LAZARUS TAXA IN MASS EXTINCTION RECOVERY INTERVALS: INDICATORS OF RECORD BIAS OR MASS RARITY?

The recovery intervals following some mass extinction events are characterized by an abundance of “Lazarus taxa” that disappear from the fossil record but subsequently reappear. Two major hypotheses have been put forth to explain this pattern. The first posits that Lazarus taxa are indicators of record failure -if the quality of the fossil record decreases from one interval to the next, the number of taxa that go unsampled should increase. The second hypothesis proposes that Lazarus taxa are instead indicative of episodes of “mass rarity” during which prolonged environmental crises reduce the populations, and hence the preservation probabilities, of many clades. Here we use the Paleobiology Database and the Macrostrat Database to evaluate relative support for these two competing hypotheses. We first measure change in the quality of the fossil record by comparing the number of Paleobiology Database marine fossil collections and Macrostrat Database marine sedimentary packages, as well as changes in major lithology (e.g. carbonate vs. siliciclastic) and the geographic locus of sampling, between stages. We then calibrate the relationship between changes in the quality of the fossil record and changes in the proportion of range-through genera (defined as the number of genera that are sampled in the previous interval and in subsequent intervals but not in the interval in question divided by the total number of genera crossing the lower boundary of the interval) for both background and mass extinction recovery intervals. If the “mass rarity” hypothesis holds true and detection probabilities in post-extinction recovery stages are lower than expected given changes in record quality, these stages should stand out as positive outliers. As expected, we find a strong inverse relationship between changes in fossil record quality and changes in proportion of range-through genera during background intervals. We further find that although the recovery from the End-Ordovician mass extinction appears to best support the “mass rarity” hypothesis, the majority of other intervals of presumed biotic crisis better support the fossil record bias hypothesis.