Abstract
Macrophytes are a key component of lagoon ecosystems, and the spatio-temporal variation of their biomass reflects the effect of environmental drivers. Understandingthese dynamics is essential for managing, preserving and anticipating the fate of such fragile ecosystems. First, using both satellite data and field observations,we showcase here a machine learning approach to accurately detect submerged vegetation in a preserved lagoon in South France, the Bagnas lagoon. Second,we investigate the statistical relationship between the surface area of Stuckenia pectinata and abiotic factors known to influence the life history of this aquatic plant.We thus identify and evaluate the factors that most impact Stuckenia pectinata population dynamics within the lagoon. We find that water temperature, salinity, anddepth are significantly related to submerged aquatic vegetation (SAV) coverage. We show that the SAV surface area reaches a peak during summer, which can varyacross years. Furthermore, we examine the lifetime distribution of SAV during each year cycle across the pond. We show that the average annual lifespan differssignificantly from one year to the next, which can be explained by annual differences in exposure time to sufficiently high water levels.