Pandora Reef, Rib Reef and Myrmidon Reef, located 16.6 km, 57.6 km and 116.9 km from the coastline respectively, were chosen for a year long series of cross-shelf coral recruitment experiments, commencing in January 1981. Settling plates approximately 15 mm thick and 300-400 cm² in total surface area (sum of upper and lower surfaces) were cut from natural, uncleaned heads of the massive corals Porites and Goniastrea spp. Plates were attached to each of four pins on a galvanized steel support rack at an angle of 37-45°. Racks were then secured on the reef by attaching them to two solid steel reinforcement bars driven into the reef.
On each reef, racks were placed on the exposed, windward E-SE forereef slope, with plates facing into the wave surge. To investigate natural recruitment, two racks of plates (four plates per rack) were deployed at both 3m and between 14 and 18m on each reef. One rack at each depth was retrieved after 6 months and the second after 12 months. A third rack was deployed at each depth at the end of the first 6 months and retrieved 6 months later.
Transplant experiments were carried out to investigate survival of coral recruits when moved to a new environment. Two additional racks of plates (three per rack) were deployed at each depth, at each reef in January 1981. After 6 months the racks were redeployed to make all possible combinations of transplants between reefs, while maintaining original depths. A further rack of plates was deployed at each depth, at each reef in January 1981, as a control for handling. Racks were retrieved after 6 months, held onboard ship for one to several days and replaced at the same site. All transplant and control plates were retrieved within 12 months of the initiation of the experiment.
After retrieval, all plates were stored frozen at -20°C until analysed. In the laboratory, plates were thawed, cleaned of sediment and debris and dried in an oven at -45°C. Each plate was then cleaned of algae and examined using a dissecting microscope. Taxonomic identification was based primarily on calical structure. Juveniles were usually identified to family, often to genus, and occasionally to species, with scanning electron microscopy (SEM) and taxonomic guides available for corals of this region.
Coral spat were assessed for post-settlement mortality as follows: skeleton intact with at least remnants of tissue; coral overgrown and damaged; and complete colony mortality. Only the last category was utilized in estimating post-settlement mortality frequencies, which were calculated simply as the proportion in this state within a given taxon.
Intra- and interspecific competition for space involving corals and other sessile epibiota was also assessed from the coral's perspective. Associates within 8 mm of the juvenile colony were recorded and examined for interactions and a static observation of overgrowth was taken as evidence of competition for space. A competition parameter was defined as the number of competitive wins over the total number of encounters within a taxon.
Placement of settled corals or relative exposure was also assessed by differentiating between juvenile corals found on upper surfaces of the plate and those settling cryptically on the lateral or lower surface. An exposure parameter was calculated as the proportion of those juveniles in the latter state.
