An increase in the ventilation of the abyssal North Pacific Ocean at the end of the last ice age

The rate at which a portion of the ocean interior exchanges gases with the atmosphere is generally described in terms of “ventilation”. Ventilation predominantly occurs where dense waters outcrop at high latitudes, pumping radioactive 14C from the atmosphere into the ocean, while simultaneously undoing the work of the ‘biological pump’—releasing the CO2 excess to the atmosphere and replenishing the oxygen shortfall resulting from the decay of organic matter in the ocean interior. As biologically sequestered carbon would be less readily released to the atmosphere if ventilation of the ocean were reduced, some explanations for the low atmospheric pCO2 of the last ice age have invoked a poorly ventilated deep glacial ocean (Toggweiler, 1999; Stephens and Keeling, 2000; Sigman and Boyle, 2000). However, paleoceanographic evidence to support this has been sparse and often ambiguous, particularly in the Pacific Ocean. In the modern North Pacific Ocean, the balance between organic matter respiration and the circulation of the ocean interior produces a broad nutrient maximum and oxygen minimum in the upper 1.5 km of the water column (Fig. 1b,c). Upward mixing of the nutrient-rich thermocline waters supplies the fertile North Pacific ecosystem with the ingredients necessary for growth, simultaneously leaking respired CO2 to the atmosphere. A lid of lowsalinity waters impedes local ventilation of deep waters at the subarctic surface, so that abyssal waters are ventilated only at the distant surfaces of the North Atlantic and Southern Ocean. The rapid attenuation of organic matter flux with depth leads to a deep sea that is better oxygenated and contains lower concentrations of remineralized nutrients and carbon, even though exchange with the atmosphere is slower, as indicated by the extremely low Δ14C (Fig. 1d). Multiproxy records from two sediment cores in the deep subarctic Pacific were developed to investigate how these patterns may have differed during the last ice age. These paleoceanographic records from a little-studied region of the global ocean provide a new perspective on the deep ocean chemistry and surface ocean fertility over the last glacial-interglacial transition (Galbraith et al., 2007). First, past values of deepwater Δ14C were estimated by separately measuring the Δ14C of co-occurring fossil benthic and planktic foraminifers. Using calibrated calendar ages of the planktic foraminifers, the benthic foraminiferal Δ14C can be decay corrected to the time of growth to give the paleo bottom water Δ14C. These values are shown in Figure 2, calculated in parts per thousand relative to the contemporary atmospheric ΔC ́cont-atm, as described in Galbraith et al. (2007). The Δ14C results clearly show that the deep North Pacific Ocean was more poorly ventilated during the Last Glacial Maximum (LGM) than it is today and that the 14C ventilation improved during deglaciation. This is similar to results from the Panama Basin at 3.2 km depth (Shackleton et al., 1988) but contrasts with measurements previously made at shallower depths in the equatorial Pacific Ocean, which show a scattered range of glacial deep ocean Δ14C that overlaps with present-day values (Broecker et al., 2004). Although it is difficult to evaluate the cause of this discrepancy, it may reflect a vertical gradient in ventilation, with ventilation similar to today in the upper 2.5 km of the water column, and relatively unventilated water below this. Intriguingly, the sole published glacial-age benthic-planktic pair from the deep South Pacific (Sikes et al., 2000) shows a more 14C-depleted value, suggesting that a lateral gradient also existed across the glacial deep Pacific Ocean at depths of 3-3.5 km, with the most poorly ventilated waters in the south. Geochemical measurements made on the sediment in which the foraminifers were buried provide additional information on the state of the glacial North Pacific. The analyses of two sedimentary components are shown in Figure 2 (see Galbraith et al., 2007 for more proxy records). The first, opal, suggests decreased diatom export during the LGM, consistent with many previously published records from the subarctic Pacific, which show reduced algal growth during the glacial (Kienast et al., 2004; Jaccard et al, 2005; Brunelle et al., 2007, and references there-