AN OCEANOGRAPHIC MECHANISM FOR CARBON CYCLE CHANGES AND GLACIATION IN THE LATE ORDOVICIAN

Major glaciation in the late Ordovician was accompanied by two phases of extinction and marked by anomamlously large positive shifts in d¹³C and d¹⁸O in marine carbonates. New detailed stratigraphic data confirm that the entire glacial period was short-lived; spanning no more than 1.5 graptolite zones in the Hirnantian. Oxygen isotope data from brachiopods and ostracod calcite suggest significant cooling of perhaps 6 to 8 ^oC at low palaeolatitudes.

We examine the implied change in carbon cycling using simplified box models and an Ocean General Circulation Model (OGCM). The oceanic carbon cycle is examined for overturning circulations related to the non-glacial and glacial periods paying particular attention to changes in ocean d¹³C signatures and atmospheric CO₂ levels. Limiting cases are explored for a circulation (i) with dense water formation on shallow shelves and (ii) with no shallow shelves. The implications for nutrient supply and primary productivity are explored.

For non-glacial conditions in the Palaeozoic, the circulation is often assumed to be shallow with deep, global anoxia. However, this circulation is inconsistent with the observed long term carbon burial, due to exhaustion of nutrients in the surface ocean. Instead, we propose a deep thermal mode controlled by densification on shallow shelves with anoxia confined to isolated regions.

For the glacial period (Hirnantian), shifts in oceanic d¹³C and atmospheric pCO₂ consistent with the geologic record are obtained by having a shallower circulation, without the need for enhanced biological productivity or changes in nutrient concentrations.