We present a new deep‐water carbonate ion concentration ([CO32–]) record, reconstructed from the “size‐normalized weight” (SNW) of the planktonic foraminifer Neogloboquadrina dutertrei in core MD06‐3047B, representing a mid‐depth site (2.5 km) in the western tropical Pacific since 700 ka. On glacial‐interglacial timescales, deep‐water [CO32–] exhibits an inverse relationship with global sea‐level elevations, consistent with the “coral reef hypothesis” that the deep Pacific carbonate system responded to variations in shelf‐carbonate production through the past 700 kyr. On longer timescales, a decoupling between deep‐water [CO32–] and δ13C around the globe can be explained by a combination of continental weathering and nutrient inputs. During the mid‐Brunhes interval (~600‐200 ka), [CO32–] reached a maximum of ~100 μmol kg–1 at the marine isotope stage (MIS) 12/11 boundary, followed by a steep decrease to a minimum of ~40 μmol kg–1 during mid‐MIS 11, representing the largest‐amplitude change in [CO32–] over the past 700 kyr. The [CO32–] maximum records the largest deglacial oceanic carbon release since 700 ka, and the [CO32–] minimum was a response to a global increase in pelagic carbonate production. From MIS 3 to 2 and from early to mid‐MIS 13, [CO32–] showed rising trends opposite to those at water depths greater than 3.4 km, implying enhanced Pacific stratification during these intervals. These findings provide new insights into the Pleistocene evolution of the carbonate system in the Pacific Ocean.

Plain Language Summary

The present study is significant in quantifying large‐amplitude changes in deep Pacific [CO32‒] (from ~100 to 40 μmol kg‒1) during the mid‐Brunhes interval. We infer that the [CO32‒] maximum at the MIS 12/11 termination was the largest deglacial oceanic carbon release during the late Pleistocene, and that it may have contributed to the "mid‐Brunhes climatic shift". The [CO32‒] minimum during the mid‐MIS 11 corresponds to the well‐known "mid‐Brunhes dissolution interval", indicating a ~50 % increase in pelagic carbonate production at that time. Moreover, the present study provides further evidence for the "coral reef hypothesis", i.e., that the deep Pacific carbonate system responded to variations in shelf‐carbonate production on 100‐kyr glacial‐interglacial timescales.