LEGACY RADIOCARBON AGES AND THE MIS 3 DATING GAME: A CAUTIONARY TALE FROM RE-DATING OF PRE-LGM SITES IN WESTERN CANADA

Meta-analyses that include compilations of thousands of radiocarbon dates are becoming more abundant in the contemporary geoscience literature. However, there can be pitfalls with the uncritical use of dates near the temporal limits of radiocarbon dating, and even minute amounts of “younger” carbon can cause a sample that is actually beyond the limit of radiocarbon dating to give a misleading finite age. We present a cautionary note on the use of pre-LGM radiocarbon dates (both conventional and AMS) compiled from earlier literature, based on case studies at four purportedly MIS 3 sites in both glaciated and unglaciated western Canada. In unglaciated central Yukon Territory, a spruce stump previously dated to ~26 ¹⁴C ka BP yielded non-finite and ~50 ¹⁴C ka BP dates by AMS (n=3), with implications for hypotheses on full-glacial boreal forest refugia in eastern Beringia. In southeast Yukon, sub-till plant macrofossils originally dated to ~24 ¹⁴C ka BP were re-dated to ~45 ¹⁴C ka BP (n=5) and thus no longer provide reliable constraints for the timing of Cordilleran ice-sheet advance near its northeast limit. Two sites in Alberta, with previously published MIS 3 radiocarbon dates on sub-till organics, similarly returned AMS ages (n=6) that were either non-finite or almost indistinguishable from the background standard. We were not able to replicate the MIS 3 radiocarbon dates that constrain the ages of any of these four purportedly interstadial sites. Our results suggest that substantial caution should be exercised when relying on MIS 3 radiocarbon dates compiled from the literature, including careful manual vetting of published ages; consideration of the stratigraphic and taphonomic context; and judicious re-dating using stringent sample preparation, modern AMS techniques, and suitable secondary standards.