Abstract
Phytoplankton-derived organic matter sustains heterotrophic marine life in regions away from terrestrial inputs such as the Southern Ocean. Fluorescence spectroscopy has long been used to characterize the fluorescent organic matter (FOM) pool. However, most studies focus only in the dissolved FOM fraction (FDOM) disregarding the contribution of particles. In order to assess the dynamics and drivers of the dissolved and particulate fractions of FOM, we used a Lagrangian approach to follow the time evolution of phytoplankton proliferations at four different sites in the Southern Ocean and compared the FOM in filtered and unfiltered seawater aliquots. We found that filtration had little effects on FOM visible spectrum fluorescence intensities, implying that most of this signal was due to dissolved fluorophores. On the other hand, protein-like fluorescence was strongly supressed by filtration, with fluorescence of particles accounting for up to 90 % of the total protein-like FOM. Photobleaching was identified as the main driver of visible FDOM composition, which was better described by indices of phytoplankton photoacclimation than by measurements of the incident solar radiation dose. In contrast, protein-like FOM intensity and fractionation were primarily related to abundance, composition and physiological state of phytoplankton proliferations. The chlorophyll a concentration from non-diatom phytoplankton explained 91 % of the particulate protein-like FOM variability. The proportion of protein-like fluorescence found in the dissolved phase was predicted by the combination of potential viral and grazing pressures, which accounted for 51 and 29 % of its variability, respectively. Our results show that comparing FOM measurements from filtered and unfiltered seawater provides relevant information on the taxonomic composition and cell integrity of phytoplankton communities. A better understanding of the commonly overlooked FOM fractionation process is essential for the implementation of in situ fluorescence sensors and will also help us better understand the processes that govern OM cycling in marine systems.