Reply to Collins and Copeland: Spontaneous gelatinization not supported by evidence

We thank Collins and Copeland (1) for raising the point that a better understanding of starch granule behavior in archaeological settings would be useful. However, we disagree with their assessment that the cooked starch granules we observed on Neanderthal teeth (2) were an effect of diagenesis.

To estimate the effect of aging on starch granules, they apply an equation that is used to describe thermal gelatinization of a starch/water system, and we believe their application is not appropriate for predictions about archaeological specimens. Their activation energy estimate derives from experiments in which starches are heated to gelatinization temperatures (≈50–80 °C), far outside the range expected for archaeological cave sediments. They propose that their calculation is accurate at lower temperatures (≈10–20 °C), but the table they refer to shows that the activation energy for gelatinization increases as the temperature decreases, meaning that much longer periods of time than they estimated would be required for low-temperature gelatinization to occur, if it occurs at all. The activation energies for low-temperature gelatinization in the absence of pressure or catalysts are not published in their reference material and are, as far as we can tell, yet unknown. In addition, because gelatinization can only occur in systems with an excess of water (3), this equation will likely not describe the behavior of individual starch granules preserved in relatively dry archaeological contexts, like highly mineralized dental calculus.

Even in the case that this thermal gelatinization process could prove applicable in some archaeological specimens, we are confident that cooked granules from Shanidar do not represent diagenetic gelatinization for three reasons. First, Henry (4) recovered starches from 44 teeth from 14 sites in Europe, the Near East, and Africa that range between ≈120,000 and 10,000 y old. There was no correlation between number of recovered native (unaltered) or gelatinized starches and sample age or temperature. Second, Henry also examined plant microfossils preserved on 57 tools from Shanidar Cave. Twenty-one of the 119 recovered starch granules were from Triticeae, but no gelatinized starches of any type were found, despite these tools being the same age or older than the dental calculus sample. Similarly, at Neotropical sites where temperatures are much higher than at Spy or Shanidar, gelatinized starches were recovered from dental calculus and not from stone tools (5, 6). Finally, the examination of starch granules by light microscopy, as opposed to chemically invasive methods like enzymatic digestion, allows one to differentiate among different kinds of damage. We did note that several of the starch granules from Shanidar and Spy showed damage that may be related to aging, but this was in the form of cracking or breaking, consistent with dry, low-temperature damage, rather than the swelling, buckling, and loss of birefringence that are the result of heating in water. These starches were included in the “damaged/encrusted” count (2). For these three reasons, we discard the possibility of spontaneous gelatinization over time as unsupported by the available evidence. This leaves cooking as the best explanation for the gelatinization of the Shanidar starch granules.