Finding the identity of animal (and plant) fats

Last week I made a note about some ongoing work at the Spanish site of El Salt, which suggested taxonomic identifications for burned traces of animal and plant fats.

I was wondering how exactly that kind of identification is done. I don’t know any details in the El Salt example, but I was able to find some recent work from Neolithic contexts that makes a similar kind of identification.

For example, Oliver Craig and colleagues (2005) tested potsherds for fatty acid residues, and then subjected those residues to isotopic analysis. The isotopic composition of different weight fatty acids (C18:0 and C16:0) may have different carbon-13 fractions from each other, a relation that varies among different animal taxa. So basically, you fraction out the 18-carbon and 16-carbon fatty acids and measure the ratio of carbon-13 to carbon-12 in the two sample components.

Craig and colleagues were able to show that milk fatty acids have a distinct ratio of carbon-13 fractions compared to body fat (adipose tissue) from the same taxa, basically milk has a lower carbon-13 fraction in the heavier 18-carbon fatty acids. They found most of their sampled potsherds to have a similar ratio, and interpreted that as evidence for the use of milk products in Neolithic central and eastern Europe.

Last year, Evershed and colleagues (2008) came to a similar result, applied to potsherds from early Neolithic sites in the Near East. Evershed published a review article on organic trace analysis in archaeology last year, from which I’ve drawn this helpful figure:

Figure 2 from Evershed 2008. Original caption: Simple saturated C16:0 and C18:0 fatty acids generated via hydrolysis of triacylglycerols (LHS) during processing and/or burial of fats (and oils), which on their own have limited diagnostic value as biomarkers. However, the plot (RHS) of the ?13C values for these fatty acids shows how the fats of the major Old World domesticated animals can be separated due to differences in the their metabolic and biosynthetic origins. The ellipses are confidence ranges (P = 0.684) and the theoretical mixing ranges. Such plots provide the basis for determining the origins of animal fat residues (adapted from Mukherjee et al. 2005).

The references I’ve found distinguish fats by mammal taxa only by contrasting pig from ruminant, so I tend to interpret the references to “deer and goat” in the El Salt press report to the fact that they’re the resident ruminants. Of course a finer statistical segregation based on more comparative sampling is also possible. Also, Evershed’s review goes into some forensic contexts, and shows that human adipose tissue has its own distinctive signature. In theory that would make it possible to find evidence of cannibalism from prehistoric contexts.


Craig OE, Chapman J, Heron C, Willis LH, Bartosiewicz L, Talor G, Whittle A, Collins M. 2005. Did the first farmers of central and eastern Europe produce dairy foods? Antiquity 79:882-894.

Evershed RP. 2008. Organic residue analysis in archaeology: the archaeological biomarker revolution. Archaeometry 50:895-924. doi:10.1111/j.1475-4754.2008.00446.x

Evershed RP and 21 others. 2008. Earliest date for milk use in the Near East and southeastern Europe linked to cattle herding. Nature 455:528-531. doi:10.1038/nature07180