sr to ascertain the values of local strontium isotope sources.
There are several sources of information on
87Sr/86Sr levels in Sweden. Fig. 2 shows 87Sr/86Sr measured on non-intrusive whole rock, mainly granites and gneisses, from various published geological studies, compiled by Karl-Göran Sjögren (Sjögren et al. 2009). Although the sample locations are biased toward southern and
western Sweden, several clear patterns emerge.
Very low values, below 0.706, are extremely rare and likely come from small, volcanic features in the landscape. Values between 0.706 and 0.713 are restricted to southwestern Scania, Öland, and the area around Karlskrona in the south-eastern corner of mainland Sweden. These are areas of bedrock primarily composed of marine sediments such as limestone and chalk. The remaining values are for the most part quite high, greater than 0.718, reflecting the very old geological terrain of the Fennoscandian Shield.
More detail is available from several studies of strontium isotope ratios from archaeological sites in Sweden. Sjögren et al. (2009) analyzed Fig. 2. 87Sr/86Sr in bedrock from various areas in southern and central
Swe-den (Sjögren et al. 2009).
167
HUMAN MOBILITY AT UPPÅKRA
87Sr/86Sr in human remains from megalithic burials in the Falbygden region of western Swe-den along with numerous bioavailable samples from the surrounding region. Frei et al. (2009) recorded strontium isotope ratios in sheep wool and soil leachates from several areas in Sweden and Denmark. Arcini and Price (in press) have measured 87Sr/86Sr in human teeth from the Viking period and various biological materials from the Swedish island of Gotland and south-ern Sweden in a study of place of origin for the inhabitants of Viking Gotland. In addition, Frei and Price (2012) have reported a large number of baseline 87Sr/86Sr values from Denmark with a range of local values generally between 0.7089
and 0.7108. Since the southwestern corner of Scania and the area around the site of Uppåkra are in a very similar geological context of glacial moraine and outwash deposits, these values are likely appropriate for this region as well. These data are summarized in Fig. 3. Again a general pattern of lower values in southwestern Scania and on Gotland observed.
We have also measured bioavailable samples from the site of Uppåkra itself, specifically cattle and pig tooth enamel. Information on the samp-les is provided in Table 1 and the data obtained can be found in Table 2. These data are graphed by species in ranked order in Fig. 5. There are slight differences between the two species. The Fig. 3. 87Sr/86Sr values for baseline samples from various studies in southern
and central Sweden. Red = biological; blue = human enamel.
168 T. DOUGLAS PRICE
mean value for 14 cattle is 0.7118 ± 0.0009 with a range from 0.7097 to 0.7134. The mean value for 6 pigs is 0.7115 ± 0.0002 with a range from 0.7113 to 0.7118. The very low variance for the pigs suggests that this range of values likely reflects the local value at Uppåkra. The variation among the cattle, particularly the one very low and the two very high values, suggests that some of these animals may be imported.
More analyses are currently underway to map the isotopic variation in the vicinity of Uppåkra.
Oxygen Isotopes in Apatite
Oxygen isotope ratios vary geographically in surface water and rainfall. The oxygen isotope ratio in the human skeleton reflects that of body water, and ultimately of drinking water (Luz et al. 1984; Luz and Kolodny 1985; Kohn 1996), which in turn predominantly reflects local rainfall. Isotopes in rainfall are greatly affected by enrichment or depletion of the heavy 18O isotope relative to 16O in water due to evaporation and precipitation (e.g. Dans-gaard 1964). Major geographic factors affect-ing rainfall values then are latitude, elevation, amount of precipitation, and distance from the source (e.g., an ocean). Rainwater, H2O, can contain either isotope, and H218O has a greater mass than H216O, and requires more energy to evaporate and to stay in the atmos-phere. As this moisture moves over land, the first precipitation contains more of the heavy isotope and as the clouds move inland (and to higher elevations) the rain becomes even more depleted in the heavier isotope. Thus oxygen isotope ratios have some potential to vary geographically and provide information on past human movement.
Oxygen isotopes in ancient human skele-tal remains are found in both tooth enamel and bone. Oxygen is incorporated into dental
enamel during the early life of an individual and it remains unchanged through adulthood.
