RHYNCHONELLIFORM BRACHIOPOD EXTINCTION SELECTIVITY PATTERNS BEFORE, DURING, AND AFTER THE LATE ORDOVICIAN MASS EXTINCTION

The Late Ordovician Mass Extinction (LOME) was closely associated with the growth and decay of south polar ice sheets, but substantial uncertainties remain regarding (1) the nature and timing of Late Ordovician climate changes and (2) the ways in which these changes caused extinctions in the marine realm. Temporal patterns of extinction selectivity can provide constraints on both of these problems. We examine Late Ordovician-Early Silurian extinction selectivity patterns using a comprehensive and taxonomically standardized database of the global paleogeographic distributions of rhynchonelliform brachiopods, comprising ~6300 georeferenced regional occurrences of ~560 genera. We use machine learning and multivariate logistic regression models to evaluate the relationship between extinction risk and aspects of geographic range, latitudinal range, depth range, and other macroecological attributes through 10 time slices. We compare resulting variable importance metrics for each interval with those for the preceding interval to highlight major changes in extinction regime. Late Katian selectivity patterns differ from those of preceding and succeeding intervals primarily in the greater prominence of latitudinal range and depth range as extinction predictors. Genera with wide latitudinal ranges experienced lower extinction rates than those with comparatively narrow latitudinal ranges throughout the entire study interval but this difference was most emphatic during the Late Katian, implying that genera with limited thermal tolerance ranges were particularly hard-hit. The Late Katian interval was also characterized by highly selective extinction of deep-water genera, especially in low latitude regions (the “Foliomena fauna”). This depth gradient in extinction intensity is opposite to that which would be expected from oceanic cooling or falling sea level alone, implying that oceanographic changes were also important agents of extinction during the first pulse of the LOME. The second extinction pulse, at the end of the Hirnantian, is not well explained by any of the predictors examined, indicating that this event, which coincided with warming, deglaciation, and reflooding of cratons, was driven by fundamentally different processes.