Isotopic signatures and hereditary traits: snapshot of a Neolithic community in Germany.

Bentley, R. Alexander ; Wahl, Joachim ; Price, T. Douglas 等

Introduction

The site of Talheim (c. 4900-4800 BC), a late Linearbandkeramik (LBK) community in the Neckar valley of Germany, yields a truly unique picture of Neolithic life because, unlike most cemeteries, Talheim very probably represents a group of people who lived al the same time and were then killed in a single event (Wahl 1985; Wahl & Konig 1987; Wild el al. 2004; Price el al. 2008).

Talheim is located about 10km south of Heilbronn and about 30km NNE of the LBK site of Vaihingen (Figure 1; Price el al. 2008; Krause 2000). The remains of 34 individuals recovered there include 18 adults and 16 children, all buried in a single pit 3m long (Wahl & Konig 1987; Figure 2). Radiocarbon dates of the bones are consistent with all individuals having lived around 4900-4800 BC (Alt et al. 1997; Wild et al. 2004). The individuals were almost certainly killed in a single attack--most of the skulls show violent injuries and 20 were killed by blows to the head or arrow-wounds from behind as if the victims had tried to flee (Wahl & Konig 1987). The 18 adults include nine males, seven females and two of undetermined sex (Wahl & Konig 1987; Alt et al. 1995).

[FIGURE 1 OMITTED]

From a previous independent analysis by Alt et al. (1995) of non-metric, odontological characteristics of the teeth (belonging to the so-called epigenetic variants, or 'discreta'), it appears that Talheim was a homogeneous and isolated population. Figure 3 shows the way in which Alt et al. (1995: Figure 2) summarised their major results, showing which Talheim individuals possess relatively rare heritable traits, numbered 164, 333, 554 and 673. The lines in Figure 3 connect individuals of particular similarity (which may include similarities involving traits other than these four). In the centre, there are three individuals (84-2, 84-23 and 83-15A) that Alt et al. (1995: 214-15) suggest could be a father and his two children. On the left, sharing traits 554 and 673, are two adult males (83-7 and 83-18B) who might be brothers (Alt et al. 1995: 214; and see below). One of these 'brothers' (83-18B) is connected by a line to ah adult woman (83-20A) who lacks 554 and 673, because they were particularly similar among other traits (Alt et al. 1995: 214). In our discussion below, we compare our isotope results with these interpreted genetic relationships published a decade ago by Alt et al. (1995; also Alt et al. 1997 for version in French). By comparing our isotope results with the skeletal morphology, we find that diet, geographic mobility and genetic relatedness were all correlated within this community.

[FIGURE 2 OMITTED]

Methods

Strontium, carbon and oxygen isotopes were analysed in the tooth enamel of the Talheim individuals, in order to characterise the variation in their geographic origins and possibly variation in diets. Samples were selected from all skeletons for which a tooth was available, including 22 human individuals. Samples of the mandible/maxilla were taken from seven of those individuals.

Strontium isotopes ([sup.87]Sr/[sup.86]Sr) in archaeological tooth enamel serve as a geographic 'signature' from an individual's childhood, when the enamel was forming (e.g. Price et al. 2002; Bentley 2006). For this region of southern Germany, our 'map' of the biologically-available strontium isotope signatures (Price et al. 1998; 2001; 2003; 2008; Bentley et al. 2003; Bentley & Knipper 2005a) shows generally that the lowlands of southern Baden-Wurttemberg, including the Neckar Valley, are characterised by [sup.87]Sr/[sup.86]Sr ratios between about 0.708 and 0.710, whereas the nearest crystalline mountains, including the Vosges, Black Forest and parts of the Odenwald, exhibit [sup.87]Sr/[sup.86]Sr above 0.711. As Bentley and Knipper (2005a: Figure 3a) found, enamel from lowland pigs have [sup.87]Sr/[sup.86]Sr between 0.708 and 0.710, and pigs from the uplands have ratios above 0.710, with one of two exceptions suspected to be from traded pigs.

Determined largely by temperature, the mean annual oxygen isotope composition ([[delta].sup.18]O relative to Standard Mean Ocean Water [SMOW]) in precipitation depends on latitude and altitude (Bowen & Wilkinson 2002), but also on topographic relief, distance from large bodies of water, and relative humidity. Geographic origins are reflected by enamel [[delta].sup.18]O values, since mammals took in much of their oxygen through ingested water (e.g. D'Angela & Longinelli 1990; Kohn 1996; Balasse et al. 2002). Due to averaging, however, seasonal 8180 variations within the tooth enamel of mammals (those with continuous dental growth) are less than in the precipitation, of the order of about 5 [per thousand] or less in cows from South Africa, for example (Balasse et al. 2002). In humans from settled communities, whose water supply is more consistent, the variations ate even smaller, with tooth enamel [[delta].sup.18]O varying about 1[per thousand] among prehistoric people from one place (Budd et al. 2004).