Thus, oxygen isotopes have the potential to be used to investigate human mobility and provenience (Bowen & Revenaugh 2003).
Oxygen has three isotopes, 16O (99.762% in nature), 17O (0.038%), and 18O (0.2%), all of which are stable and non-radiogenic. Oxygen isotopes are conventionally reported as the per mil difference in the ratio of 18O to 16O between a sample and a standard. This value is designated as δ18O. This value can be meas-ured in either carbonate (CO3)-2 or phosphate (PO4)-3 in apatite in tooth and bone. Less sam-ple is needed for carbonate, preparation is less demanding, and results between laboratories are more comparable (e.g., Bryant et al. 1996;
Sponheimer and Lee-Thorp 1999; Chenery et al. 2012). The standard used is commonly VSMOW (Vienna Standard Mean Ocean Water) for phosphate, or PDB (PeeDeeBee dolomite) for carbonate oxygen.
Oxygen isotope ratios in modern precipita-tion have been mapped for Sweden (Burgman et al. 1987) and are shown in Fig. 3. δ18OSMOW values for precipitation range from -14 in the extreme north of Sweden to between -8‰ and -10‰ in the province of Scania. These δ18O
-SMOW values in Scania correspond to a range in δ18OPDB in enamel between approximately -5‰ and -7‰ (Chenery et al. 2012).
Carbon Isotopes in Apatite
Carbon isotope ratios between 13C and 12C are reported relative to a reference material and reported conventionally as δ13C in per mil (‰
or parts per thousand). Carbon isotope ratios are measured in human remains as either organic collagen or the mineral apatite. Most of the pal-eodiet work using carbon isotopes has focused on the organic collagen component of the
skel-169
HUMAN MOBILITY AT UPPÅKRA
eton. Carbon also is present in the mineral, or carbonate, portion and contains related infor-mation on diet (e.g., Krueger & Sullivan 1984;
Lee-Thorp 1989; Ambrose & Norr 1993). Tooth enamel – and the carbonate and phosphate minerals where carbon is bound – forms during childhood. Bone collagen provides a record of adult diet; tooth enamel is a record of the diet of early childhood. Another important difference in stable carbon isotope ratios between these two components of bone tissue lies in the source of the carbon. Experimental studies have demon-strated that collagen carbon comes largely from dietary protein, while the apatite carbon in bone and enamel more accurately reflects the isotopic composition of the total diet (e.g., Ambrose &
Norr 1993). The results of these experiments permit more detailed reconstruction of the iso-topic composition of prehistoric human diets.
Because the two tissues, apatite and collagen, record different components of diet there is no direct comparison of the two values. In gener-al, however, more negative values reflect more terrestrial diets composed of C3 plants. Less negative values are a result of marine foods or C4 plants or both in the diet. In the Iron Age of Scandinavia the only C4 food of potential significance would have been millet. Fish and other marine resources may have placed a large role in human diets in some areas.
Human Enamel and Isotopes at Uppåkra
We measured strontium, carbon, and oxy-gen isotopes in the tooth enamel from the remains of 10 individuals found at Uppåkra.
Basic information on the samples and the bur-ial context is provided in Table 4. Isotopic measurments are listed in Table 5. Summary statistics for the isotopic measurements from these human remains are presented in Table 6.
The average value for 87Sr/86Sr from the 10 human samples was 0.7132 ± 0.0024 with a wide range of values from 0.7111 to 0.7191.
A bar graph of the ranked 87Sr/86Sr values pro-vides some sense of the variation present (Fig.
5). There is a regular gradient from the lowest value at 0.7111 to the eighth sample at 0.7129.
There are two remaining high values at 0.7157 and 0.7191 that are clearly distinct from the remainder and without question represent non-local individuals.
The two high 87Sr/86Sr individuals have dis-tinctive depositional contexts. The tooth from sample 27 (87Sr/86Sr = 0.7157) comes from a disarticulated mandible, perhaps from one Fig. 4. Oxygen isotope ratios in modern precipita-tion in Sweden (Burgman et al. 1987, Fig. 1).
170 T. DOUGLAS PRICE
of the four victims of the fire that destroyed House 24 adjacent to the ceremonial house.
