Abstract
Predator-prey interactions have been a central theme in population ecology for the past century, but real-world data sets only exist for recent, relatively short (<100 years) time spans. This limits our ability to study centennial/millennial-scale predator-prey dynamics. We propose that regional radiocarbon databases can be used to reconstruct a signal of predator-prey population dynamics in deep time, overcoming this limitation. We support our argument with examples from Pleistocene Beringia and the Holocene Judean Desert.