[FIGURE 3 OMITTED]

In southern Germany, the 1991 average of [[delta].sup.18]O in precipitation varied geographically from -8.2[per thousand] to -11.2[per thousand], becoming more negative moving eastwards away from the Atlantic Ocean, and southwards with increasing altitude towards the Alps (Tutken et al. 2004). This geographic gradient was probably broadly similar during the Neolithic, even if perhaps 2-3 more negative overall (Mayer & Schwark 1999: Figure 4). From their regional sampling of prehistoric pig enamel, Bentley and Knipper (2005a) found that [[delta].sup.18]O in Iron Age pigs was moderately well correlated with values from contour map values for [[delta].sup.18]O in modern precipitation. Interestingly, this correlation was not present for the Neolithic pigs (Bentley & Knipper 2005a), which could reflect more varied water sources for Neolithic livestock; resolving this question may require future studies of palaeogroundwaters (Edmunds & Tyler 2002).

Within the set of [C.sub.3] plants ([C.sub.4] plants ate not relevant to this study), carbon isotope signatures ([delta][sup.13]C, relative to the Pee Dee Belemite [PDB] carbonate standard) correlate positively with temperature and sunshine, and negatively with humidity and precipitation (e.g. Heaton 1999). For humans, [delta][sup.13]C values in tooth enamel carbonate reflect an average of the whole diet, offset by--9.4[per thousand] such that a pure [C.sup.3] vegetarian would have a [delta][sup.13]C value of about--13[per thousand] in enamel (Ambrose & Norr 1993). Assuming that the Talheim individuals lived contemporaneously in a single community, experiencing similar levels of temperature, humidity, and insolation, other possible causes of [delta][sup.13]C variations include meat consumption (e.g. DeNiro & Epstein 1978), altitude (e.g. Korner et al. 1991), and forest density (e.g. Heaton 1999). With so many possible causes, small variations in [delta][sup.13]C can be difficult to explain, but nonetheless clear patterns can be observed. Among medieval pig teeth from southern Germany, [delta][sup.13]C was generally greater (p < 0.01) in upland pigs than in lowland pigs by about 1[per thousand] (Bentley & Knipper 2005a: Figure 3a), which might be an effect of altitude and/or lack of forest canopy in the medieval uplands.

Procedures

For strontium isotope analysis, each enamel sample (5-20mg) was dissolved in 5 M HN[O.sup.3], and then purified by extraction chromatography using Eicrom[R] Sr-spec resin (e.g. Bentley et al. 2003 for details). We then analysed [sup.87]Sr/[sup.86]Sr by thermal ionisation mass spectrometry (TIMS) at Cornell University and the National Oceanography Centre, Southampton (see Table 1). We then measured [[delta].sup.18]O and [delta][sup.13]C in the carbonate (C[O.sub.3]) component of the tooth enamel, following a tested procedure (Koch et al. 1997; Balasse et al. 2002), in which about 5mg of tooth enamel was mechanically cleaned of all dentine, powdered and soaked overnight in 5 per cent acetic acid to remove post-burial carbonate contamination, then analysed via a Kiel III automated cryogenic distillation system interfaced with a ThermoFinnigan Mat 253 gas-source mass spectrometer at the Bloomsbury Environmental Isotope Facility (BEIF), University College London (cf. Bentley & Knipper 2005a). Repeated analyses of the NBS 19 standard yielded a precision better than 0.1[per thousand], (1 s.d.) for [[delta].sup.18]O and 0.05[per thousand] for [[delta].sup.18]O.

Results

Our results are shown in Tables 1 and 2. With no local animals among the archaeological remains, we do not attempt to define a strict 'local' [sup.87]Sr/[sup.86]Sr range for Talheim. Nonetheless, we get a good sense of the local range for this community from the clear clustering ofvalues (cf. Price et al. 2008). Firstly, the seven human bone samples from Talheim (Table 2) have a mean [sup.87]Sr/[sup.86]Sr of 0.70906 [+ or -] 0.00005, which yields a [+ or -] 2 s.d. range of 0.70895 0.70916. Secondly, the tire sampled children's teeth (useful if the Talheim children were locally raised) have a mean of 0.70910 [+ or -] 0.00005, yielding a [+ or -]2 s.d. range of 0.70899 - 0.70921. By taking the combined extent of these two ranges, we conservatively estimate the local [sup.87]Sr/[sup.86]Sr range to be 0.70895-0.70921. This range is shown in Figure 4, which also shows the ranges derived from archaeological pig teeth at Vaihingen (0.70913-0.70979) and Ilsfeld (0.70947-0.71005), two nearby sites (Bentley et al. 2004; Bentley & Knipper 2005a). Notably, the Talheim range is considerably tighter than either of these other ranges, which adds further support to the Talheim sample consisting of contemporaries from a single community.

