Deep Pacific carbonate ion concentration and CaCO3 compensation since the last ice age

Atmospheric carbon dioxide levels were roughly 30% lower during recent glacial periods relative to interglacials. Various mechanisms have been put forth to explain glacial atmospheric CO2 drawdown, including changes in deep ocean circulation. An alteration of circulation patterns that serves to reduce upwelling of CO2-rich waters is one such model. A net transfer of carbon to the deep ocean accomplishes a speciation shift of the ΣCO2 pool whereby deep sea [CO32-] drops, seafloor CaCO3 dissolves, excess alkalinity is added to the ocean, and sea surface pCO2 is consequently lowered. While this process is stepwise over the course of glaciation, its reversal is compressed into the relatively short deglacial interval, and hence a large CaCO3 preservation spike is predicted at that time. Deep water [CO32-] reconstructions are required to evaluate this 'CaCO3 compensation' hypothesis. Benthic foraminiferal Zn/Ca and B/Ca ratios reflect the amount of dissolved metals in ambient bottom waters, but the partition coefficients of both trace metals are strongly influenced by seawater saturation state with respect to CaCO3 (ΔCO32-). Thus, glacial/interglacial changes in deep ocean carbonate ion concentrations (and consequently the link between atmospheric CO2 and CaCO3 compensation) may be inferred by applying benthic foraminiferal trace metal content as a paleo-ΔCO32- proxy. Previous work in the deep eastern tropical Pacific yielded the first Zn/Ca-based ΔCO32- record, which showed a [CO32-] spike during the last deglaciation that is consistent with the CaCO3 compensation hypothesis. Here we expand on this work by reconstructing ΔCO32- in another core from the same region, but from 1200 m shallower water depth and with better temporal resolution. Cibicidoides wuellerstorfi Zn/Ca and B/Ca are measured alongside other trace metals including Cd, Mg, Sr, Li, U, Mn, Al, and Fe at a 5 cm (1 kyr) resolution from core RC13-140 (2o52'N, 87o45'W, 2246 m). We initially focus on the first 180 cm (~35 kyr) of the core and use complementary δ18O data to investigate phasing of the deep ocean ΔCO32- signal with respect to ice volume and climate changes, in particular to test the predication of a prominent deglacial [CO32-] excursion.