Propagation method and species drive survival patterns across reef zones in coral seeding on the Great Barrier Reef

Introduction: Coral reef restoration increasingly relies on scalable methods, yet outcomes vary across species, propagation techniques, and habitats. Coral seeding, where coral propagules are settled on deployment units before outplanting, provides a flexible approach that accommodates both asexual (e.g., microfrags) and sexual (e.g., spat) propagation. Objectives: To improve predictability and efficiency of coral seeding, we tested how propagation method, species and habitat shape early survival after seeding. Methods: We conducted a multi-species coral seeding experiment at Davies Reef (central Great Barrier Reef), deploying microfrags and spat on tabs within seeding devices across ten sites spanning lagoon, back, flank, and front reef zones. Survival was monitored for ~ 10 months. Analyses included time-to-mortality, growth and generalized mixed models testing the effects of zone, flow, benthic composition and density dependence at the tab-level. Results: Microfrags outperformed spat in survival and reached ~ 10× larger mean size. Species effects zone-specific: spat survival declined at exposed (flank/front) sites, whereas microfragments remained comparatively robust. Reef zone improved model fit relative to flow alone, while site-level benthic composition did not predict survival. Microhabitat effects accounted for ~ 30% of variance, with higher survival on tabs dominated by crustose coralline algae (CCA) and lower on macroalgae-dominated surfaces. Positive density dependence was detected for Galaxea fascicularis and Montipora turtlensis , but not for Acropora loripes . Conclusion: Propagation method, species, and reef zone jointly shape coral survival, but centimetre-scale microhabitat factors are key. Microfragmentation provides more reliable early survival and growth, whereas spat contribute genetic diversity. Implications for practice: Reef-zone context should guide deployment. Exposed zones should be avoided for spat but are suitable for microfrags. Settlement substrates should minimise macroalgae and prioritise CCA. Species-specific seeding densities are recommended: higher densities benefit Galaxea fascicularis and Montipora turtlensis but not Acropora loripes . Given high within-site variability, deploying many devices at fewer well-chosen sites and incorporating fine-scale monitoring will improve outcome predictability. Combine propagation methods strategically, deploy microfrags for reliable early cover and spat to sustain genetic diversity and adaptive potential. Lastly, practical proxies such as reef zone and tab-level substrate checks are more reliable predictors of survival than coarse site-level benthic summaries.