Définition d’un indice composition pour le phytoplancton en Manche- Atlantique, à partir des données du micro-phytoplancton du REPHY et des réseaux régionaux, et des mesures complémentaires de la flore phytoplanctonique acquises avec des technique...

This deliverable is composed of two reports which propose indice of phytoplankton composition, based respectively on the results of pigment and genetic diversity methods, and the results of microscopy method. Proposals based on the results of pigment and genetic diversity methods. The present work is a continuation of several studies carried out in the past (Siano and Delmas, 2013, 2015) on investigating an indicator of phytoplankton diversity using two methods complementary to optical microscopy: the genetic diversity analyzed with metabarcoding (exhaustive analysis of the diversity and relative abundance of OTUs, i.e. groups of sequences identified by their genetic resemblance) and the pigmentary diversity (analysis of the diversity and abundance of pigments produced by photosynthetic cells). As part of the present study we aim to (i) to deepen our knowledge on the spatio‐temporal structure of phytoplankton assemblages in relation with environmental forcings for three size classes (0.2‐3 μm, 3‐20μm, >20μm, here respectively defined by convention as picophytoplankton, nanophytoplankton and microphytoplankton), and (ii) to verify the hypothesis that certain groups of the phytoplankton could be representative and descriptor of their corresponding size class. This work relies on data collected during four sampling campaigns: three of them at spatial scale (Pelgas 2012, 2013, and Phytec campaigns) and one at temporal scale (Dynapse). During Pelgas 2012 and 2013 campaigns, the sampling was done in the Gironde plume during May, on three coast‐tooffshore transects. The Phytec campaign was carried out in the Iroise sea (Bay of Brest) and the south area of Britain, including the Loire plume, bays of Quiberon, Vilaine, and Concarneau. Finally, the Dynapse campaign was carried out in the bay of Concarneau, between March and July 2012. The statistical analyses performed on metarbarcoding data show a clear separation of the three size classes studied and suggest that picophytoplankton is less variable in the time space than nano‐ and microphytoplankton. This is related to the recurrent presence in this size fraction of some OTUs such as Micromonas sp, Ostreococcus sp and Chrysochromulina rotalis. Within picophytoplankton, we found a good correlation between certain OTUs of the Mamiellophyceae group and chlorophyll b, the pigment characteristic of this class of microalgae. This correspondence confirms the good intercalibration of pigment and genetic diversity data. Chlorophyll b appeared to be a good descriptor of biomass and variability of picophytoplankton. Indeed, within the picophytoplankton, chlorophyll b is positively correlated to the total chlorophyll a, the latter being an indicator of the total picophytoplankton biomass. The phytoplankton variability, studied through the genetic and pigmentary analysis, is not explained by measured environmental parameters, certainly because of insufficiently pronounced environmental gradients during the sampling campaigns. Given the results obtained, we recommend testing the relevance of OTUs (especially those identified within picophytoplankton) and chlorophyll b as descriptors of the state of water masses on a more contrasted gradient of pressures. A sampling should be carried out on a well‐established environmental gradient beforehand in order to test the hypothesis that chlorophyll b and relative abundances of OTUs Micromonas and Bathycoccus prasinos could be used for the detecting and quantifying environmental conditions that fall anthropogenic influences. Proposals based on the results of microscopy method. As part of the implementation of the Water Framework Directive (WFD), the evaluation of the ecological status of coastal waters is based in particular on the monitoring of phytoplankton composition, indice not defined to date in France. As part of the observation and surveillance network for phytoplankton and hydrology (REPHY), established in 1984, counts of all phytoplankton species sampled in water are made by optical microscopy. This study aims to propose one or more indices of phytoplankton composition, based on counts made in microscope, to evaluate the quality of coastal waters of the Channel‐Atlantic coast, as part of the WFD. In their article, Spatharis and Tsirtsis (2010) propose a method for the selection of composition indices, based on the study of the monotonicity and linearity behavior of the relationship between the indice and the logarithm of the total abundance. In this study, the list of indices and the selection method proposed by Spatharis and Tsirtsis (2010) are taken and applied to the data from 16 monitoring locations spread across the Channel‐Atlantic coast. This study selected six indices: the indices "Odum", "Menhinick", "Camargo", "Sheldon," "E3" and "Simpson E." Further study of the relationship indice vs pressures, other than total abundance, using other statistical tools would be needed to confirm the relevance of these indices. In addition, for the indice to be operational within the WFD framework, additional work on temporal aggregation of measurements, on the definition of reference values and therefore on the calculation of EQR, are still required in connection with thematic experts.