Enhanced δ¹³C and δ¹⁸O Differences Between the South Atlantic and South Pacific During the Last Glaciation: The Deep Gateway Hypothesis

Abstract

Enhanced vertical gradients in benthic foraminiferal δ13C and δ18O in the Atlantic and Pacific during the last glaciation have revealed that ocean overturning circulation was characterized by shoaling of North Atlantic sourced interior waters; nonetheless, our understanding of the specific mechanisms driving these glacial isotope patterns remains incomplete. Here we compare high‐resolution depth transects of Cibicidoides spp. δ13C and δ18O from the Southwest Pacific and the Southwest Atlantic to examine relative changes in northern and southern sourced deep waters during the Last Glacial Maximum (LGM) and deglaciation. During the LGM, our transects show that water mass properties and boundaries in the South Atlantic and Pacific were different from one another. The Atlantic between ~1.0 and 2.5 km was more than 1‰ enriched in δ13C relative to the Pacific and remained more enriched through the deglaciation. During the LGM, Atlantic δ18O was ~0.5‰ more enriched than the Pacific, particularly below 2.5 km. This compositional difference between the deep portions of the basins implies independent deep water sources during the glaciation. We attribute these changes to a “deep gateway” effect whereby northern sourced waters shallower than the Drake Passage sill were unable to flow southward into the Southern Ocean because a net meridional geostrophic transport cannot be supported in the absence of a net east‐west circumpolar pressure gradient above the sill depth. We surmise that through the LGM and early deglaciation, shoaled northern sourced waters were unable to escape the Atlantic and contribute to deep water formation in the Southern Ocean.

Data Constraints on Glacial Atlantic Water Mass Geometry and Properties

“…Whether or not glacial NADW had lower preformed phosphate than estimated here, its preformed values were likely still lower than the Pacific, so if there was a greater fraction of NADW in the global ocean, it would have contributed to a more efficient soft tissue pump and to lower atmospheric CO 2 (Kwon et al, ). Alternatively, if SOW filling the global deep ocean had lower preformed phosphate than estimated for the Atlantic here, perhaps by a SOW mass that was distinct from that ventilating the deep Atlantic (e.g., Sikes et al, ), it could have contributed to a stronger biologic pump. If AAIW filling the other basins also had lower preformed phosphate than modern, like Atlantic AAIW, it may have also played a role.…”

A Southwest Pacific Perspective on Long‐Term Global Trends in Pliocene‐Pleistocene Stable Isotope Records

“…These findings are supported by South Atlantic δ 13 C values being equal to, or lower than, southwest Pacific values at~2.6 Ma to recent, despite enhanced NCW/NADW production during more recent intervals during the Pleistocene (Bell et al, 2015), thus, eluding to the initiation of the circulation system closely resembling the modern. Cooling of AAIW and the establishment of modern δ 13 C gradients between the deep southwest Pacific and South Atlantic coincide with the period of global cooling surrounding the time period associated with higher amplitude glacial-interglacial cycles at~1.5-1.0 Ma (Lisiecki & Raymo, 2005) and may reflect the more extreme intrabasinal isolation due to lower sea level (e.g., Lisiecki & Raymo, 2005) and a shallow sill depth of the Drake Passage (e.g., Sikes et al, 2017;Figure 5). Warming of AAIW and a marked increase in subantarctic South Pacific δ 18 O values demonstrates the complexity associated with the interval of time between~2.6 and 1.5 Ma, which possibly relates to varying poleward heat transport from the subtropics but also southern high-latitude cooling and expansion of continental ice sheets (McClymont et al, 2016).…”

“…Finally, persistently higher δ 13 C in the southwest Pacific compared to the South Atlantic after~2.6 Ma is consistent with Pleistocene glacial-interglacial gradients and the reflection of regional sources of AABW (Weddell Sea versus Ross Sea sector; e.g., Bostock et al, 2013;McCave et al, 2008;Ninnemann & Charles, 2002;Figure 4d). In so that, NADW was not reaching the deep Pacific either because the sill depth of the Drake Passage was shallower or that NADW was flowing at too shallow of depth to pass from the Atlantic to the Pacific (Sikes et al, 2017). Cooling of AAIW and the establishment of modern δ 13 C gradients between the deep southwest Pacific and South Atlantic coincide with the period of global cooling surrounding the time period associated with higher amplitude glacial-interglacial cycles at~1.5-1.0 Ma (Lisiecki & Raymo, 2005) and may reflect the more extreme intrabasinal isolation due to lower sea level (e.g., Lisiecki & Raymo, 2005) and a shallow sill depth of the Drake Passage (e.g., Sikes et al, 2017;Figure 5).…”

Deglacial carbon cycle changes observed in a compilation of 127 benthic <i>δ</i><sup>13</sup>C time series (20–6 ka)

“…Processes that potentially explain atmospheric CO 2 change during glacial cycles include the efficiency of the biological pump (Martínez-Garcia et al, 2009;Galbraith and Jaccard, 2015), circulation changes (Ferrari et al, 2014;Schmittner and Lund, 2015;Lacerra et al, 2017;Menviel et al, 2017;Sikes et al, 2017;Wagner and Hendy, 2017), or a combination of multiple processes (Bauska et al, 2016;Skinner et al, 2017). Different processes could influence the carbon cycle on different timescales (Bauska et al, 2016;Kohfeld and Chase, 2017) and/or in different regions (e.g., Gu et al, 2017) and complicate interpretations of which processes are most responsible for atmospheric CO 2 change.…”