[FIGURE 4 OMITTED]

Even allowing that our estimate of the local range is approximate, several [sup.87]Sr/[sup.86]Sr values are obviously well above it (Figure 4). Suggesting an upland component to the diet during childhood, these higher [sup.87]Sr/[sup.86]Sr values are not local to the Talheim community. We call this Group 3, which includes two females (83-10B, 83-20A) and two males (83-7 and 83-18B). The two males in Group 3 ate in fact the above-mentioned 'brothers' identified by Alt et al. (1995). Although the two males do not share any special genetic similarity with the Group 3 females, all four members of Group 3 were close to 30 years old (Figure 4). One of the two Group 3 women (83-20A) has by far the highest [[delta].sup.18]O (27.21) of the Talheim samples.

Among the rest, who have signatures within or very near the local range, we can identify two clear groups by plotting [sup.87]Sr/[sup.86]Sr vs. 8180 (Figure 5a), such that 12 individuals in what we call 'Group 1' have a mean [[delta].sup.18]O of 26.10 [+ or -] 0.14, and the five individuals of 'Group 2' have a mean of 25.42 [+ or -] 0.14. As Figure 5a shows, Groups 1 and 2 ate quite distinct (Table 3), with the [[delta].sup.18]O gap between them about as large as the variation within each group.

Adding the carbon isotope data, we also plot [[delta].sup.18]O vs. [delta][sup.13]C (Figure %) and [sup.87]Sr/[sup.86]Sr vs. [delta][sup.13]C (Figure 5c), while retaining the same group symbols used in Figure 5a. In Figure % the data fall into two arrays: a horizontal array with relatively uniform [sup.87]Sr/[sup.86]Sr, and a diagonal array in which [sup.87]Sr/[sup.86]Sr correlates linearly with [delta][sup.13]C ([r.sup.2] = 0.89 for 10 data points). Group 3 falls exclusively within the diagonal array, and Groups 1 and 2 fall within the horizontal array, including where it overlaps with the lower part of the diagonal array.

[FIGURE 5 OMITTED]

In Figure 6, we again plot [sup.87]Sr/[sup.86]Sr vs. [delta][sup.13]C as in Figure 5c, but this time using the data symbols to represent the genetic traits from Figure 3. The result is striking: all six of the individuals possessing genetic trait 164 and/or 333 plot along the horizontal array, and the three with trait 554 plot along the diagonal array. Taking the diagonal array to contain 10 points and the horizontal array to contain 10 points, the binomial distribution probability p that these patterns occurred randomly is significantly small for trait 554 occurring with the diagonal array (p < 0.05), as well as traits 164 and/or 333 occurring with the horizontal array (p < 0.02) and not with the diagonal array (p < 0.03). With only three instances of trait 554, its absence from the horizontal array could still be due to chance (p = 0.20). The fourth trait, 673, does not show a pattern on Figure 6. However, the four males for which 673 occurs exhibit a range of [[delta].sup.18]O values (25.99 [+ or -] 0.11) that is narrow enough to be significantly associated with the trait (p < 0.05). Finally, among the six remaining individuals without any of these four traits, five occur within the diagonal array (Figure 6).

[FIGURE 6 OMITTED]

Discussion

Three groups were defined from strontium and oxygen isotopes (Figure 5a, Table 3). Group 1 contains no adult females: only males, young children, and an 11-year-old girl. Since we did not analyse isotopes in all the Talheim females (some tooth samples not available), it is always possible that one or more of the females not sampled could belong to Group 1. However, none of the four adult females that we did analyse (84-4, 83-20a, 83-10b and 83-22D) fell within Group 1, and given that 10 out of 15 (67 per cent) of the other samples were in Group 1, the probability that all four adult females would rail outside Group 1, by chance alone, is less than 2 per cent. In any case, this result is at least consistent with the women from this group having been extracted during the attack (e.g. Wild et al. 2004). Since adult women are present in Group 2 and Group 3, it appears that the women of Group 1 were selectively removed, and given the circumstances, presumably captured. It would seem that Group 1 represents the local Talheim community, since it contains all the young children we analysed. These would presentan interesting question: why were the women of Groups 2 and 3 killed, while the women of Group 1 were abducted?

