Cooling after Warming: pCO2 Undershoot Forced By Organic Carbon Burial Following the End-Triassic Mass Extinction

Abstract:

In the lowermost Jurassic, following the end-Triassic mass extinction, coupled positive δ³⁴S and δ¹³C excursions indicate elevated rates of organic carbon burial. These coupled excursions have been most widely interpreted to result from ocean anoxia driven by warming induced via CO₂ effusion from the Central Atlantic Magmatic Province (CAMP). Yet, recent stratigraphic correlations suggest the positive excursions corresponded with an interval – identified by paleosol proxy data – during which pCO₂ dropped below pre-extinction baseline levels following the cessation of CAMP activity. If so, the positive δ¹³C excursion could not have been directly related to CAMP induced warming, nor the pCO₂ undershoot driven by enhanced drawdown of carbon by weathering of fresh CAMP basalts. The enhanced weathering of silicate rock can cause a reduction in pCO₂, but it cannot increase the ratio of C_org to C_carb burial – a prerequisite for an increase in δ¹³C value of the ocean. Here we use a carbon cycle model to test whether the establishment of anoxic marine conditions can account for both the positive δ¹³C excursion and pCO₂ undershoot indicated by the paleosol proxy record. Anoxia leads to high rates of organic carbon burial because it is sustained by enhanced remineralization of phosphate within sulfidic bottom waters. Enhanced return of phosphate to the marine water column fuels further productivity, resulting in the stabilization of anoxia even in the face of rising pO₂ and a declining supply of phosphate from weathering. We parameterize this behavior in the model by imparting hysteresis to the C:P burial ratio of sediments such that it remains high even as pCO₂ declines past baseline levels, resulting in a pCO₂ undershoot accompanying the positive carbon isotope excursion. This interpretation of the data suggests that the negative feedbacks that respond to warming may overcompensate, resulting not in a smooth exponential decay, as the recovery is often modeled, but rather in a non-linear response.