Although the carbon-reservoir problem with bulk-sediment radiocarbon dates from lakes has long been recognized, many synoptic studies continue to use chronologies derived from such dates. For four sites in central North America, we evaluate chronologies based on conventional radiocarbon dates from bulk sediment versus chronologies based on accelerator mass spectrometry (AMS) radiocarbon dates from terrestrial plant macrofossils. The carbon-reservoir error varies among sites and temporally at individual sites from 0 to 8000 yr. An error of 500–2000 yr is common. This error has important implications for the resolution of precise event chronologies.