Group 2 is equally interesting, but for a different reason. The five members of Group 2--a man in his 20s, a woman ofabout 20, a boy and a girl both about 11 years old, and a mature woman of about 50--have a pattern of genetic relatedness that suggests they were a single family. Of the three individuals at the centre of Figure 3, Group 2 contains the 'father' (84-2) and the 11-year-old 'daughter' (83-15A) mentioned above as identified by Alt et al. (1995), but not the 'son' (84-23), who fell into Group 1 and yielded the highest [delta][sup.13]C (-11.83) of the entire Talheim sample. However, the 11-year-old boy (83-15B) in our Group 2 seems just as likely to be the son, as he shares traits 164 and 333 with the potential father, and in fact Alt et al. (1995) had linked them both by a line of genetic similarity (Figure 3). As one would expect for the potential mother (84-4, upper left in Figure 3), she has relatively little genetic similarity with the 'father', which would mean that the 'children' inherited traits 164 and 333 paternally. The 'grandmother' (83-22D) would be on the mother's side, since she too lacks traits 164 and 333, and her [delta][sup.13]C value is identical to hers (-13.8). Although this interpretation is conjectural, it should be no surprise to discover an immediate family within a single community, and the particular Group 2 membership of a man, a woman, two similarly aged children, and a woman of the previous generation seems unlikely to be a chance combination. Furthermore, a re-examination of the skeletal morphology by Joachim Wahl found no striking differences among the Group 2 individuals, consistent with their being close family members.

From strontium isotopes, it would appear that the females in Group 3 (83-10B, 20A), one of the males (83-18B) and less certainly the other (83-7), had grown up in a locality distant from the childhood home of the rest of the massacred community. The men were determined as being of close kindred by Alt et al. (1995), possibly even brothers. They both rail within the diagonal array in Figure 5c, and their [[delta].sup.18]O values (25.98 and 25.91) are nearly identical. However, because their [sup.87]Sr/[sup.86]Sr values (0.70941 vs. 0.71067) are quite different from each other, they probably did not grow up in exactly the same locality. One possible explanation is that they were not brothers but father and son (given their difference in age at death), who grew up in different villages, of whose different [sup.87]Sr/[sup.86]Sr values reflect the different places and times of enamel formation as they moved across a territorial range, perhaps as livestock herders.

The diagonal array containing Group 3 (Figure 5c) suggests a single group within the community which falls along a mixing line (cf. Bentley 2006; Montgomery et al. 2007; Phillips & Koch 2002), with one end-member in the regional uplands, with higher [sup.87]Sr/[sup.86]Sr and less-negative [delta][sup.13]C values (Bentley & Knipper 2005a). Constructing a proper mixing model is daunting however, because it would have to include upland vs. lowland meat, plants and milk- at least six different categories of food resources, each with its own Sr concentration, [sup.87]Sr/[sup.86]Sr and [delta][sup.13]C value--and the myriad possible concentrations of those resources.

Nevertheless, following Montgomery et al. (2007), we can consider more simply that a transhumant group had two diets--possibly a summer diet from the uplands, and a winter diet in the lowlands. Both diets may have been quite similar and quite rich in animal protein, in reference to the carbon and nitrogen isotope results from Herxheim (LBK, southern Germany) by Durrwachter et al. (2006), who found 'surprisingly little variation' in human diet. Analyses of archaeological pig tooth enamel (Bentley & Knipper 2005a) reasonably indicate an end-member for the lowland/winter diet would be about [sup.87]Sr/[sup.86]Sr = 0.70910 and [delta][sup.13]C = -13.8[per thousand] (lower left of Figure 5c), whereas an upland/summer end-member could be about [sup.87]Sr/[sup.86]Sr = 0.712 and [delta][sup.13]C = -12.5[per thousand] (towards the upper right). Hence the diagonal array in Figure 5c may represent a group that practiced transhumance, with different proportions of upland/lowland resources accounting for their spread along the array. We present supporting lines of evidence for transhumance below.

