We tracked flows of carbon and nitrogen during an experimental phytoplankton bloom in a natural estuarine assemblage in Randers Fjord, Denmark. We used 13 C-labeled dissolved inorganic carbon to trace the transfer of carbon from phytoplankton to bacteria. Ecosystem development was followed over a period of 9 d through changes in the stocks of inorganic nutrients, pigments, particulate organic carbon and nitrogen, dissolved organic carbon (DOC), and algal and bacterial polar-lipid-derived fatty acids (PLFA). We quantified the incorporation of 13 C in phytoplankton and bacterial biomass by carbon isotope analysis of specific PLFA. A dynamic model based on unbalanced algal growth and balanced growth of bacteria and zooplankton adequately reproduced the observations and provided an integral view of carbon and nitrogen dynamics. There were three phases with distinct carbon and nitrogen dynamics. During the first period, nutrients were replete, an algal bloom was observed, and carbon and nitrogen uptake occurred at a constant ratio. Because there was little algal exudation of DOC, transfer of 13 C from phytoplankton to bacteria was delayed by 1 d, compared with the labeling of phytoplankton. In the second phase, the exhaustion of dissolved inorganic nitrogen resulted in decoupling of carbon and nitrogen flows caused by unbalanced algal growth and the exudation of carbon-rich dissolved organic matter by phytoplankton. During the final, nutrient-depleted phase, carbon and nitrogen cycling were dominated by the microbial loop and there was accumulation of DOC. The main source (60%) of DOC was exudation by phytoplankton growing under nitrogen limitation. Heterotrophic processes were the main source of dissolved organic nitrogen (94%). Most of the carbon exudated by algae was respired by the bacteria and did not pass to higher trophic levels. The dynamic model successfully reproduced the evolution of trophic pathways during the transition from nutrient-replete to -depleted conditions, which indicates that simple models provide a powerful tool to study the response of pelagic ecosystems to external forcings.Understanding the transfer of carbon and nutrients between the environment and autotrophs and heterotrophs is key to furthering our knowledge on biogeochemical cycling and ecosystem functioning and how both relate. Ecologists have long distinguished two trophic pathways in the pelagic environment-the herbivorous or classical food web and the microbial loop-but now acknowledge the existence of a continuum of trophic structures with the herbivorous food 1 Corresponding author (K.vdMeersche@nioo.knaw.nl).
AcknowledgmentsWe thank Joop Nieuwenhuize for analytical and logistic support, our Eurotroph colleagues for a stimulating research environment, and two anonymous reviewers for constructive feedback. We thank Wim Vyverman and Luc De Meester for constructive remarks. The modeling part of this research was performed in the frame of a master's thesis at Ghent University (K.V.d.M.).