Neandertals noshed on mammoth meat?

In an article in the Jul 2005 issue of Journal of Human Evolution, Hervé Bocherens and colleagues contribute new isotopic values for bone collagen from three Neandertals, a review of results from prior studies of Neandertal isotope ratios, and an interpretation of the values obtained for the Saint-Césaire specimen based on comparisons with contemporary hyena bones.

The latter part of the study is the most interesting: the Neandertal bones have a substantially higher proportion of nitrogen-15 (15N) than the hyenas. This difference is assumed to come from a difference in the diet. The 15N levels of both species are much higher than plant-eaters, so the difference is apparently not the inclusion of some plant food by Neandertals. Instead, the difference must come eating animals with different 15N enrichment.

But there are good reasons not to believe a word of it; read on for why.

How it works

I've discussed carbon isotope analysis before, click the link for details about how it is related to diet. The bottom line is that carbon-12 is preferentially taken up by plants during photosynthesis, but less so in grasses and certain other plants. Therefore the ratio of carbon-12 to the other stable isotope, carbon-13 is an indicator of the plant foods an animal (or its prey) ate.

Nitrogen stable isotopes are slightly different. Plants vary in the uptake of nitrogen-15 (compared to nitrogen-14) depending on whether they obtain their nitrogen mainly from the action of symbiotic bacteria or whether they get it directly from nitrates in the soil. Proteins are nitrogen-rich, so that most of the nitrogen content of an animal comes from the proteins in the food it eats (Ambrose 1993). This results in a trophic level effect, with each level preferentially taking up a 2-4 percent higher proportion of 15N than its food.

Previous studies of Neandertals have found that they have isotope ratios consistent with a high trophic level. This has been asserted to be evidence that Neandertals were hunters, and that they ate a very high proportion of meat compared to plant foods: upwards of 95 percent meat or more (e.g. Richards et al. 2000).

The current study builds on those previous observations by comparing Neandertals in more detail with another carnivore (hyenas) and attempting to figure out the proportions of different meats necessary to arrive at the isotope ratios in Neandertal bones.

From the discussion:

The differences observed between the amount of different prey consumed by Neanderthal and hyaena provide insights about hunting strategies of Chatelperronian Neanderthals in Saint-Césaire. Spotted hyaena is an opportunistic predator and scavenger with dietary preference for large and medium size ungulates (e.g. Cooper et al., 1999 and Silvestre et al., 2000). The isotopic signatures of spotted hyaenas in southwestern France around 36,000 years ago indicate that horse was the most abundant prey species, whereas bison, aurochs, red deer, giant deer, and reindeer were relatively abundant, woolly rhinoceros were less consumed, and mammoth were the least consumed of large herbivores. Spotted hyaenas from Belgium and Great-Britain of the same age exhibit similar isotopic signatures (Bocherens et al., 1995 and Bocherens et al., 1997), suggesting that this dietary pattern probably holds for north western Europe. Among these available prey, Neanderthals consumed much less reindeer and much more rhinoceros and mammoth than hyaenas. The low proportions of mammoth and rhinoceros in the diet of hyaena, a famed scavenger, indicates that available carcasses of these large herbivores were relatively rare in the landscape. Thus, the high proportions of these animals in the diet of Neanderthals indicate that they were obtained through another strategy than simply scavenging. Active hunting of these large herbivores by Saint-Césaire Neanderthals is thus strongly suggested by the isotopic evidence. There is some zooarchaeological evidence of proboscideans and rhinoceros hunting by Neanderthals (e.g. Scott, 1980, Auguste, 1995, Auguste et al., 1998, Locht and Patou-Mathis, 1998, Bratlund, 2000, Patou-Mathis, 2000, Patou-Mathis, 1999, Conard and Niven, 2001 and Moncel, 2001). However, mammoth and rhinoceros remains do not dominate Neanderthal faunal assemblages even if they are usually present in small numbers (e.g. Patou-Mathis, 2000) (Bocherens et al. 2005:82).

So why don't Neandertal sites have more mammoth bones in them? Most likely answer: how were they going to carry a whole mammoth bone anywhere?

This discrepancy between significant consumption of very large herbivores by Neanderthals and the remains of very large herbivores being scarce in Neanderthal sites might be due to transport decisions: filleted meat could have been transported to the occupation sites, leading to an underrepresentation of the role of large-bodied animals in Middle Paleolithic diet (Rabinovitch and Hovers, 2004) (Bocherens et al. 2005:82).