Figure 6 shows that certain non-metric traits are restricted to certain isotopically-delineated groups. This implies a correlation between diet, geographic origin and genetic relatedness. One possible explanation for this is that Talheim was differentiated by hereditary subsistence occupations (cf. Bogucki 1988)--perhaps specialised cultivators (horizontal array) and stockherders (diagonal array, including Group 3)? At nearby Vaihingen, there were significantly more 'non-local' strontium isotope signatures among people buried in the ditch surrounding the settlement than people buried within the settlement (Bentley et al. 2003). Although the non-locals may have been immigrants from other villages, they might also have acquired their non-local signatures through herding livestock, as the Neolithic cattle, sheep and goats from Vaihingen also show a range of [sup.87]Sr/[sup.86]Sr values (Bentley et al. 2004). In fact, transhumance at Vaihingen was recently confirmed by analysing cow enamel samples at regular intervals along the tooth, yielding a continuous [sup.87]Sr/[sup.86]Sr record for the first two years of the cow's life. The results (Bentley & Knipper 2005b) from three Vaihingen cows show that one was clearly taken into the uplands during the summer to pasture, while the other two were taken to different places during the summer. Hence it could be that cultivators and pastoralists in the LBK shared the same communities yet maintained social, hereditary and possibly even 'ethnic' distinctions between their groups.

Conclusions

We analysed strontium, oxygen and carbon isotopes in human tooth enamel samples from 23 Talheim individuals. Firstly, when we plot [[delta].sup.18]O vs. [sup.87]Sr/[sup.86]Sr (Figure 5a), the data form three groups, one (Group 3) with [sup.87]Sr/[sup.86]Sr values from upland sources, and the others (Groups 1 and 2) with similar, lowland [sup.87]Sr/[sup.86]Sr values but distinct in their 8180 values. Secondly, when we plot [delta][sup.13]C vs. [sup.87]Sr/[sup.86]Sr (Figure 5c), we find two arrays of data points, of which the diagonal array is suggestive of increasing proportions of upland sources to the diet, potentially due to herding livestock in the uplands.

When compared to the previous results of skeletal morphological analysis by Alt et al. (1995), the isotope analyses reveal three patterns:

(1) Group 2 (Figure 5a) consists of an adult male, a boy and a girl all inferred by Alt et al. (1995) to be closely relate& as well as an adult female and an much older female who ate not closely related to them. Given the isotope clustering, the relatedness between the man and the children, and the suggestive combination of members, we suggest that Group 2 was a nuclear family.

(2) Group 1 (Figure 5a) has no adult women, yet it contains four adult men, one 11-year-old girl, two boys aged 6-8, and all five of the sampled young children. The presence of young children (absent in Groups 2 and 3) suggests Group 1 was the local community. The lack of adult females (present in Groups 2 and 3) suggests the women of Group 1 were spared, and apparently abducted, by those who killed the others.

(3) The four members of Group 3, all adults, appear to be genetically related. Falling along a diagonal array in Figure 5c, three of them show distinctive upland [sup.87]Sr/[sup.86]Sr ratios associated with their upbringing. A possible explanation is that specialised cultivators represent the horizontal array and stockherders represent the diagonal array, with both specialisations being passed down hereditarily.

Our evidence for the violent abduction of young females from Group 1 rather dramatically supports Eisenhauer's (2003) case for patrilocality at Talheim, and also bolsters the isotopic (Bentley et al. 2002) and genetic (Seielstad et al. 1998) evidence for patrilocality in Neolithic Europe in general. Anthropologically-speaking, this fits logically with the evidence for transhumance in the LBK (Bogucki 1988; Kienlin & Valde-Nowak 2003; Bentley & Knipper 2005b) and the possible nuclear family at Talheim. The nuclear family is most viable within patrilineal kinship systems, whereby men seek to control the inheritance of their descendants (e.g. Fox 1983: 27-53). Transhumance is also highly correlated with patrilocality (e.g. Holden & Mace 2003), in the sense that livestock and land ate valuable elements of property that men will seek to control (Holden et al. 2003). The fact that these independent lines of evidence are so consistent with each other encourages us that they truly represent aspects of kinship and community in Neolithic Europe.

Acknowledgements

For providing access to a clean lab and TIMS assistance for the stronrium isotope analyses, we thank Dr Rex Taylor, Dr Matthew Cooper and especially Tina Hayes of the National Oceanography Centre (NOC) in Southampton, and Professor William White of the Corneli University Earth Sciences Department. The Sr, C & O isotope measurements and analysis, conducted by RAB at UCL and NOC, were made possible by a research grant from the Leverhulme Trust (F/07134/Z), excepting those measurements made by RAB at Cornell in April 2000, which were funded by a NSF Doctoral Dissertation Improvement Grant (#0073721).

Received: 2 August 2006; Revised: 22 September 2006; Accepted: 18 January 2007

References

ALT, K.W., W. VACH &J. WAHL. 1995. Verwandtschaftsanalyse der Skelettreste aus dem bandkeramischen Massengrab von Talheim, Kreis Heilbronn. Fundberichte aus Baden-Wurttemberg 20: 195-217.

