The middle Cretaceous oceans were characterized by widespread anoxia, high sea levels, reef/carbonate-bank extinctions, and, at least in the Pacific, by intense widespread mid-plate volcanism. The demise of reefs on numerous Pacific atolls has left a large number of guyots. This synthesis explores the possibility that the middle Cretaceous demise of mid-Pacific and other carbonate-bank communities, and related extinctions elsewhere, reflects sporadic, possibly volcanogenic, upwelling of anoxic water. The upwelling would have had to be rapid enough to overwhelm surface reoxygenation, and frequent/widespread enough to suppress recolonization. According to the Hallock-Schlager hypothesis, upwelling of nutrient-rich waters destroyed the carbonate-bank communities; nutrient excess and anoxia may well have worked together.
The globally widespread presence of at least mid-depth anoxia (±H2S?) is reasonably well established (the Cenomanian/Turonian "Oceanic Anoxic Event" perhaps was an extreme example), as is the mortality of modern benthic invertebrates at O2 levels below about 0.15 ml/l. Recolonization times for modern reef corals imply that upwelling catastrophes, if localized, need have recurred on approximately decade time scales. Previously postulated Cretaceous sea-level fluctuations, raising sea level at rates on the order of 5-10 m in 104-105 yr, would then have sufficed to carry the intermittently stressed reef surfaces below the ∼5- to 10-m-deep high-carbonate-production zone, below which reef recovery chances diminish. Thus, the episode of frequent (localized but widespread) upwelling events need have lasted only about 104-105 yr. Although intermittent deep-water ventilation is known to have occurred on Milankovitch time scales during the Cretaceous, thereby bringing anoxic and/or nutrient-rich waters to the surface, geologic (and ocean modeling) evidence is necessary to establish whether a significant amount of rapid upwelling did (or could) occur.
Because mid-plate submarine volcanism was widespread and intense in the middle Cretaceous Pacific, volcanogenic/hydrothermal heating and entrainment of nutrient-rich anoxic waters is one possible mechanism for catastrophic buoyant upwelling, rudistid bank extinction, and guyot formation. The efficacy of this process depends principally on the buoyancy (that is, heat energy) flux and the buoyancy (Brunt-Väisälä) frequency. Typical hydrothermal plumes along the Mid-Oceanic Ridge axis rise only to 150- to 400-m heights above the sea floor, but "megaplumes" rising 1,000 m are known.
Simple plume models imply that only basalt eruptions comparable to the largest historically known would be sufficiently energetic (1012-1014 J/s) to bring bottom water to the ocean surface from normal Mid-Oceanic Ridge axial depths (3 km) over much of the present ocean; however, these models are conservative through neglect of momentum overshoot and gas exsolution. To the extent that higher eruption rates and lower Brunt-Väisälä frequencies (due to lower vertical temperature gradients) prevailed in the middle Cretaceous Pacific, anoxic deep waters would more readily have reached the ocean surface. Most Cretaceous guyots are now in the North Pacific but were carried across the equator by Pacific plate motion; normal tendencies to upwelling and oxygen deficiency along the equator would have made the carbonate banks most vulnerable during their equatorial transit.