STATISTICAL EVIDENCE FOR SELECTIVITY IN THE PLANKTIC FORAMINIFERA

Planktic Foraminifera have previously been shown as a premiere example of selective extinctions. This is especially true for the K/Pg, when they go from a large and diverse group to a few small, compressed globigerine forms. To test the significance of this pattern, we utilize a 24 variable morphometric database, comprehensive at the species level. These measurements include a combination of ratios, size metrics, angles, and the number of chambers in the final whorl. By performing a Principle Component Analysis (PCA), the repeated pattern of small compressed forms evolving to a more complex, diverse, large assemblage is clearly illustrated. To investigate selectivity, the mean and variance of each morphometric variable were calculated separately, and binned into roughly half a planktic foraminiferal zone. In an extinction, directional selection is defined as a significant change in the mean morphology from one bin to the next. Similarly, stabilizing selection is a decrease in the variance of the morphometric distribution, as the extreme morphologies are eliminated. This appears in the morphometric data as ‘shifts’ in the mean and the variance during extinction events. These ‘shifts’ are expressed by the first differences of the mean and variance. By holding the species’ originations and extinctions to the real values and bootstrapping the morphometric values, the authors can develop stochastically derived curves of the mean and variance, which are sensitive to the changing species diversity. The measured data is then examined for significant deviations from this null model. At the K/Pg, there is directional selection in the total area of the test and the length-to-width ratio, as well as stabilizing selection in the apical angle. While this analysis is focused on the hallmark extinction, the K/Pg, there are significant events during other intervals. The PCA was also interrogated for events, but through the creation of the composite axis, several disparate changes in morphology were combined and the signal of selectivity was largely lost. This highlights the need to study selectivity in individual variables, not as composites. This statistical approach highlights specific unique events which drive significant selection in the Planktics, and allows for a more focused investigation into oceanographic causes.