Complementary constraints from carbon (13C) and nitrogen (15N) isotopes on the glacial ocean's soft-tissue biological pump

A three-dimensional, process-based model of the ocean’s carbon and nitrogen cycles, including C and N isotopes, is used to explore effects of idealized changes in the soft-tissue biological pump. Results are presented from one preindustrial control run (piCtrl) and six simulations of the Last Glacial Maximum (LGM) with increasing values of the spatially constant maximum phytoplankton growth rate μmax, which accelerates biological nutrient utilization mimicking iron fertilization. The default LGM simulation, without increasing μmax and with a shallower and weaker Atlantic Meridional Overturning Circulation and increased sea ice cover, leads to 280 Pg more respired organic carbon (Corg) storage in the deep ocean with respect to piCtrl. Dissolved oxygen concentrations in the colder glacial thermocline increase, which reduces water column denitrification and, with delay, nitrogen fixation, thus increasing the ocean’s fixed nitrogen inventory and decreasing δNNO3 almost everywhere. This simulation already fits sediment reconstructions of carbon and nitrogen isotopes relatively well, but it overestimates deep ocean δCDIC and underestimates δNNO3 at high latitudes. Increasing μmax enhances Corg and lowers deep ocean δ CDIC, improving the agreement with sediment data. In the model’s Antarctic and North Pacific Oceans modest increases in μmax result in higher δ NNO3 due to enhanced local nutrient utilization, improving the agreement with reconstructions there. Models with moderately increased μmax fit both isotope data best, whereas large increases in nutrient utilization are inconsistent with nitrogen isotopes although they still fit the carbon isotopes reasonably well. The best fitting models reproduce major features of the glacial δCDIC, δ N, and oxygen reconstructions while simulating increased Corg by 510–670 Pg compared with the preindustrial ocean. These results are consistent with the idea that the soft-tissue pump was more efficient during the LGM. Both circulation and biological nutrient utilization could contribute. However, these conclusions are preliminary given our idealized experiments, which do not consider changes in benthic denitrification and spatially inhomogenous changes in aeolian iron fluxes. The analysis illustrates interactions between the carbon and nitrogen cycles as well as the complementary constraints provided by their isotopes. Whereas carbon isotopes are sensitive to circulation changes and indicate well the three-dimensional Corg distribution, nitrogen isotopes are more sensitive to biological nutrient utilization.