Accueil > Actualités > Séminaires > Séminaire de Lars Stemmann au LOCEAN

Séminaire

Titre : Global in situ imaging of large plankton and particles by the Underwater Vision Profiler in the upper kilometre of all oceans
Nom du conférencier : Lars Stemmann
Son affiliation : Professeur Sorbonne Université, Laboratoire d’Océanographie de Villefranche sur Mer
Laboratoire organisateur : LOCEAN
Date et heure : 13-12-2018 15h00
Lieu : Campus de Jussieu, salle de séminaire du LATMOS, tour 45/46, 4eme étage
Résumé :

Large plankton and detrital particles play key roles in the biological pump but they are difficult to measure simultaneously in situ. The Underwater Vision Profiler (UVP) allows quasi-continuous profiling of the water column to detect organisms and non-living particles larger than 100 μm. Images of objects >500μm can be classified semi-automatically into taxonomic and functional categories using the newly developed ECOTAXA web application. Since 2008, more than 6000 UVP profiles have been performed in the upper kilometre of all oceans and in several seas (Mediterranean Sea, Baffin Bay, Baltic Sea) providing millions of images. Until now approximately 25% of the profiles have been sorted in a consistent manner using the same taxonomic framework (47 categories). The efficiency of the UVP in capturing motile zooplankton was tested by comparing UVP copepod counts to net collected copepod abundances. This unique data set allow us to estimate zooplankton biomass and particle flux on a global scale. Zooplankton biomasses have been calculated using relationships between mass and size and our estimates are consistent but in the higher range of those found in the literature. Global estimates of calculated particle flux is also consistent with estimates obtained by sediment traps or thorium analysis. Large protists (rhizarians), bacterial filaments and copepods together contribute more than 80% of the organism’s abundance in the upper layer (0-200m). In the deep layer (500-900m), copepods and rhizarians are the major contributors to the living stock. Results show that plankton communities in both layers are spatially structured following surface phytoplankton production patterns, particle accumulation and geochemical conditions. However, the spatial pattern tends to be less marked in the deep layer. This global dataset prefigures what will arise from an observation network of cameras deployed during synoptic campaigns and in the future by autonomous platforms.

Contact :

francesco.dovidio@locean-ipsl.upmc.fr