Little is known about the dynamics of marine food chains spanning primary to higher trophic levels on centennial and longer timescales, especially where the supply of dissolved iron limits primary productivity. To elucidate the long-term dynamics of biological productivity in the Coastal Oyashio (CO), which is a major pathway for transporting dissolved iron into the western North Pacific from winter to spring, we reconstructed the lower trophic level productivity over the last 3000 years in the CO. Our results demonstrate that the concentrations and mass accumulation rates of both Chl-a (chlorophyll a and its derivatives) and biogenic opal used as proxies of primary productivity, and steryl chlorin esters (SCEs) used as that of zooplankton productivity, show a millennial-scale increasing trend and centennial-scale variability beginning ca. AD 400. SCEs were positively correlated with Chl-a, indicating that changes in zooplankton productivity were induced by bottom-up control of primary productivity. The Chl-a and SCEs showed synchronous centennial-scale patterns with a relative abundance of sea-ice-associated diatom species transported by CO, and with a ventilation index in the Okhotsk Sea Intermediate Water. This synchronous pattern indicates that lower trophic-level productivity during the spring bloom responded to the intensity of iron-replete CO.

Atmospheric supply of iron (Fe) to the ocean has been suggested to regulate marine productivity in large parts of the world’s ocean. However, there are still large uncertainties regarding how the atmospheric inputs of dissolved Fe (DFe) influence the seawater DFe concentrations and thus net primary production (NPP). Here, we use an atmospheric chemistry model and two ocean biogeochemistry models with high (Model H) and low (Model L) sensitivities to atmospheric sources of DFe to explore the responses of ocean biogeochemistry to different types of atmospheric inputs of DFe: mineral dust and combustion aerosols. When both Fe content in mineral dust of 3.5% and Fe solubility of 2% are prescribed in sensitivity simulations, the ocean models overestimate DFe concentration in the surface ocean downwind from the North African and East Asian dust plumes. Considering different degrees of atmospheric Fe processing reduces the overestimates of DFe concentration in the North Atlantic and North Pacific. The two ocean biogeochemistry models show substantially different magnitudes of responses to the atmospheric input of DFe. The more detailed Model H shows a much higher sensitivity of NPP to the change in combustion aerosols than to mineral dust, regardless of relative inputs of the sedimentary sources. This finding suggests that pyrogenic Fe-containing aerosols are more important sources of atmospheric bioavailable Fe for marine productivity than would be expected from the small amount of DFe deposition, especially in the Pacific and Southern oceans.