--1997. La reconstitution "genetique" de la population de la fosse rubanee de Talheim, Alllemagne. Cahiers de l'Association pour la Promotion de la Recherche Archeologique en Alsace 3:1-8.

AMBROSE, S.H. & L. NORR. 1993. Experimental evidence for the relationship of the carbon isotope ratios of whole diet and dietary protein to those of bone collagen and carbonate, in J.B. Lambert & G. Grupe (ed.) Prehistoric Human Bone: Archaeology at the Molecular Level: 1-37. Berlin: Springer.

BALASSE, M., S.H. AMBROSE, A.B. SMITH & T.D. PRICE. 2002. The seasonal mobility model for prehistoric herders in the south-western cape of South Africa assessed by isotopic analysis of sheep tooth enamel. Journal of Archaeological Science 29: 917-32.

BENTLEY, R.A. 2006. Strontium isotopes from the Earth to the archaeological skeleton: a review. Journal of Archaeological Method and Theory 13: 135-87.

BENTLEY, R.A. & C. KNIPPER. 2005a. Geographical patterns in biologically available strontium, carbon and oxygen isotope signatures in prehistoric SW Germany. Archaeometry 47: 629-44.

--2005b. Transhumance at the early Neolithic settlement at Vaihingen (Germany). Antiquity: http://www.antiquity.ac.uk/projgall/bentley/index.html

BENTLEY, R.A., T.D. PRICE, J. LUNING, D. GRONENBORN, J. WAHL & P.D. FULLAGAR. 2002. Human migration in early Neolithic Europe. Current Anthropology 43: 799-804.

BENTLEY, R.A., R. KRAUSE, T.D. PRICE & B. KAUFMANN. 2003. Human mobility at the early Neolithic settlement of Vaihingen, Germany. Archaeometry 45: 471-86.

BENTLEY, R.A., T.D. PRICE & E. STEPHAN. 2004. Determining the 'local' [sup.87]Sr/[sup.86]Sr range for archaeological skeletons. Journal of Archaeological Science 31: 365-75.

BOGUCKI, P. 1988. Forest Farmers and Stockherders. Cambridge: Cambridge University Press.

BOWEN, G.J. & B. WILKINSON. 2002. Spatial distribution of [delta] [sup.18]O in meteoric precipitation. Geology 30: 315-8.

BUDD, P., A. MILLARD, C. CHENERY, S. LUCY & C. ROBERTS. 2004. Investigating population movement by stable isotope analysis: a report from Britain. Antiquity 78: 127-40.

D'ANGELA, D. & A. LONGINELLI. 1990. Oxygen isotopes in living mammals' bone phosphate: further results. Chemical Geology 86: 75-82.

DENIRO, M.J. & S. EPSTEIN. 1978. Influence of dier on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta 42: 495-506.

DURRWACHTER, C., O.E. CRAIG, M.J. COLLINS, J. BURGER & K.W. ALT. 2006. Beyond the grave: variability in Neolithic diets in Southern Germany? Journal of Archaeological Science 33: 39-48.

EDMUNDS, W.M. & S.W. TYLER. 2002. Unsaturated zones as archives of past climates: toward a new proxy for continental regions. Hydrogeology Journal 10: 216-28.

EISENHUER, U. 2003. Jungerbandkeramische Residenzregeln: Patrilokalitat in Talheim, in J. Eckerr, U. Eisenhauer & A. Zimmermann (ed.) Archaologische Perspektiven: Analysen und Interpretationen im Wandel; Festschrift fur Jens Luning zum 65. Geburtstag: 561-73. Rahden: Leidorf.

Fox, R. 1983. Kinship and Marriage: an Anthropological Perspective. Cambridge: Cambridge University Press.

HEATON, T.H.E. 1999. Spatial, species, and temporal variations in the [sup.13]C/[sup.12]C ratios of [C.sub.3] plants: implications for paleodiet studies. Journal of Archaeological Science 26: 637-49.

HOLDEN, C. & R. MACE. 2003. Spread of cattle led to the loss of matrilineal descent in Africa, a coevolutionary analysis. Proceedings of the Royal Society B 270: 2425-33.

HOLDEN, C.J., R. SEAR & R. MACE. 2003. Matriliny as daughter-biased investment. Evolution and Human Behavior 24:99-112.

KIENLIN, T.L. & P. VALDE-NOWAK. 2003. Neolithic transhumance in the Black Forest Mountains, SW Germany. Journal of Field Archaeology 29: 29-44.

KOCH, P.L., N. TUROSS & M.L. FOCEL. 1997. The effects of sample treatment and diagenesis on the isotopic integrity of carbonate in biogenic hydroxylapatite. Journal of Archaeological Science 24: 417-29.

