Simulated Future Trends in Marine Nitrogen Fixation Are Sensitive to Model Iron Implementation

Biological nitrogen fixation is an important oceanic nitrogen source, potentially stabilizing marine fertility in an increasingly stratified and nutrient-depleted ocean. Iron limitation of low latitude primary producers has been previously demonstrated to affect simulated regional ecosystem responses to climate warming or nitrogen cycle perturbation. Here we use three biogeochemical models that vary in their representation of the iron cycle to estimate change in the marine nitrogen cycle under a high CO₂ emissions future scenario (RCP8.5). The first model neglects explicit iron effects on biology (NoFe), the second utilizes prescribed, seasonally cyclic iron concentrations and associated limitation factors (FeMask), and the third contains a fully dynamic iron cycle (FeDyn). Models were calibrated using observed fields to produce near-equivalent nutrient and oxygen fits, with productivity ranging from 49 to 75 Pg C yr⁻¹. Global marine nitrogen fixation increases by 71.1% with respect to the preindustrial value by the year 2100 in NoFe, while it remains stable (0.7% decrease in FeMask and 0.3% increase in FeDyn) in explicit iron models. The mitigation of global nitrogen fixation trend in the models that include a representation of iron originates in the Eastern boundary upwelling zones, where the bottom-up control of iron limitation reduces export production with warming, which shrinks the oxygen deficient volume, and reduces denitrification. Warming-induced trends in the oxygen deficient volume in the upwelling zones have a cascading effect on the global nitrogen cycle, just as they have previously been shown to affect tropical net primary production.