Genomic-to-space measurements reveal large-scale ocean nutrient stress

Global ocean phytoplankton growth and primary production are intimately linked to nutrient fluctuations from seasonal to millennial time scales. Rapid recycling compromises the utility of surface nutrient or phytoplankton stocks for delineating the biogeography of global ocean nutrient stress. Here, field-measured hydrography and ‘omics biomarkers of nutritional status are coupled to a satellite remote sensing metric of cell physiology to mechanistically evaluate monthly to multi-decadal shifts in global phytoplankton nutrient stress. We observe a clear biogeography in nutrient stress aligned with variations in the nutricline depth and distinctly elevated stress in nitrogen- compared to phosphate-limited waters. Regions where cells are switching to rare forms of alternative nutrients are most stressed. Temporal modes of stress are dominated by seasonal changes, but strong signatures of natural climate cycles are also apparent. Surface ocean warming over the last twenty years has led to broad increases in nutrient stress with one notable exception. Southern hemisphere oligotrophic regions experienced declines in nutrient stress that we attribute to changes in ocean nitrogen fixation. Our integrated hydrography, genomic, and satellite remote sensing of phytoplankton physiology has uncovered contemporary changes in global phytoplankton nutrient stress.

Methods
The symbol Q has been historically used to denote the phytoplankton carbon to chlorophyll ratio (Q=C:Chl), a quantity that registers the combined effects of light and nutrient availability on phytoplankton physiology (8, 39). In this work, we define a new quantity, Q’, which represents the component of C:Chl variability attributable to nutrient stress where Qobs is the satellite-derived C:Chl ratio and Qphoto is the C:Chl values estimated for a given mixed layer light environment in the absence of nutrient limitation.  In other words, Q’ is the satellite observed C:Chl normalized to the photoacclimation component of C:Chl.  Here, Qobs was estimated directly from satellite retrievals of Chl and phytoplankton carbon biomass (C or Cphyto) and Qphoto is estimated following Behrenfeld et al. This simulation integrates effects of diel variability in the underwater light field due to time of day and vertical mixing. This includes phytoplankton exposure to darkness during dawn/dusk and periods of vertical mixing that extend below the sunlit euphotic zone, which can significantly impact cellular synthesis of Chl, and thus Q.   The model can be represented by a series of simple expressions that require only three globally available input quantities where Qphoto is the combination of a baseline solution (QDM) corresponding to C:Chl under deep mixing conditions (MLD > 6 optical depths), and a shallow-mixing correction (DQSM) that only applies when the MLD is shallower than 6 optical depths (i.e., the term converges to unity when MLD exceeds 6 optical depths). PAR is the daily integrated broadband (400-700nm) surface irradiance, KPAR is the diffuse attenuation coefficient for PAR, and MLD is the mixed layer depth. The exponent in the shallow mixing correction contains the term IML which is defined as the median irradiance within the mixed layer. Patterns of Qphoto estimated from this model broadly match the behavior of other photo-physiological indices (e.g., the photosynthetic assimilation efficiency, cellular fluorescence, in situ Q, etc.) and patterns in expected C:Chl .