KOHN, M.J. 1996. A predictive model for animal [delta][sup.18]O: accounting for diet and physiological adaptation. Geochimica et Cosmochimica Acta 60:4811-29.

KORNER, C., G.D. FARQUHAR & S.C. WONG. 1991. Carbon isotope discrimination by plants follows latitudinal and altitudinal trends. Oecologia 74: 623-32.

KRAUSE, R. (ed.) 2000. Die bandkeramischen Siedlungsgrabungen bei Vaihingen an der Enz, Kreis Ludwigsburg, Baden-Wurttemberg. Ein Vorbericht zu den Ausgrabungen von 1994-1997. Berichte der Romisch-Germanischen Kommission 79: 5-105.

MAYER, B. & L. SCHWARK. 1999. A 15,000-year stable isotope record from the sediments of Lake Steisslingen, Southwest Germany. Chemical Geology 161: 315-37.

MONTGOMERY, J., J.A. EVANS & R.E. COOPER. 2007. Resolving archaeological populations with St-isotope mixing diagrams. Applied Geochemistry 22: 1502-14.

PHILLIPS, D.L. & P.L. KOCH. 2002. Incorporating concentration dependence in stable isotope mixing models. Oecologia 130:114-25.

PRICE, T.D., G. GRUPE & P. SCHROTER. 1998. Migration in the Bell Beaker period of central Europe. Antiquity 72:405-11.

PRICE, T.D., R.A. BENTLEY, J. LUNING, D. GRONENBORN & J. WAHL. 2001. Prehistoric human migration in the Linearbandkeramik of Central Europe. Antiquity 75: 593-603.

PRICE, T.D., J.H. BURTON & R.A. BENTLEY. 2002. Characterization of biologically available strontium isotope ratios for the study of prehistoric migration. Archaeometry 44:117-35.

PRICE, T.D., J. WAHL, C. KNIPPER, E. BURGER-HEINRICH, G. KURZ & R.A. BENTLEY. 2003. Das bandkeramische Graberfeld vom 'Viesenhauser Hof' bei Stuttgart Muhlhausen: Neue Untersuchungsergebnisse zum Migrationsverhalten im fruhen Neolithikum. Fundberichte aus Baden-Wurttemberg 27: 23-58. Stuttgart.

PRICE, T.D., J. WAHL & R.A. BENTLEY. 2008. Isotopic evidence for mobility and group organization among Neolithic farmers at Talheim, Germany, 5000 B.C. European Journal of Archaeology.

SEIELSTAD, M.T., E. MINCH & L.L. CAVALLI-SFORZA. 1998. Genetic evidence for a higher female migration rate in humans. Nature Genetics 20: 278-80.

TUTKEN, T., T.W. VENNEMANN & H.-U. PFRETSCHNER. 2004. Analyse stabiler und radiogener Isotope in archaologischem Skelettmaterial: Herkunftsbestimmung des karolingischen Maultiers von Frankental und Vergleich mit spatpleistozanen Grossaugerknochen aus den Rheinablagerungen. Prahistorische Zeitschrift 79:89-110.

WAHL, J. 1985. Ein neolithisches Massengrab bei Talheim, Landkreis Heilbronn. Archdologische Ausgrabungen in Baden-Wurttemberg 1984: 30-32.

WAHL J. & H.G. KONIG. 1987. Anthropologisch-traumatologische Untersuchung der menschlichen Skelettreste aus dem bandkeramischen Massengrab bei Talheim, Kreis Heilbronn. Fundberichte aus Baden Wurttem berg 12: 65-193.

WILD, E.M., P. STADLER, A. HAUSSER, W. KUTSCHERA, P. STEIER, M. TESCHLER-NICOLA, J. WAHL & H.J. WINDL. 2004. Neolithic massacres: Local skirmishes or general warfare in Europe? Radiocarbon 46: 377-85.

R. Alexander Bentley, Department of Anthropology, Durham University, 43 Old Elvet, Durham DH1 3HN, UK (Email: r.a.bentley@durham.ac.uk)

Joachim Wahl, RP Stuttgart, Landesamt fur Denkmalpflege, Osteologie, Stromeyersdorfitrasse 3, D-78467, Konsmnz, Germany (Email: Joachim.Wahl@rps.bwl.de)

T. Douglas Price, Department. of Anthropology, University of Wisconsin, 1180 Observatory Dr., Madison, WI 53706-1393, USA (Email: tdprice@wisc.edu)

Tim C. Atkinson, Department of Earth Sciences, University College London, Gower Street, London WC1E 6BT, UK (Email: t.atkinson@ucl.ac.uk)

Table 1. Isotope analyses of human tooth enamel samples from Talheim.
All carbon and oxygen isotope measurements were made at BEIF, and the
Sr isotope measurements were done at the National Oceanography Centre
in Southampton, except those indicated ([section]) that were done at
Keck Isotope Geochemistry Lab, Cornell University. Measurement errors
(1 s.e.) corresponding to the last digits of the value are shown in
parentheses.

