Impacts of specific or general biological parameterisations in calibration of larval transport models for the anchovy and sardine in the Mediterranean Sea

Highlights: • High densities of sardine and anchovy larvae estimated above continental shelves. • Different biological formulations in larval transport model were tested. • Feeding larval growth resulted in significant dispersal differences. • Wave-driven flow impact was strong for the winter-spawner sardine. The early life pelagic period of marine species plays a key role in shaping population dynamics, as it is influenced by a range of spatio-temporal environmental and biological factors. Over the past decades, numerous studies have developed empirical relationships and mathematical formulations describing the evolution of larval traits over time, sometimes incorporating environmental drivers. These formulations are commonly used to parametrise Largrangian models that simulate the transport of early life stages under realistic oceanographic conditions. When multiple formulations are available for a given species, as it is the case for the European anchovy (Engraulis encrasicolus), a summer spawner, and the European sardine (Sardina pilchardus), a winter spawner, the choice of parameterisation can become a critical source of uncertainty. In this study, we investigate the influence of the biological parametrisation (growth and behaviour) and environmental forcing (waves) on larval transport for both species. We compiled and reviewed available formulations from the literature and implemented them in a series of Lagrangian simulations, varying one parameterisation at a time. First, we characterised trajectory statistics using a reference parameterisation set. We then compared larval duration, travelled distance, and spatial distribution at the late post-flexion stage. Our results show that both species exhibit similar aggregation areas of late-stage larvae in the Gulf of Lion and along the Ebro shelf across their respective spawning seasons. However, their broader spatial distributions differ substantially, particularly in offshore regions such as the Balearic Sea. Among the tested factors, larval growth under feeding conditions had the strongest influence on transport trajectories, primarily by extending the time spent in highly dynamic pelagic environments. These findings highlight the importance of carefully selecting biological parameterisations in Lagrangian modelling studies, while also comparatively identifying subtle and minor influence from environmental and biological conditions. When multiple formulations are available, prioritising those derived from local observations (e.g., the northwestern Mediterranean Sea) is essential for improving model realism.