House 24 is thought to have been an elite res-idence based on the quantity of glass vessels and jewelry. The individual represented by sample 21 (87Sr/86SrSr = 0.7191) comes from the only inhumation yet known at Uppåkra, from a burial excavated in 1934 not far from House 24 and the ceremonial structure. This male individual, ca. 40 years of age, was bur-ied with three pots and a comb. Light isotope analysis of the bone collagen of the individual produced distinctive δ15N values thought to indicate freshwater fish consumption, a diet not observed in other individuals from Uppåkra.
A question remains about the place of ori-gin of the eight lower human values. Are they local or non-local? These ratios are for the most part higher that the expected local value based on the pigs. They are largely within the range of the cattle from the site. This question must remain unanswered until more information is available regarding isotope sources within the region of Uppåkra, but at present it would
appear that all these individuals fall within the range of the animals analyzed from the site.
Carbon and oxygen isotopes provide some additional information.
Table 3. Summary statistics for strontium, carbon, and oxygen isotope ratios for ten sam-ples of human tooth enamel from Uppåkra, Sweden.
87Sr/86Sr δ13C δ18O
Mean 0.7132 -14.5 -5.0 1 StDev 0.0024 0.5 0.9 Minimum 0.7111 -15.7 -6.8 Maximum 0.7191 -13.9 -3.3
The oxygen isotope ratios measured in the 10 enamel samples from the site averaged -5.0‰
± 0.9 and ranged between -3.3‰ and -6.8‰, largely within the expected range for south-western Sweden based on modern precipita-tion. A plot of δ18O vs. 87Sr/86Sr provides fur-ther information (Fig. 6). In this plot fur-there is a rather tight cluster of six individuals around 0.712 and -5‰. There are four individuals outside this cluster, including the two samples Fig. 5. Bargraph of ranked 87Sr/86Sr values for fauna and humans from Uppåkra.
171
HUMAN MOBILITY AT UPPÅKRA
with very high 87Sr/86Sr values (samples 27 and 21) identified previously. These two indi-viduals have δ18O values similar to the other six suggesting they might be from southern Sweden as well but from an area with higher
87Sr/86Sr values, perhaps in the region of the Fennoscandian Shield in eastern Scania or southwestern Sweden. There are two individ-uals with 87Sr/86Sr values similar to the local cluster of six but with much higher (sample 30) and lower δ18O (sample 28) values. These individuals may also be non-local and have come from the north (more negative) and south (less negative) respectively.
Carbon isotopes show relatively little var-iation among the Uppåkra samples with a mean of -14.5‰ ± 0.5, with a range from -13.9‰ to -15.7‰. A plot of δ13C vs. δ18O is informative (Fig. 7). The one very negative value (sample 30) is quite distinct and suggests a more terrestrial diet for this individual with a local 87Sr/86Sr signature. This individual also has the least negative δ18O suggesting more northerly origins that might fit with a more terrestrial diet.
Conclusions
The preliminary isotopic data from the Swedish Iron Age site of Uppåkra provides an intriguing glimpse at past mobility and diet. Strontium and oxygen isotope ratios help to identify local vs. non-local status and together suggest that four of the individuals (21, 27, 28, and 30) of the ten in the sample are non-local to the site. Important to remember that identify-ing non-locals is rather straightforward, while determining place of origin is much more dif-ficult because of the wide spread occurrence of certain 87Sr/86Sr values and the poorly under-stood variation in δ18O. If the oxygen data are reliable, they suggest that two individuals, 30 and 28, may be coming from the north and south respectively. It is also of interest to note that some of the cattle appear to have been mobile while the pigs appear largely local.
While it is clear that the preliminary study has answered our initial questions concerning the reliability, variation, and proveniencing of the Uppåkra human remains, significant more research needs to be done to fill in this first glimpse. Several steps are essential. The inves-tigation of bioavailable 87Sr/86Sr in the region of Uppåkra must be done to determine the Fig. 6. A scatterplot of δ18O vs. δ13C for 10 human enamel
samples from Uppåkra. Fig. 7. A scatterplot of δ18O vs. δ13C for 10 human enamel samples from Uppåkra.
