MARINE INVERTEBRATE EXTINCTION SELECTIVITY AT THE CRETACEOUS/PALEOGENE MASS EXTINCTION

Numerous studies have demonstrated that extinction selectivity occurred during background times and at mass extinction events, usually by testing for correlation of taxon traits (ecological, anatomical, geographic, etc.) with extinction rate or survivorship. Certain studies have shown that traits which reduced extinction probability during background times did not operate at the Cretaceous/Paleogene (K/Pg) mass extinction event (e.g., Jablonski, 1986), leading some to conclude that mass extinction events exhibit less ecological selectivity than background extinction times (e,g., Brusatte et al., 2008). However, trait inference for fossil taxa can be difficult or even impossible for some traits, and the absence of significant correlations does not suggest absence of selectivity as the trait causing selectivity may not have been observed. Using a method that is agnostic about taxon traits, I have demonstrated that during both background and mass extinction times, extinction of marine invertebrate genera has been nonrandom with respect to species richness categories of genera. I have also shown that extinction selectivity at the species level cannot always be extrapolated from selectivity of genera; the appearance of random genus extinction with respect to species richness results when species have experienced high selectivity. This result means that at times of high species selectivity, species richness of genera will appear to have not buffered against extinction probabilities of genera. Via this method the K/Pg mass extinction appears to be a time of very high species extinction selectivity. Such selectivity could be a result of regional extinction rate heterogeneity, taxon specific extinction rate heterogeneity, or clustering of extinctions in genera consisting of species which possess extinction-biasing traits (“traditional” or ecological selectivity). As it presently stands, there is compelling evidence that the K/Pg extinction event was highly selective, and thus any proposed explanation for the K/Pg mass extinction must also explain the high extinction selectivity.