Testing a site-prioritisation framework to maximise survival of seeded corals

Updated: 2026-06-16

We compiled coral spat survival data for eight common Indo-Pacific coral species from seeded coral deployments of the Reef Restoration and Adaptation Program (RRAP) and the BHP-AIMS Australian Coral Reef Resilience Initiative. Using these survival data, we tested how environmental conditions influence early survival and assessed the effectiveness of a site-selection framework for coral outplanting proposed by Humanes et al. (2025). Improving survival during the vulnerable early life stage is essential for successful coral restoration, and different species are expected to respond differently to local biotic and abiotic conditions. Data were sourced from multiple coral deployment studies conducted between 2021 and 2023 across four regions of the Great Barrier Reef, including both inshore and mid-shelf reefs (Moore Reef, Davies Reef, Keppel Islands, and Heron Island). For each deployment, coral larvae were produced from locally collected gravid colonies, which were then transported to the National Sea Simulator (Seasim) at AIMS where they were spawned. These larvae were reared in aquaria, settled onto concrete tabs, and placed into ceramic seeding devices. These devices were then deployed by scuba divers at reef sites at depths between 3 and 9 m. Depending on the deployment, devices were tethered to a rod or stake, fixed directly onto the substrate, or freely deployed. Spat survival was recorded as the number of tabs that had living coral spat after 240–420 days. Environmental data were compiled for each site, including reef slope, water velocity at the seabed, sedimentation, rugosity, and benthic cover (macroalgal cover, hard coral cover, percentage of consolidated substrate, and percentage of free consolidated substrate). Reef slope and water velocity were extracted from satellite bathymetry and wave models. Sedimentation and rugosity were categorised from site photographs. Benthic cover was estimated using point-based analysis of quadrat photographs. The data were analysed using generalised linear mixed-effects models with a binomial (or betabinomial) error structure. Key environmental variables related to spat survival were identified for each species using model selection. We accounted for differences in deployment duration and included a random effect of size. Then, we calculated site scores based on the framework proposed by Humanes et al. (2025) and tested whether site scores could successfully predict spat survival. We used bootstrapped out-of-sample predictions to test how reliably can site scores be used to select sites for future seeded coral deployments. This data corresponds to the manuscript "Testing a site-prioritisation framework to maximise survival of seeded corals" submitted to the Journal of Applied Science.