No evidence for sample contamination or diet offset for pre-Columbian chicken dates from El Arenal

In their recent paper on ancient DNA from Pacific chickens, Thomson et al. (1) claim in their supporting information that there were “issues with previous radiocarbon dates” published by Storey et al. (2). Factual errors in Thomson et al.’s (1) discussion of the radiocarbon dating must be rectified. I demonstrate that there are no “issues” as claimed by the authors for the chicken bone dates for El Arenal, Chile, which produced reliable pre-Columbian dates at 2 SD, and indicate a pre-Columbian introduction of chickens to South America.

In their supporting information, Thomson et al. (1) infer sample contamination because preparations “…were performed without ultrafiltration of high molecular weight collagen, or the removal of exogenous organic matter through XAD-2 purification.” In fact, ultrafiltration is decidedly not considered to be the standard for collagen preparation, and studies challenge its efficacy for protein purification, as well as the ultrafilter’s own potential for contamination of the protein sample (3). The XAD-2 purification method referred to has rarely been used for bone protein since the late 1990s.

Thomson et al.’s (1) statement that the oldest sample (NZA26115) was very small and characterized by a low collagen yield is misleading. NZA26115 was a small sample of bone, approximately 45 mg, which was all that remained after mitochondrial DNA analyses. NZA26115 produced a normal collagen yield, yet even at quite normal percent yield per weight of sample, produced just enough collagen for ¹⁴C dating and a ¹³C analysis, which at −20.9‰ is not indicative of marine diet offsets (4).

Thomson et al.’s (1) claim that “Previous work has shown that neither C:N ratios, nor δ¹³C:δ¹⁵N ratios are indicators of reliable radiocarbon dates” is incorrect. Since the mid-1980s, C:N ratios and δ values have been the standard by which the world’s most well-regarded radiocarbon laboratories judge the preservation of a protein sample and its likelihood of a reliable radiocarbon date (5).

Thomson et al. (1) then suggest that dietary marine carbon in Chilean chicken bones produced erroneously older dates, and recommend the use of ¹³C measurements on individual amino acids. This process would no more clarify the fact that El Arenal chickens did not have marine diet components than do our published replies to previous claims of an unidentified ¹⁴C-depleted diet source affecting El Arenal chicken dates (4). The δ¹³C range over the three dated chicken bones (−20.9‰ to −19.45‰) is well within the expected range for wholly terrestrial diets. The 1.4‰ difference over the three bones could actually be an indication of varying amounts of C4 plants in chicken diets, such as from δ¹³C-enriched maize, a diet source found in excavations at El Arenal. The δ¹³C and δ¹⁵N data for El Arenal chickens clearly indicate a terrestrial diet with low protein, but especially, the terrestrial range value of δ³⁴S (2.16‰) unequivocally establishes that there is no evidence for a marine component in the diet. Thomson et al.’s (1) claims misrepresent potentials for sample contamination or diet offsets affecting ¹⁴C results for El Arenal chickens reliably dated to the pre-Columbian period and consistent with the archaeological content and other dates obtained from the site (4).