Niggling doubts

Is there any reason to doubt the result? I would like to know more about the reasons for the isotopic differences among the herbivores. The carbon isotopes are a function of the intake of different plants: in particular, plants with different carbon cycles. Browsers like elephants eat mostly plants with a three-carbon photosynthetic cycle; this results in a low carbon-13 ratio. In contrast, grazers eat a higher proportion of grasses, which have a four-carbon cycle; this leads to a higher proportion of carbon-13 in the animals' bones.

But what about the nitrogen ratios? These are generally interpreted in terms of trophic level. But what accounts for the extreme difference between the mammoths and the other herbivores? Richards et al. (2000) cite earlier work by Bocherens in suggesting that mammoths may have targeted certain plant species that had a higher 15N ratio. Maybe so.

On the other hand, what if 15N enrichment were related to longevity in some way? Isotopic ratios in teeth differ from those in bones, presumably because infants before weaning are at a "higher trophic level," consuming only the mother's milk. Teeth develop early in life, therefore their isotope levels reflect the diet composition early in life. It is interesting that the two highest 15N ratios among the herbivores are found in woolly rhinoceros and mammoth. Could these ratios be connected not to diet but to development in these large, long-developing species? And if so, could Neandertals follow a similar pattern?

Another wrench

And there is the question of fish. In the abstracts to the Third International Mammoth Conference in 2003, Bocherens and colleagues had an abstract on the subject of Neandertals eating mammoth that included this line:

However, an uncertainty remains about the possible contribution of freshwater fish, which may be similar in isotopic signature to mammoth meat.

A full text search of the JHE paper yields the word "fish" zero times. So what happened with that? Fish have relatively high nitrogen-15 levels because the aquatic food chain is generally longer than the terrestrial one, and this high 15N ratio may also characterize aquatic birds (Richards et al. 2001). Likewise, freshwater ecosystems have relatively low carbon-13 ratios, because of the availability of carbon from geological sources. Richards and colleagues (2001) focus on the combination of high 15N and low C13 as an indicator of freshwater resource use. And the Neandertals are substantially lower in 15N than Mesolithic humans who used aquatic foods extensively (Richards et al. 2001). So would it take much fish to make Neandertals look different from hyenas? Probably not.

My take

It would be nice if the isotope people would get together and settle this little problem. Richards et al. (2001) found higher 15N in Upper Paleolithic Europeans than in Neandertals; they concluded an increase in aquatic resource consumption. But one might as easily assert that later Europeans ate more mammoth -- we do, after all, find a lot more mammoth bones associated with later sites than with Neandertals. How could you tell a high proportion of mammoth from a low proportion of fish?

From these data you can't. The bottom line is that all these estimates are working on only two observations: carbon and nitrogen isotope ratios. But they are trying to estimate parameters along at least four dimensions: trophic level, aquatic vs. terrestrial, C4 vs. C3 plant consumption, and 15N plant selectivity (the explanation for the mammoth level). The computer graphic estimating the proportion of different animals in the Neandertal and hyena diets is especially silly: how can two observations estimate seven parameters simultaneously?

A little luck restricts the range of possibilities a lot: Bocherens et al. (2005) can settle on a high mammoth consumption because only the mammoths as high in 15N than Neandertals. But even this requires a severe limiting assumption: that there were no aquatic resources in the Neandertal diet.

With assumptions like that, who needs conclusions?

UPDATE: An earlier reference by Drucker and Bocherens (2004) discusses the freshwater aquatic resources; I've discussed the story in a later post.

References:

Bocherens H, Drucker DG, Billiou D, Patou-Mathis M, Vandermeersch B. 2005. Isotopic evidence for diet and subsistence pattern of the Saint-Césaire I Neanderthal: review and use of a multi-source mixing model. J Hum Evol 49:71-87.

Ambrose SH. 1993. Isotopic analysis of paleodiets: methodological and interpretive considerations. In Investigations of Ancient Human Tissue, edited by M. K. Sandford. Gordon and Breach Science Publishers, Langhorne, PA. pp. 59-130.

Richards MP, Pettitt PB, Trinkaus E, Smith FH, Paunovic M, Karavanic I. 2000. Neanderthal diet at Vindija and Neanderthal predation: the evidence from stable isotopes. Proc Nat Acad Sci USA 97:7663-7666. Full text online

Richards MP, Pettitt PB, Stiner MC, Trinkaus E. 2001. Stable isotope evidence for increasing dietary breadth in the European mid-Upper Paleolithic. Proc Nat Acad Sci USA 98:6528-6532. Full text online