EMPTY HOUSE: IS THE LILLIPUT EFFECT IN EARLY TRIASSIC AND EARLY JURASSIC GASTROPODS A CONSEQUENCE OF REDUCED SPECIES-RICHNESS WITHIN GENERA?

We used a global database of gastropod occurrences to investigate trends in gastropod size across the end-Permian and end-Triassic extinction events. We found that the size of the largest, median, or smallest species within a genus is not significantly associated with extinction of the genus. In fact, genera with extremely large (and small) species exhibit slightly lower rates of extinction, most likely because they also tend to be species-rich. Within surviving genera, there is a non-significant tendency for the size of the smallest species to increase and the size of the largest species to decrease. Species-richness within a genus also tends to decrease between pre- and post-extinction intervals. Moreover, change in maximum size is correlated with change in species-richness and inversely correlated with maximum size in the time interval preceding extinction. Thus, reduction in maximum size (and contraction of overall size range) of gastropods across mass extinctions may be largely explained by our finding that although species-rich genera tend to survive, they are represented by fewer species in the aftermath of mass extinction and the genera with the largest species prior to extinction tend to experience the greatest reduction in maximum size. The reduction in maximum size is particularly striking to observers because body size tends to be approximately log-normally distributed; although each log unit of change across the size spectrum may be equally significant from a biological standpoint, changes at the high end are most apparent to human observers. The few surviving genera characterized by a single species before and after the mass extinctions generally exhibit size decreases, hinting at selection for smaller size within lineages in the aftermath of extinction. However, there are too few monospecific genera within our database to draw any robust conclusions on this point.