What ocean biogeochemical models can tell us about bottom-up control of ecosystem variability

Processes included in earth system models amplify the impact of climate variability on phytoplankton biomass and, therefore, on upper trophic levels. Models predict much larger relative interannual variability in large phytoplankton biomass than in total phytoplankton biomass, supporting the goal of better constraining size-structured primary production and biomass from remote sensing. The largest modelled variability in annually averaged large phytoplankton biomass is associated with changes in the areal extent of relatively productive regions. Near the equator, changes in the areal extent of the high-productivity zone are driven by large-scale shifts in nutrient fields, as well as by changes in currents. Along the poleward edge of the Subtropical Gyres, changes in physical mixing dominate. Finally, models indicate that high-latitude interannual variability in large phytoplankton biomass is greatest during spring. Mechanisms for producing such variability differ across biomes with internal ocean processes, such as convection complicating efforts to link ecosystem variability to climate modes defined using sea surface temperature alone. In salinity-stratified subpolar regions, changes in bloom timing driven by salinity can produce correlations between low surface temperatures and high productivity, supporting the potential importance of using coupled atmosphere ocean reanalyses, rather than simple forced ocean reanalyses, for attributing past ecosystem shifts.