Gradual and abrupt changes during the Mid-Pleistocene Transition

During the Mid-Pleistocene Transition (MPT), the dominant glacial-interglacial cyclicity as inferred from the marine δ18O records of benthic foraminifera (δ18Obenthic) changed from 41 kyr to 100 kyr years in the absence of a comparable change in orbital forcing. Currently, only two Mg/Ca-derived, high-resolution bottom water temperature (BWT) records exist that can be used with δ18Obenthic records to separate temperature and ice volume signals over the Pleistocene. However, these two BWT records suggest a different pattern of climate change occurred over the MPT-a record from North Atlantic DSDP Site 607 suggests BWT decreased with no long-term trend in ice volume over the MPT, while South Pacific ODP Site 1123 suggests that BWT has been relatively stable over the last 1.5 Myr but that there was an abrupt increase in ice volume at ∼900 kyr. In this paper we attempt to reconcile these two views of climate change across the MPT. Specifically, we investigated the suggestion that the secular BWT trend obtained from Mg/Ca measurements on Cibicidoides wuellerstorfi and Oridorsalis umbonatus species from N. Atlantic Site 607 is biased by the possible influence of Δ[CO32−] on Mg/Ca values in these species by generating a low-resolution BWT record using Uvigerina spp., a genus whose Mg/Ca values are not thought to be influenced by Δ[CO32−]. We find a long-term BWT cooling of ∼2-3°C occurred from 1500 to ∼500 kyr in the N. Atlantic, consistent with the previously generated C. wuellerstorfi and O. umbonatus BWT record. We also find that changes in ocean circulation likely influenced δ18Obenthic, BWT, and δ18Oseawater records across the MPT. N. Atlantic BWT cooling starting at ∼1.2 Ma, presumably driven by high-latitude cooling, may have been a necessary precursor to a threshold response in climate-ice sheet behavior at ∼900 ka. At that point, a modest increase in ice volume and thermohaline reorganization may have caused enhanced sensitivity to the 100 kyr orbital cycle.