The Janzen-Connell hypothesis predicts increased survival of early life stages with decreasing conspecific density and increased distance from conspecific adults. Evidence of Janzen-Connell effects in maintaining diversity is common in plant communities, but its importance in sessile marine invertebrates communities remains unclear. Under controlled aquarium conditions, we examined the effect of density-dependence and adult conspecific water treatments (absent/present) on propagule settlement success and settler post-settlement survival, along with associated spatial patterns, for six broadcast-spawning, reef-building coral species from three families. We also tested if settlement success was linked to increasing propagule species diversity for three coral species from two families. We found that the probability of settlement was density independent and not influenced by adult present water treatments Yet, adult present water treatments and settler density did have a synergistic negative effect on the probability of short-term settler survival for all species examined. Settlers also showed greater spatial aggregation as their numbers increased, but were less aggregated in adult present water treatments compared to those in adult absent water treatments. We further show evidence of significant species interactions among propagules, as settlement in single-species trials was four-fold higher compared to mixed-species trials. Our findings from controlled experimental arenas indicate that the early establishment of corals was predominantly limited by density-dependent settler-adult interactions among conspecifics and propagule-propagule interactions among heterospecifics. Thus, the proximity to established conspecific adults, settler density and species diversity of propagules are relevant drivers of local coral community diversity and structure. Based on these outcomes, we suggest that the Janzen-Connell hypothesis, with demonstrated importance for plants, is partially upheld for reef corals.

Experiments were conducted at Heron Island Research Station, on the southern Great Barrier Reef, during the November 2017 coral spawning season. We collected multiple colonies of six common, broadcast-spawning coral species (Acropora hyacinthus, A. millepora, A. valida, Astrea curta, Leptoria phrygia, and Porites cylindrica). Experiments were conducted in two outdoor raceways. Each experimental replicate was a one litre plastic cylinder container with a settlement tile placed into raceways following a randomized block design. The containers were immersed in constant flow-through seawater to maintain ambient water temperatures. Containers were filled with one of two experimental water treatments: flow-through reef seawater (“adult absent water treatment”) or water from tubs holding two conspecific adult colonies (“adult present water treatment”), and then actively swimming coral larval propagules were added to the treatment containers. With this set-up, we conducted three experiments: settlement, post-settlement survival, and diversity.

For settlement, we established five replicate containers per combination of three crossed variables: six coral species (A. hyacinthus, A. millepora, A. valida, As. curta, L. phrygia and P. cylindrica), three propagule densities (10, 50 and 100), and two adult water treatments (absent or present). The post-settlement survival experiment extended from the settlement experiment, so density (or 'the number of settlers') was a continuous variable crossed with coral species (5 levels, no P. cylindrica) and adult water treatments (2 levels). For the diversity experiment, we tested whether increasing propagule species diversity affected settlement success using three species (A. millepora, A. valida and L. phrygia), and three propagule species diversity treatment levels at a set density (one, two and three). We used the five replicate containers from the settlement experiment, testing settlement at density 50 without an adult, as the single species diversity treatment (control). For the two and three propagule species diversity treatments, we added 50 (25 of each species) and 51 (17 of each species) propagules to each of seven replicate containers.

Settlers were counted visually on all settlement tile orientations (top, bottom and sides) using a dissection microscope and mapped onto a photographic image of each tile surface using imaging software. Settlement was scored as the number successfully settled versus the number that died on each tile from the initial starting density. Post-settlement survival was scored as the number that survived versus the number that died on each tile from the settlement experiment number settled. The diversity experiment scored settlement (number successfully settled versus the number that died) for each species on each tile.

See ReadMe file. The data for the spatial analyses are not provided here as they are included in the manuscript.