In a more prospective component of this review, the suggested causes for AZ surface isolation are considered in light of the subarctic North Pacific (SNP), where the paleoproxies of productivity and nutrient consumption indicate similar upper ocean biogeochemical changes over glacial cycles, although with different timings at deglaciation. Among the proposed initiators of glacial AZ surface isolation, a single mechanism is sought that can explain the changes in both the AZ and the SNP. The analysis favors a weakening and/or equatorward shift in the upwelling associated with the westerly winds, occurring in both hemispheres. This view is controversial, especially for the SNP, where there is evidence of enhanced upper water column ventilation during the last ice age. We offer an interpretation that may explain key aspects of the AZ and SNP observations. In both regions, with a weakening in westerly wind-driven upwelling, nutrients may have been “mined out” of the upper water column, possibly accompanied by a poleward “slumping” of isopycnals. In the AZ, this would have encouraged declines in both the nutrient content and the formation rate of new deep water, each of which would have contributed to the lowering of atmospheric CO₂. Through several effects, the reduction in AZ upwelling may have invigorated the upwelling of deep water into the low latitude pycnocline, roughly maintaining the pycnocline’s supply of water and nutrients so as to (1) support the high productivity of the glacial SAZ and (2) balance the removal of water from the pycnocline by the formation of Glacial North Atlantic Intermediate Water. The proposed return route from the deep ocean to the surface resembles that of Broecker’s (1991) “global ocean conveyor,” but applying to the ice age as opposed to the modern ocean.