Competing effects of seasonal hydrodynamics and temperature-dependent development shape marine larval dispersal

Abstract. Marine larval dispersal emerges from the interaction between physical transport and biological development. Yet biophysical dispersal models commonly prescribe pelagic larval duration independently of the thermal histories experienced by larvae, potentially overlooking an important coupling between environmental conditions, development, and transport. This simplification may be particularly consequential for species with prolonged spawning seasons, because larvae released at different times encounter distinct circulation regimes and thermal environments. Here, using Japanese anchovy (Engraulis japonicus) as a case study, we coupled a Lagrangian particle-tracking model with two representations of temperature-dependent larval development, a growing degree-day (GDD) approach and an optimal growth temperature (OGT) approach, to examine how physical transport and biological development jointly shape larval dispersal. A fixed 30-day travel duration was simulated to isolate seasonal variation in hydrodynamic transport. Larvae released from Haizhou Bay on six dates between May and July during 2013–2018 experienced markedly different trajectory-specific thermal histories, resulting in substantial variation in development-dependent travel duration. Under the fixed-duration scenario, median dispersal distance increased from 128.2 km on 1 May to 168.8 km on 15 July, indicating progressively stronger hydrodynamic transport over the spawning season. In contrast, the GDD approach predicted a decrease in travel duration from 62.1 to 25.5 d and a corresponding decrease in median dispersal distance from 226.5 to 154.2 km, reversing the hydrodynamically driven seasonal trend. The OGT approach produced a nonlinear seasonal response because larval growth declined when experienced temperatures exceeded the thermal optimum, further demonstrating that the thermal–development relationship can qualitatively alter predicted dispersal patterns. Together, these results demonstrate that larval dispersal emerges from competing physical and biological controls, with spawning phenology coupling the two by simultaneously altering seasonal circulation and trajectory-specific thermal histories. Representing travel duration as an environmentally responsive biological trait, rather than a fixed parameter, may therefore improve predictions of larval dispersal and population connectivity under climate-driven shifts in spawning phenology. More broadly, our findings suggest a spatial dimension to the phenological match–mismatch framework, whereby shifts in spawning phenology may alter not only when larvae encounter suitable conditions but also where these encounters occur.