Productivity changes across the mid-Pleistocene climate transition

We use benthic foraminiferal accumulation rates as a proxy for productivity changes during the mid-Pleistocene climate transition (MPT) (~1.2 Ma to 0.4 Ma). Our data are chosen to test the hypothesis that longer-term cooling and the onset of 100 kyr cyclicity are linked to atmospheric CO2 draw-down associated with an increase in primary productivity. To this end, we have constructed records from a global array of seven sites spanning major ocean basins and representing different hydrographic regimes (e.g., high and low latitudes, upwelling versus the deep western warm pools). We compare our data to published productivity proxy records from each site to identify limitations and uncertainties in the reconstructions. Results indicate that there is evidence for productivity increases during the onset of the MPT (1.2-1.0 Ma), but the changes are not globally synchronous and likely reflect regional hydrographic variability. On the orbital scale, productivity maxima tend to occur more closely related to glacial than interglacial intervals overall, but the relationships are not consistent. High interglacial productivity characterizes low latitude sites some of the time. In the obliquity band, high interglacial productivity in the eastern equatorial Pacific coincides with low interglacial productivity in the Southern Ocean, supporting a high to low latitude link via intermediate water circulation distribution of nutrients. On the regional scale, our records contribute new evidence for changes in Northern Hemisphere frontal systems during the MPT and for a close link between surface ocean production of organic matter and consumption on the ocean floor in the western tropical Atlantic. Pyrite counts at the two Southern Ocean sites provide supporting evidence for sluggish thermohaline overturn during the mid-point of the MPT at ~900 ka. Taken together, our records do not show a globally synchronous productivity signal that would support the biological pump as a driver for potential CO2-induced climate cooling during the MPT. Instead, we document complex regional variations in the carbon cycle, reflecting a combination of both biological and physical processes both on the longer as well as on the orbital time-scale.