Isotopes and Red Herrings: Comments on Milner et Al. and Liden et Al
Hedges, R. E. M., Antiquity
These comments are primarily in response to the paper by Milner et al. who argue in favour of a far less clear cut interpretation of the body of isotopic data of human bone collagen isotopic composition at the Mesolithic--Neolithic transition, and, if applying the interpretation according to current assumptions, to give more cautious generalisations about the main dietary sources. The argument is driven by perceived discrepancies between estimating the major dietary resource from isotopic evidence and from field excavation data.
However, it is difficult to document the discrepancies unless the different kinds of evidence are available in the same place. For example, where Neolithic shell mounds and/or evidence for fishing are very clear, there are virtually no human isotopic data available that conflicts with. a marine dietary resource. Furthermore it is not at all clear how one might assess the contribution of marine resources (defined as fraction of dietary protein, for the sake of argument) from such evidence as the quantity of surviving fish bones, or shells in mounds. Moreover, Milner et al's argument is based most strongly on the situation around the Baltic and North Sea, while the isotopic evidence they discuss is from sites mainly in the western UK. However, those isotopic results could be expanded to include data from the Channel Islands, Brittany and the W coast of France, arguably the coast of Portugal, and in fact the same picture is also applicable to those Neolithic sites studied in the Mediterranean e.g. (Richards et al. 2001; Papathanasiou 2003).
My main purpose, however, is to reflect on the interpretation of the isotopic evidence from human bone. Part of the argument of Milner et al is based on there being too few human individual results from which to generalise with certainty. This can only be countered at a very detailed level--but the fact is that scores of human Neolithic bones from dozens of coastal sites over a very wide range of European coastline contain, without exception, terrestrial-value isotopic collagen. It is quite safe to assert that Mesolithic human bone frequently exhibits collagen isotopic composition enriched in [sup.13]C compared to accepted terrestrial values, which can be interpreted, as explained by Milner et al, as due to the consumption of between 25 and 90 per cent protein--depending on the value measured but often in the higher range--from marine sources. In contrast, all Neolithic--period human bone so far measured, from coastal sites, although comparatively uncommon, exhibits collagen isotopic composition which does not significantly differ from terrestrial values.
There is no dispute in Milner et al concerning results which exhibit high marine resources. The problem really arises for the situation where, on non-isotopic evidence, significant marine protein is apparently consumed, but is not registered as a detectable enrichment of [delta][sup.13]C in human bone collagen, and it is this point which is the crux of the matter.
There are four main issues that bear on the interpretation of partial marine diets:
1. How well the terrestrial "end point" can be defined (in terms of [delta][sup.13]C values). Put very roughly, a change of 1 [per thousand] corresponds to about 10 per cent change in marine intake. If the terrestrial end-point is uncertain to [+ or -] 1 [per thousand], this allows a difference of about 20 per cent in deciding what is a minimum detectable level of marine consumption. This is likely to be a worst case--but points up the need to present relevant data and consider the errors involved, when aiming for quantitative statements at diets with a minor marine component.
2. Variation in the enrichment of [sup.13]C in food chains. Milner et al present data which show that estuarine foodchains are liable to be less enriched than offshore and/or high energy littoral environments. This could apply to molluscs, crustacea and detritivorous fish. …