 [[delta]. [[delta].
 Age sup.13]C sup.18]O
Individual # Sex (years) (PDB) (SMOW)

83-3B Male 20-40 n.d n.d
83-3C Child 8-10 -12.62 (1) 25.90 (3)
83-6 Child ~2 -13.84 (1) 26.38 (3)
83-7 Male 30-40 -13.48 (8) 25.98 (5)
83-8 Male 50-60 -12.82 (9) 26.14 (7)
83-1OB Female? 20-40 -13.19 (3) 25.47 (6)
83-11 Male 20-30 -13.29 (7) 25.97 (8)
83-12 Male 20-30 -13.72 (11) 26.22 (10)
83-13 Boy 6-7 -13.69 (4) 26.03 (5)
83-14B Child 2-4 -13.06 (2) 26.07 (4)
83-15A Girl ~11 -12.84 (9) 25.46 (9)
83-15B Boy? 10-12 -13.14 (8) 25.55 (7)
83-18B Male? 20-40 -13.07 (2) 25.91 (1)
83-20A Female 20-30 -13.53 (9) 27.21 (11)
83-22C1 Male? 20-40 -12.91 (9) 25.94 (6)
83-22D Female? 40-60 -13.79 (9) 25.32 (21)
83-22J Child 4-6 -13.78 (3) 26.09 (3)
84-2 Male 20-30 -13.41 (2) 25.22 (3)
84-4 Female ~20 -13.84 (2) 25.53 (2)
84-23 Boy? ~8 -11.83 (1) 26.15 (2)
84-28 Girl ~11 -12.90 (6) 26.07 (8)

 Genetic
 Group characters
Individual # [sup.87]Sr/[sup.86]Sr (Fig. 5a) (Alt etal. 1995)

83-3B 0.70895 n.d.
83-3C 0.70917 (2) ([section]) 1
83-6 0.70912 (1) 1
83-7 0.70941 (2) 3 554, 673
83-8 0.70911 (1) 1
83-1OB 0.70993 (2) ([section]) 3
83-11 0.70904 (1) 1 333
83-12 0.70906 (1) 1
83-13 0.70922 (1) 1
83-14B 0.70904 (1) 1
83-15A 0.70884 (1) 2 333
83-15B 0.70876 (1) 2 333, 164
83-18B 0.71067 (1) ([section]) 3 554, 673
83-20A 0.70972 (1) 3
83-22C1 0.70894 (1) 1 673
83-22D 0.70902 (1) 2
83-22J 0.70905 (1) 1
84-2 0.70909 (1) ([section]) 2 333, 164
84-4 0.70913 (2) ([section]) 2 554
84-23 0.70932 (1) 1 333, 164, 673
84-28 0.70912 (1) 1 164

Table 2. Strontium isotopes in human bone samples from Talheim.

Grave # part [sup.87]Sr/[sup.86]Sr

84-2 Mandible 0.70914
83-22J Mandible 0.70908
84-23 Mandible 0.70906
83-6 Occipital 0.70905
83-14B Mandible 0.70905
84-4 Mandible 0.70899
? Maxilla 0.70899

Table 3. Characteristics of the three groups defined in the text and
in Figure 5a, including mean isotope value of the group, and p values
for differences between group means (t-test, two-tailed, assuming equal
variances).

 [sup.87]Sr/[sup.86]Sr [sup.18]O

Group 1 (n = 11) 0.70911 [+ or -] 0.00010 26.09 [+ or -] 0.14
Group 2 (n = 5) 0.70897 [+ or -] 0.00016 25.42 [+ or -] 0.14
Group 3 (n = 4) 0.70993 [+ or -] 0.00054 26.14 [+ or -] 0.75

t-tests (p values)

1 vs. 2 0.051 0.00000032
2 vs. 3 0.0061 0.067
1 vs. 3 0.00018 0.81

 [sup.13]C

Group 1 (n = 11) -13.13 [+ or -] 0.22
Group 2 (n = 5) -13.40 [+ or -] 0.78
Group 3 (n = 4) -13.32 [+ or -] 0.22

t-tests (p values)

1 vs. 2 0.57
2 vs. 3 0.73
1 vs. 3 0.57