Mesozooplankton play an important role in pelagic ecosystems. Linking primary producers and higher trophic levels, they are also involved in the biological carbon pump through the production of particles sinking into the deep ocean. This carbon flux is amplified by the diel vertical migrations (DVM) of some mesozooplankton taxa. Recent methods, like quantitative imaging, produce an enormous quantity of data at the individual scale. Then, data can be used in a trait-based approach in order to better understand how an ecosystem works. Additionally, high throughput sequencing of environmental DNA permits the rapid identification of species at a community-level. Thus, metabarcoding can provide a complementary vision to the functional approach. In this study, we investigate variations of the morphological and taxonomic facets of mesozooplankton diversity, and link its realized and potential traits with environmental and/or ecosystem’s functioning changes. We analysed 69,950 individual images of mesozooplankton, obtained by the Zooscan, together with the 225 most abundant unique 18S V1V2 sequences, acquired by metabarcoding. Data were sampled monthly, during the day and at night, between March-2016 and May-2017, in the first 200 m of the water column at the Bermuda Atlantic Time-series Study. A morphological space was created based on 18 morphological descriptors from the entire set of images to identify the main morphological traits that vary in the mesozooplankton community. Our results showed that environment variations structured the taxonomic composition of mesozooplankton, but did not affect the overall morphology of the community. Three distinct communities were defined based on taxonomic composition, and succeeded one another throughout the study period. A co- occurrences network was built from metabarcoding data and 6 groups of taxa (or "modules") were identified. These were linked to changes in ecosystem’s functioning and/or in the community’s morphology. DVM importance was confirmed by the existence of a module with taxa preferentially sampled during the night, and by a darker opacity for migratory organisms that was linked to higher carbon export.