Introduction
The Early Paleozoic, notably the Ordovician Period, hosts one of the largest marine biodiversification events in Earth history, the Great Ordovician Biodiversification Event (GOBE). Not long after this proliferation of marine fauna, however, the second-largest mass extinction in Earth history occurred, the Late Ordovician Mass Extinction Event (LOME; (Harper et al., 2014)). The LOME resulted in the loss of ~85% of marine species during two distinct extinction pulses, with the first occurring at the Katian-Hirnantian boundary, and the second in the late Hirnantian (Brenchley et al., 2001; Harper et al., 2014). Traditionally, the first LOME pulse has been associated with rapid global cooling and widespread glaciation that resulted in major eustatic sea-level fall, creating widespread marine habitat loss and ecologic shifts (Harper et al., 2014). The second LOME pulse has been associated with sea-level rise and an expansion of anoxic (potentially euxinic, anoxic, and sulfidic water column) conditions (Hammarlund et al., 2012). However, recent studies have invoked widespread anoxia/euxinia for both LOME pulses, indicating that the redox conditions surrounding this event might be more complex than initially understood (Zou et al., 2018). Associated with the LOME was a major perturbation in the global carbon cycle, recorded as a positive excursion in the marine carbon isotope record known as the Hirnantian carbon isotope excursion (HICE) (Brenchley et al., 2003).
The primary causal mechanisms for the HICE has previously been attributed to changes in carbonate weathering regimes during eustatic sea-level fall (Kump et al., 1999) and/or enhanced burial of organic matter associated with increased preservation resulting from decreased marine oxygenation (Brenchley et al., 2003; Hammarlund et al., 2012; Jones and Fike, 2013). Traditional sulfur (S) isotope approaches have investigated the dynamics and extent of euxinic marine conditions during the LOME. Pyrite sulfur isotope (δ34Spyr) profiles across multiple globally distributed paleobasins show nearly synchronous positive excursions, suggesting a global perturbation, albeit with local overprints to explain the variable magnitudes ~15-40‰. Overall, the positive δ34Spyr shift may reflect increased pyrite burial under widespread sulfidic conditions (Hammarlund et al., 2012; Jones and Fike, 2013). However, the single reported seawater sulfate (δ34SCAS) record shows little variation in this global redox proxy throughout the HICE, suggesting a minimal change in global pyrite burial rates over this interval and thus a limited global expansion of euxinic conditions (Jones and Fike, 2013).
Additionally, non-traditional paleoredox proxies have been applied to assess the extent of widespread reducing conditions as a potential kill mechanism for both LOME pulses. Specifically, uranium and molybdenum (δ238U and δ98Mo) stable isotope records have been interpreted to reflect global to regional changes in marine redox, and have identified possible shifts towards more reducing conditions—but at differing times during the Hirnantian (Zhou et al., 2015; Bartlett et al., 2018). Models based on δ238U data from eastern Laurentia suggest up to 15% of the total seafloor area experienced anoxic conditions just before the end of the Ordovician (Bartlett et al., 2018). Meanwhile, δ98Mo data from South China have been interpreted to record local shifts from suboxic to euxinic conditions prior to and during the early Hirnantian (Zhou et al., 2015). However, each of these marine redox interpretations are based on singular datasets within widely different depositional environments. Moreover, delineating the timing of these inferred changes in redox relative to one another remains problematic, as the lack of a universally accepted, fully integrated Upper Ordovician biostratigraphic scheme limits the resolution of correlations (see supplemental information). Additionally, these two paleoredox proxies have different specific responses due to their position on the redox ladder, i.e. changes in increasingly reducing marine conditions, which creates additional complications for understanding the onset of non-sulfidic anoxia (δ238U) versus euxinia (δ98Mo).
To better elucidate critical gaps in our understanding of the mechanistic underpinnings for the LOME, we present new I/(Ca+Mg) ratios and δ34SCAS datasets from three widely distributed Upper Ordovician carbonate successions. Our study localities were deposited in different paleocean basins and record the HICE along with major fluctuations in eustatic sea level (Finney et al., 1997; Young et al., 2010; Ghienne et al., 2014; Kiipli and Kiipli, 2020). These new-paired geochemical datasets provide a more comprehensive understanding and specificity of marine redox conditions (local and global) and climate that led to the second-largest extinction event in Earth history.
Background