An eco-physiological interpretation of phytoplankton primary production and bulk community properties within the photic layer of marine water columns - Part II: DEEP Model Development & Validation

Using the framework and key insights laid out in our companion manuscript (Behrenfeld et al., 2026), here we describe the development of a new Depth Explicit Eco-Physiological Production (DEEP) model. The DEEP model is built upon the foundations of a previous net primary production (NPP) model [the Carbon-Based Productivity Model (CbPM)] but provides explicit representation of cellular resource allocation, light-dependent macronutrient recycling, and the suppression of chlorophyll synthesis under very low light. The effect of each mechanism on NPP is interrogated and quantified using a zero-dimensional ecosystem model and cellular resource allocation model. The resulting analytical expressions are then synthesized into the DEEP model to recreate the subsurface structure of community-averaged biomass (C phyto ), chlorophyll (Chl), chlorophyll-to-carbon ratios (θ), and NPP. As a by-product of this calculation, surface nutrient inventories (S ∞ ) and nutricline depth (Z nut ) are also produced. Inclusion of the mechanisms described above in the DEEP model produces close correspondence with observations and reproduces the observed vertical structure in biological properties (Chl, θ, C phyto , and NPP) across a wide spectrum of open ocean conditions. The DEEP model represents a marked improvement in model performance compared to other contemporary models, reducing NPP error by 16% compared to the next best performing algorithm, and reducing NPP error by up to 54% compared to the ensemble mean. The greatest reduction occurs at study sites representative of stably stratified, oligotrophic conditions. This suggests that resource allocation, light-dependent recycling, and photoacclimation under extreme low light all significantly impact production in the oligotrophic gyres of the world ocean.