Coral-seeding devices with fish-exclusion features reduce mortality on the Great Barrier Reef

Full description

In an 8-month field study, we assessed the efficacy of coral-seeding devices with refugia from fishes (i.e., ‘fish-exclusion’ features) to limit predation-driven mortality of Acropora digitifera spat and microfragments.


Mature gravid Acorpora digitifera (Dana 1846) colonies were collected from Davies Reef (central mid shelf Great Barrier Reef) ahead of Autumn coral spawning (February to March) in 2021. The corals were transported back to the Australian Institute of Marine Science National Sea Simulator (SeaSim) facility and kept in temperature controlled outdoor aquaria. Egg-sperm bundles were collected during spawning from 3-13 colonies and cross-fertilized. The coral larvae were settled onto conditioned, aragonite coral “frag plugs” (Ocean Wonders) after 7 days. The plugs contained a mixed community of crustose coralline algae and bacteria biofilms that are known to induce settlement of Acorpora corals; these were conditioned in aquaria for 2-months prior to settlement. Five broodstock colonies were selected for microfragmentation 3-months after spawning. Colonies were chiseled into fragments (~10 cm length), then cut into many microfragments (8 mm2). Coral spat and microfragments were held in aquaria until deployment; time in aquaria was equivalent to 4 months for spat and 1 month for microfragments.


Three devices with nominally different levels of grazing exclusion and coral protection were selected for this experiment: (1) a featureless device (open to grazing); (2) a caged featureless device (complete exclusion of grazing); and (3) a device with engineered fish-exclusion features (protection from grazing and our device of interest). The experimental device was designed at AIMS as a part of the Reef Restoration and Adaptation Program (RRAP) and represents the first design of ceramic devices to support field deployment of juvenile corals grown on standard 20-mm (Ø) coral plugs.


Devices with corals were then deployed at Davies Reef onto 4 replicate sites in August 2021. Sites were selected based on their leeward location, shallow depth (< 3 m at high tide), and preexisting population of A. digitifera corals. Devices were surveyed over 3 time points (2, 90, and 240 days) post deployment. Data collection included assessments of coral survival, fish abundance, fish grazing, benthic composition, sedimentation, and wave energy, at the 4 sites. All experimental materials were removed from the reef after 8 months.


Coral survival and grazing data were analysed using generalized linear mixed effect models (GLMMs) in R statistical software. Redundancy analyses and linear regression models were also used to explore the effects of ecological variables on survival and grazing. Zero-inflated GLMMs were used to compare fish abundance and feeding across sites.

Lineage

Maintenance and Update Frequency: notPlanned

Statement: Survival and grazing data were evaluated through image analysis and the manual identification and counting of individual corals or fish bites on plugs of each device. Fish abundance was observed by divers during stationary point count surveys. GoPro video footage was used to obtain fish feeding and grazing information in 0.25 m2 quadrats around devices. Fish bites were manually counted during video analysis. 0.25 m2 quadrats were used to identify and estimate the cover of benthic community constituents growing near devices. Quadrats were imaged and analysed manually using the Reef Cloud online data base (https://reefcloud.ai/). Clod cards were used to compare relative water flow among sites. Water flow data was expressed as dissolution over time and using the Diffusion Factor (Jokiel and Morrissey 1993). ‘SedPods’ and ‘TurfPods’ were used as sediment collection devices (Field et al 2013). Sediment processing followed standard water and wastewater methods from the American Public Health Association 2018. The files include data from all timepoints that were used in the analyses. Missing and disturbed devices were not included in the data. All data were quality checked after entry.

Notes

Credit
Negri A

Credit
Hoogenboom M

Credit
James Cook University (JCU), Australia

Credit
AIMS@JCU, Australia

Credit
Randall C