Concentration and distribution of phytoplankton nitrogen and carbon in the Northwest Atlantic and Indian Ocean: A simple model with applications in satellite remote sensing

• Published in: Frontiers in Marine Science 2022-11, Vol.9
• Main Authors: Maniaci, Giuseppe, Brewin, Robert J. W., Sathyendranath, Shubha
• Format: Article
• Language: English
• Subjects: carbon; chlorophyll-a; nitrogen; phytoplankton; Redfield; satellite
• ISSN: 2296-7745; 2296-7745
• DOI: 10.3389/fmars.2022.1035399

Despite the critical role phytoplankton play in marine biogeochemical cycles, direct methods for determining the content of two key elements in natural phytoplankton samples, nitrogen (N) and carbon (C), remain difficult, and such observations are sparse. Here, we extend an existing approach to derive phytoplankton N and C indirectly from a large dataset of in-situ particulate N and C, and Turner fluorometric chlorophyll-a (Chl-a), gathered in the off-shore waters of the Northwest Atlantic and the Arabian Sea. This method uses quantile regression (QR) to partition particulate C and N into autotrophic and non-autotrophic fractions. Both the phytoplankton C and N estimates were combined to compute the C:N ratio. The algal contributions to total N and C increased with increasing Chl-a, whilst the C:N ratio decreased with increasing Chl-a. However, the C:N ratio remained close to the Redfield ratio over the entire Chl-a range. Five different phytoplankton taxa within the samples were identified using data from high-performance liquid chromatography pigment analysis. All algal groups had a C:N ratio higher than Redfield, but for diatoms, the ratio was closer to the Redfield ratio, whereas for Prochlorococcus , other cyanobacteria and green algae, the ratio was significantly higher. The model was applied to remotely-sensed estimates of Chl-a to map the geographical distribution of phytoplankton C, N, and C:N in the two regions from where the data were acquired. Estimates of phytoplankton C and N were found to be consistent with literature values, indirectly validating the approach. The work illustrates how a simple model can be used to derive information on the phytoplankton elemental composition, and be applied to remote sensing data, to map pools of elements like nitrogen, not currently provided by satellite services.