Between 1 and 4 October 1985, settlement plates of coral rock were attached to moorings at Myrmidon Reef on the outer shelf, John Brewer Reef on the mid-shelf, Pandora Reef on the inner shelf and in an inshore non-reef habitat at Bowling Green Bay. Two replicate moorings were deployed at each site. At Pandora Reef, moorings were anchored in 15 m of water, approximately 2 km from the reef and about 0.5 km apart, well away from the plume of the reef. Settlement plates were attached 12 m below the surface. At Myrmidon and John Brewer Reefs, moorings were anchored in 60 m and 50 m water depth respectively, and 150 m to 200 m apart. The plates were attached at 15 m. Bowling Green Bay moorings were anchored in a water depth of 12 m, near the mouth of the Haughton River, about 300 m apart and the settlement plates were attached at 8m. All settlement plates were retrieved between 14 and 17 February, 1986 and those to be used to evaluate settlement were immediately bagged and frozen. The remaining plates from John Brewer Reef were transferred to tanks on the research vessel for use in transfer experiments. Cross-shelf differences were studied by transferring a minimum of 8 randomly selected plates from John Brewer Reef to moorings at the other sites, with the return of some plates to John Brewer to control transfer effects. Cross reef and fish predation effects were studied by transferring caged and uncaged plates to the fore reef slope (10 m depth), mid reef flat (5 m) and 3 back reef lagoon sites (5-9 m) at John Brewer Reef. A total of 50 plates were attached to racks, 5 to each rack, 35cm apart, held at a 45° angle and placed in pairs with one rack from each pair caged. Racks were deployed in 5 m of water at the first back reef site. The second back reef site was located in a similar area to the first, but the racks were covered with a layer of branching Acropora to reduce the predation pressure of large fishes. The third back reef site was located on a large sand flat at 9 m depth at least 20 m from the closest hard substratum, to determine whether non-reef fishes are effective predators of oysters. Transplant experiments were retrieved between 14 and 17 April 1986. and individually bagged and frozen. In the laboratory all plates were thawed, cleaned of sediment and dried at 50°C. Oysters were identified and measured using calipers. The condition of oysters (dead or alive) on the reef transplant plates was noted and dead specimens were examined for the presence of gastropod drill holes. Voucher specimens were retained and are housed at the Australian Institute of Marine Science.
