Going with the flow: Leveraging reef-scale hydrodynamics for upscaling larval-based restoration.

Anthropogenic pressures are impacting coastal marine ecosystems, necessitating large-scale interventions to accelerate recovery. Propagule-based restoration holds the potential for restoring shallow coastal systems at hectare scales by harnessing natural dispersal. However, predicting propagule dispersal remains challenging due to the complex hydrodynamic nature of coastal marine ecosystems and the complex behaviors of marine propagules. To improve predictions of fine-scale larval dispersal patterns, we developed a 3D reef-scale (~30-m resolution) dispersal model for Lizard Island, Australia, with the aim to predict the effect of island-scale hydrodynamics on the distribution of coral spawn slicks and larvae. Using in situ field observations and dispersal simulations, we assessed the model's capability to (1) forecast hydrodynamic conditions, (2) predict coral spawn slick convergence zones for collection efforts, and (3) identify optimal locations and timeframes where high particle residence time may enhance local settlement following larval delivery to damaged reefs. Predictions of convergence zones in the upper water column aligned well with field observations of coral spawn slicks. At the reef benthos, the model captured variability in current speed and direction at ~58% of studied locations. At other locations, the model did not resolve hydrodynamic conditions due to sheltering effects and associated hydrodynamic processes occurring at a scale below 50 m. At locations where the model performed well, propagules could remain within a 1-ha area around the delivery site for 5-15 h depending on locations and the timing of larval release. These high retention conditions were infrequent but occurred at least once at 15 of the 25 studied sites. Observations of local currents a posteriori confirmed model predictions, showing periods of little water movement lasting from 6.5 to 15 h. Overall, our study highlights fine-scale dispersal modeling as a key tool for scaling up larval-based reef restoration, while also acknowledging the need for better predictions of local conditions in complex, shallow environments. Applications of fine-scale modeling, coupled with local knowledge of reproductive timing and larval behavioral ecology, assist with the mass collection of propagules upon release and in identifying areas and times of optimal larval deployment to achieve the greatest impact.