Aquatic ecosystems worldwide have been negatively affected by eutrophication, many of them driven by increasing nutrient inputs from untreated domestic sewage and industrial and agricultural wastewater. During their life cycle, Alexandrium minutum and Scrippsiella donghaienis can produce resistant and revivable cysts that can accumulate in the sediments of the coastal marine environment for hundreds of years. As a working hypothesis, we presumed that these two genera have preserved their adaptation to such ecosystem changes in the form of cysts in the old sediments. From revivified cysts, some cultures of toxic and non-toxic dinoflagellates (A. minutum and S. donghaienis) were analyzed by a metabolomic approach to study the possible adaptive responses of phytoplankton to these modifications of coastal ecosystems.

Strains of Scrippsiella donghaienis were obtained from sediments isotopically dated to 1986 (ancient) and 2006 (modern) while those of A. minutum dated from 1996 (ancient) and 2006 (modern). In total 84 samples were obtained, with each experimental condition (combinations of species, age, growth phase) being cultured in triplicate in phosphate depleted conditions. Extracts of different growth phases of the ancient and modern strains of A. minutum and S. donghaienis were analyzed by liquid chromatography coupled to high resolution mass spectrometry and retreated using a workflow on a wide variety of hydrophilic and lipophilic metabolites. More specifically, two workflows have been used for univariate hypothesis testing and PCA multivariate modelling on: (i) a chemometrics platform, Agilent Mass Profiler Professional (MPP). Metabolic fingerprints of the various stages made it possible to highlight the distribution of the common and / or age-specific metabolites of the culture but also those common and specific to the different growth phases (exponential or stationary and supplemented or limited in phosphorus), and (ii) a collaborative portal dedicated to metabolomic data processing, analysis and annotation "Workflow4Metabolomics" consolidate the metabolites of the physiological and adaptive responses obtained. It is interesting to note that independent of the workflow used (MPP or Galaxy), metabolomic profiles of cultures revived from different ages were more different than those of different growth phases. While A. minutum did not show specific metabolites for the different growth phases, S. donghaienis contained specific metabolites for all growth phases, including P deficiency linked metabolites. A molecular network analysis gave an increased understanding of these metabolites produced in phosphate deficiency condition