Potential of optical and ecological proxies to quantify phytoplankton carbon in oligotrophic waters

Abstract. Satellite ocean color observations provide two proxies to estimate the phytoplankton carbon concentration, C_phyto, then used as input to models quantifying growth rates and primary production, namely the phytoplankton chlorophyll-a concentration, Chl-a, and the particulate backscattering coefficient, b_bp. Variability in phytoplankton community composition, pigment assemblages and contribution of non-algal material all interplay in the relation between these proxies and C_phyto, so that no ubiquitous relationship exists between them. It is accordingly still unclear which of Chl-a or b_bp is best suited to quantify C_phyto, or whether they both are yet each in specific trophic conditions, especially for low-productivity oligotrophic waters. Here we use a data set from the eastern Indian Ocean that includes phytoplankton cell counts, phytoplankton pigments, particulate organic carbon (POC) and inherent optical properties (IOPs) to perform a comparative assessment of C_phyto derived from either Chl-a or b_bp or cell counts combined with allometric relationships. We found significant correlations (r² > ~0.5–0.6) between the three C_phyto estimates and IOPs, Chl-a or POC when samples from all depths down to 150 m are included. When only the top 25 m are included (amenable to ocean color remote sensing), no significant relationships were found, except between the cytometry-derived C_phyto and both Chl and POC. The b_bp-derived C_phyto showed the smallest variability across the entire data set. These results warn about applying to satellite ocean color observations relationships derived from data collected throughout the euphotic layer.