ONE PULSE OR TWO? COMBINING SIMULATIONS OF SEQUENCE-STRATIGRAPHIC ARCHITECTURE AND DIVERSIFICATION PROCESSES TO MODEL THE EXPRESSION OF THE LATE ORDOVICIAN MASS EXTINCTION

The Late Ordovician mass extinction (LOME) is commonly expressed as two clusters of last occurrences in the fossil record: a first cluster appearing to coincide with the intensification of Gondwanan glaciation near the Katian-Hirnantian boundary and a second appearing to coincide with partial deglaciation in the mid-Hirnantian. Consequently, glacioeustatic sea-level fluctuations have been proposed as a principal driver of the LOME. However, sequence stratigraphic architecture exerts a strong control on the pattern of first and last occurrences in the fossil record. A single pulse of extinction may be expressed as two or more clusters of last occurrences if major stratigraphic surfaces (e.g., surfaces of forced regression, flooding surfaces) occur during or shortly before the extinction episode. Thus, it is possible that the apparently double-pulsed nature of the LOME is the stratigraphic expression of a single prolonged (10⁴–10⁵ years) extinction pulse.

We use a modelling approach to test whether stratigraphic architecture is responsible for the two clusters of last occurrences during the LOME. Using the sedimentary basin model Sedflux 2.1, we simulate five alternative Late Ordovician sea-level histories in a hypothetical ice-distal basin. We combine these with a branching model of evolution and extinction, as well as ecological gradients, to simulate patterns of last occurrences for each scenario. From this, we can determine how likely a single extinction event would be expressed as two pulses in the rock record.

Our models demonstrate that a single protracted pulse of extinction can generate the two clusters of last occurrences seen in the Hirnantian. Moreover, these clusters are synchronous with glaciation and deglaciation regardless of the timing of the extinction during the glacial cycle. Of the single-pulse scenarios analyzed, a pulse of extinction during deglaciation most consistently produces two clusters of last occurrences. Future work will compare these simulations to patterns of last occurrences from the empirical Hirnantian fossil record to produce the most likely pattern of extinction for the LOME, allowing us to better interpret the drivers and selective signature of the extinction.