172 T. DOUGLAS PRICE
proximity of the Tornqvist line and the higher strontium isotope ratios associated with that geological feature. More human samples need to be analyzed to determine if the patterns observed in this preliminary study are strong and reproducible. More consideration of the age, sex, status, and archaeological context of the sampled individuals is required in order to better understand the nature of mobility in this ancient time and place.
Acknowledgements
This report owes its completion to a number of individuals, including Karl-Göran Sjögren for his permission to reproduce Fig. 2, and to several individuals for collaboration on vari-ous projects in Sweden that have resulted in the baseline data we have available. That list includes Karl-Göran, Caroline Arcini, Karin Frei, Ola Magnell, and Lars Larsson. Thanks to all.
References
Ambrose, S.H. & Norr, L. 1993. Experimental evidence for the relationship of the carbon iso-tope ratios of whole diet and dietary protein to those of bone collagen and carbonate, Lambert, J. B. & Grupe, G. (eds.), Prehistoric Human bone.
Archaeology at the Molecular Level. Berlin. 1–37.
Benson, L., Cordell, L., Vincent, K., Taylor, H., Stein, J., Farmer, G. & Kiyoto. F. 2003. Ancient maize from Chacoan great houses. where was it grown? Proceedings of the National Academy of Sciences 22. 13111–13115.
Bentley, R.A. 2006. Strontium isotopes from the earth to the archaeological skeleton. a review.
Journal of Archaeological Method and Theory 13.
135–187.
Bowen, G.J., & Revenaugh, J. 2003. Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39. 1299.
Bryant, J., Koch, P., Froelich, P., Showers, W. &
Genna, B. 1996. Oxygen isotope partitioning
between phosphate and carbonate in mamma-lian apatite. Geochimica et Cosmochimica Acta 60. 5154–5148.
Burgman, J.O., Calles, B. & Westman, F. 1987.
Conclusions from a ten-year study of oxygen-18 in precipitation and runoff in Sweden, Isotope Techniques in Water Resource Development, IAEA-SM-299/107. 597–590.
Chenery, C.A., Pashley, V., Lamb, A.L., Sloane, H.J.
& Evans, J.A. 2012. The oxygen isotope relation-ship between the phosphate and structural carbo-nate fractions of human bioapatite. Rapid Com-munications in Mass Spectrometry 26. 309–319.
Dansgaard, W., Johnsen, S.J., Moller, J. & Langway Jr., C.C. 1969. One thousand centuries of climate record from Camp Century on the Greenland Ice Sheet. Science 166. 377–381.
Ekström, G. 1936. Skånes moränområden. Svensk Geografisk Årsbok 12. 7–77.
Frei, K.M., Frei, R., Mannering, U., Gleba, M., Nosch, M.L. & Lyngstrøm, H. 2009. Provenance of ancient textiles — a pilot study evaluating the strontium isotope system in wool. Archaeometry 51. 252–276.
Frei, K.M. & Price, T.D. 2012. Strontium Isotopes and Human Mobility in Prehistoric Denmark.
Journal of Anthropological and Archaeological Sci-ences 4. 103–114.
Graversen, O. 2009. Structural analysis of superpo-sed fault systems of the Bornholm horst block, Tornquist Zone, Denmark. Bulletin of the Geo-logical Society of Denmark 57. 25–49.
Gustafsson, M.E.R. & Franzén, L.G. 2000. Inland transport of marine aerosols in southern Sweden.
Atmospheric Environment 34. 313–325.
Hedman, K.M., Curry, B.B., Johnson, T.M., Ful-lagar, P.D. & Emerson, T.E. 2009. Variation in strontium isotope ratios of archaeological fauna in the Midwestern United States. a preliminary study.Journal of Archaeological Science 36. 64–73.
Houmark-Nielsen, M. 1999. A lithostratigraphy of Weichselian glacial and interstadial deposits in Denmark. Geological Society Denmark, Bul-letin 46. 101–114.
Knudson, K.J. 2008. Tiwanaku influence in the south central Andes. strontium isotope analysis and Middle Horizon migration. Latin American Antiquity 19. 3–24.
Kohn, M.J. 1996. Predicting animal δ18O. Accoun-ting for diet and physiological adaptation Geochi-mica et CosmochiGeochi-mica Acta 60. 4811–4829.