Fine-scale oceanic structures (1–100 km) are short-lived (days to weeks) dynamic features. They create strong gradients of physical and chemical conditions that provide in turn a diversity of habitats for phytoplankton organisms. Some dedicated in situ studies evidenced that environmental contrasts in fine-scale features like eddies, filaments, and fronts reflect into distinct phytoplankton assemblages. These local observations suggest a possible generic role of the fine scales for enhancing biodiversity. However, due to the difficulty of detecting information on the phytoplankton community at fine resolution and on extended areas, basin-scale studies are challenging and rare, and in large part based on models. In order to address this knowledge gap, here we propose a novel method that combines in situ observations, satellite observations, and statistical tools. As a case study, we analyze satellite observations collected in the western Mediterranean Sea during April-May 2023 at the same time as the oceanographic cruise BioSWOT-Med. Satellite datasets consist in i) ocean color observations that allow the estimation of biomass of six phytoplankton functional types (PFT), and ii) high resolution altimetry data from Surface Water and Ocean Topography (SWOT) satellite. Phytoplankton community composition was derived from Gaussian mixture models applied to the relative biomass of the six PFT. This statistical approach identifies dominant assemblages (clusters) and assigns, for each observation, a probability of belonging to each cluster. Phytoplankton clusters were analyzed in conjunction with fine-scale features identified from altimetry data using Finite-Time Lyapunov Exponents (FTLE), a Lagrangian diagnostic highlighting regions of strong fluid deformation. Although not all fine-scale structures coincided with abrupt shifts in assemblage composition, our results reveal a good spatial correspondence between fine-scales structures and boundaries separating phytoplankton communities. The phytoplankton observations with the lowest probability of belonging to a cluster were mostly observed at boundaries. A second clustering, based on the observations with the lowest probability of belonging to a cluster, revealed distinct community types than the one observed in the first clustering. Our results are consistent with in situ observations of phytoplankton assemblages made during the BioSWOT-Med cruise and generalize to the NW Mediterranean basin to role of the fine-scale for enhancing phytoplanktonic diversity, suggesting a general approach that can be applied to other basins.