Spectrally-resolved biogeochemical models of the ocean describe the penetration of different wavelengths of light along the water column as they are attenuated by optically active constituents. Phytoplankton is one of such optically active elements but the optical properties of the community are variable depending on the pigment composition and the cellular size of the populations, both ultimately driven by changes in taxonomic composition and physiological state. Therefore, representing the optical properties of the full phytoplankton community and their sources of variability into spectrally-resolved models is a crucial task for connecting phytoplankton biomass to phytoplankton light absorption (a_PH (λ)). In this work, we explored two alternative ways of portraying the variability in light-harvesting coefficients into bio-optical models. On one side, we used a stoichiometric approach, where the coupled physical-biogeochemical model MITgcm-Regulated Ecosystem Model version 2 (REcoM2) simulated a comprehensive pigment signature of the phytoplankton community and made the absorption coefficients variable as a function of the content in photoprotective carotenoids. On the other hand, we used an approach based on the description of the functional diversity of the community, where the OGSTM-Biogeochemical Flux Model (BFM) included several phytoplankton functional types (PFT´s) with fixed absorption coefficients that covered most of the observed variability of such coefficients in nature. For the two approaches, we evaluated to what extent they were capable of reproducing the bio-optical relationships between a_PH (λ) and phytoplankton biomass commonly observed in nature. Representing the variability of absorption coefficients through the impact of photoprotective pigments simulated exponents of the relationship comparable to those observed in the global ocean. The accurate representation of the variability of light absorption coefficients through the role of photoprotection impacted significantly the underwater light field and the simulated net primary production. Hence our results show the important implications of simulating accurately the variability of phytoplankton absorption coefficients in the ocean.