Micronekton are the mid-trophic level of the ecosystem and contribute to active carbon export to the deep ocean through diel vertical migrations. Better characterization of micronekton functional groups depending on relationships to environmental variables is useful for the management of marine resources, the conservation of biodiversity and a better understanding of climate change impacts. For this purpose, regionalization of global ocean into homogeneous provinces is an approach that is generating increasing interest. However, published regionalizations efforts (i) derived from environmental forcings, that do not specifically focus on micronekton and (ii) derived from acoustic backscatter, which do not allow direct estimates of micronekton biomass. Here, we propose to fill the gap between biophysical regionalizations and micronekton biomass. We notably defined biophysical biomes using global environmental variables known to affect micronekton: temperature of the epipelagic layer, temperature stratification, and net primary production (NPP). Six biophysical biomes were defined with a clustering method. A characterization of these biophysical biomes with simulated micronekton from the SEAPODYM-LMTL model displayed biome-specific relationships between biomass and the environmental variables used in the clustering (i.e. biomasses mostly structured by NPP and temperature). Biophysical biomes also displayed specific vertical structures suggested by modelled micronekton functional groups ratios. Then, a validation of biophysical biomes’ boundaries was performed to identify potential vertical structure reorganization in acoustic backscattering response from adjacent biomes. The regionalization identified homogeneous areas in terms of acoustic vertical structure, which were also different between adjacent biomes. Finally, a comparison with another biomes’ definition computed from micronekton biomasses suggested that environmental variables can account for only some of the variability of the micronekton structures.
Read full abstract