J Archaeol Sci 22 (1)

Considers current practice of submitting archaeological specimens for blood residue analysis. The accuracy of techniques derived from modern medical and forensic procedures has yet to be verified. A central question relates to blood protein survival over time and the relative capability of the different laboratory methods to detect, identify, and discriminate accurately between the ancient degraded proteins. The paper is intended to address some of these issues, and to sensitise both researchers and archaeologists to the need for caution in reporting and interpreting unconfirmed test results on ancient blood proteins.

In fossil mammal bones and teeth from Marillac (40,000-45,000 years BP), collagen has been found to be preserved. This collagen presents a similar amino acid composition to modern collagen. An isotopic difference between collagen of herbivorous and carnivorous species is clearly documented in this paper and it is shown that the decrease in the amount of collagen in bones and teeth through diagenesis has not altered the original isotopic signal. It has been possible to demonstrate the carnivorous diet of Neanderthals. In addition, more subtle isotopic differences have been documented between the different species within each trophic group consumed. These differences can be linked to physiology, or to particular kinds of food of each animal species. Another isotopic signal due to nursing effects has also been observed. A young mammal suckling its mother's milk is one trophic level higher than an adult of the same species. Consequently, the collagen of deciduous teeth records an isotopic signal from the food consumed before weaning. Also, in species where most of the tooth growth occurs before weaning time, the collagen of permanent teeth records, in part, information about the milk diet. Thus the 15N composition of such teeth is different from the isotopic composition of bones in the same individual. The variation in the 15N composition of bone collagen along the stratigraphical sequence of the site shows that this isotopic signal may also be used for palaeoclimate reconstruction; for example, evidence of an arid period is seen in layer 7 in Marillac.

Replies to a recent article (see 95/943) criticising an earlier paper by the current authors. Refutes the critique on the grounds that it did not adequately consider the goals of the analysis or zooarchaeology. It is thought more important to address empirical issues of correctness than abstract issues of statistical validity. The distinction is felt to be important when choosing statistical approaches.

Botanical data from an ethnoarchaeological study of cereal and pulse crops (Greece) are used to explore alternative sources of mixed crop samples. In addition to more or less pure `monocrops', deliberately mixed `maslins' were grown to exploit the tendency of individual maslin components to perform more or less well under different growing conditions. The composition of these maslins was highly variable. Crop processing may introduce systematic bias into the composition of crop samples and is also deliberately used to manipulate the relative proportions of maslin components in a highly flexible manner. Both monocrops and maslins contain low-level contamination by other cultigens, and it is shown that these were mainly introduced with the seed corn and not through crop rotation or mixing on the threshing floor. Minor contaminants resembling the dominant cultigens in growth habit, seed size, etc, are not only hard to remove but also tend to be tolerated. The implications of these observations are discussed for the interpretation of mixed archaeobotanical crop samples.

Consideration of the mortality distributions of 206 European and Near-Eastern Neanderthals (forty associated skeletons and 166 isolated elements), compared to those of eleven recent human ethnographic and palaeodemographic samples and two non-human mammalian samples, indicate that there is a clear representational bias in the total sample, with too few infants and older adults plus too many adolescents and prime-age adults. Manipulations of the Neanderthal data produce immature mortality distributions within the ranges of the recent human samples, but they maintain the high prime-age adult and low older adult mortality. This high young adult mortality rate is likely to be the product of a combination of demographic stress, the effects of pooling across temporally and geographically diverse and fluctuating populations, the need for full mobility among all individuals and hence a dearth of older individuals dying in shelters. The first three factors support the interpretations of high levels of adaptive stress previously suggested by palaeopathological analyses of their remains.