Subterranean transport and deposition of quartz by ants in sandy sites relevant to age overestimation in optical luminescence dating

An artificial layered sandy site was created using a combination of native sand and colored sand (3 parts native uncolored quartz, 1 part colored quartz) in Apalachicola National Forest near Tallahassee, Florida. Twelve layers of sand, each 1 by 1 m in horizontal extent by 10 cm in thickness were emplaced to a depth of 2 m followed by implantation at the surface of a Florida harvester ant (Pogonomymrex badius) colony (the lower two layers were 50 cm thick). The colony excavated a nest, and after 7 months, the sand layers were excavated to the base to test the hypothesis that sand grains were moved upward within the ant nest without reaching the surface. The ants penetrated 11 of the 12 colored layers reaching a depth of 130 cm. Thirty nine sticky-acetate peels of ant chamber floors were collected and colored sand grains were counted under a microscope. More than 16,000 grains were identified in layers that did not originally host them. Of these, more than 80% were unambiguously moved upward. This means that possibly as many as 54,000 upwardly mobile grains were present (ratio of 3:1 uncolored to colored). In relation to optical luminescence (OSL) dating, this means that grains that would not have been optically zeroed by transport to the surface (defined here as subterranean-transported) were present in abundance, and that if the site was ancient, there would have been found many grains that were older than the layers they presently reside in, even if only one colony of harvester ants had disturbed the layers. This is in addition to the fact that backfilling of chambers and tunnels may contribute even more significantly to the presence of a subterranean-transported component of an OSL sample. We conclude that ants can significantly affect the age distributions in sandy archaeological sites. Multiple examples of such disturbances have been documented in the literature. Most relevant to our results are recent studies of the OSL chronology of Pre-Clovis-age and Palaeoindian age archaeological sites in sandy environments in North America that may have been compromised by ant bioturbation of quartz sand grains. Here we have examined in detail the potential effects of one episode of ant nest-building on the age overestimation of affected sediments. From this we found that as few as 12 episodes of bioturbation involving backfilling of chambers in the same volume of sand could lead to the presence of 1 contaminant grain per 50 grains of sample.