PHYLOGENETICALLY RANDOM EXTINCTIONS AND THE PRESERVATION OF PHYLOGENETIC DIVERSITY DURING THE LATE ORDOVICIAN MASS EXTINCTION

Mass extinctions remove high proportions of species and higher taxa in geologically short intervals, yet their impact on the underlying tree of life can vary. Species are unevenly distributed within the modern tree of life, with many species-poor and few species-rich clades. As such, the internal branches of the modern tree are in some cases preserved by very few species, meaning clustered extinctions can remove much more evolutionary history then predicted by species richness alone. Whether this pattern is pervasive through time, and whether the loss of evolutionary history controls the long-term impact of extinction events, has not been directly tested because of the absence of phylogenetic trees that focus on the relationships of species and genera within large clades in the geologic past.

Here, we use a preliminary tree of the evolutionary relationships of strophomenid brachiopod genera to analyze whether extinctions are clustered phylogenetically during geologic intervals ranging from the Upper Ordovician through the Silurian, including the Hirnantian interval containing the Late Ordovician mass extinction. The phylogeny currently consists of 61 species representing 31 families/subfamilies, and will eventually be expanded to include the remaining strophomenid genera. Two metrics of the phylogenetic relatedness of taxa, mean pairwise distance and mean nearest taxon distance, suggest that extinctions in the Lower Katian, Upper Katian and Hirnantian are randomly distributed across the tree, whereas extinctions in the Early Silurian are clustered. Few sister taxa go extinct in Ordovician intervals, and proportional extinction generally predicts the loss of phylogenetic diversity (the sum of the branch lengths within the tree) during this time. These results are consistent with previous analyses on the clustering of extinctions within families, suggesting an important change in the predictor of extinction risk following the mass extinction. Given this consistency, we expect the pattern to strengthen as the remaining strophomenid genera are added to the tree. Incorporating phylogenetic relationships into studies of mass extinction can reveal how the loss of evolutionary history can impact recovery and evolution downstream of major extinction events.