Modified north west shelf fish traps were used to sample reef fishes at 40-45 m depth along the northern slope (flat habitat) and north-eastern slope (broken habitat) of Rib Reef and in two habitats of different rugosity on the back reef at Davies Reef. Sampling was carried out at Rib Reef between 23 April and 1 May, 1992 and at Davies Reef between 23 and 30 May, 1992.At both reefs, echo-sounding profiles (perpendicular to the reef crest) were run from shallow water (<20 m) through each trap location and out onto the off-reef floor using a paper-recording sounder to give a permanent record of each habitat. A video camera was towed close to the bottom, on transects normal to the depth contours and passing within metres of each trap to identify the type of substrate and benthic organisms in the area where traps were deployed. Each transect was centred on a trap and was approximately 100 m long, generally running from approximately 35 m depth to 45 m. For most traps, two parallel transects were run either side of the trap.A maximum of 12 traps were deployed at any one time and the position of each trap was recorded with a GPS. Traps were baited with approximately 1 kg of mulched Western Australian pilchards (Sardinops neopilchardus), placed in crab pot style bait canisters. Fresh bait canisters were placed in traps every time that the traps were set. Traps were set a minimum distance of 100 m apart and ranged up to a several hundred metres apart in order that the capture fields of each trap would not overlap.Setting and hauling of traps was carried out in the early morning and evening. 'Night' set traps were hauled from 0530h, with total soak times varying from 12 to 14 hours. Sorting and processing of trap catches was usually completed by 0830h and usually, as each trap was emptied it was refilled with a fresh bait canister and redeployed for a 'Day' set. 'Day' sets were hauled from 1630h onwards, with total soak times varying between 9 and 11 hours. Sorting and processing was usually completed by 1830h and traps redeployed. At both reefs, simultaneous comparisons between the two habitats were carried out on five days (n = six traps per habitat) whilst on the remaining four days, all 12 traps were placed in the habitat with the greatest catch rates of the target species. In contrast to Rib Reef, however, the timing of these two different series was randomised among days at Davies Reef. This was done to avoid the possibility of a consistent change in catchability of fish over the sampling period confounding the results of the two series.The catch from each trap was placed in bins of running seawater. Each fish was identified to species, measured to the nearest millimetre (standard length, length to caudal fork, and body depth), tagged and released or frozen for studies of age and growth.In addition to the traps at Davies Reef, drop-lines were deployed in the 40 m depth stratum among the traps, on three days. The drop-lines were similar to those developed by a professional line fisherman (Mr Paul Whelan) for catching red snappers. Each of the six drop-lines had 5-hooks (medium-sized tuna circle) and were baited with squid. Fishing activity was divided into morning (0800 - 12ooh), afternoon (1430 -1700h) and night (2000- 2330h) sessions. These sampling times were determined by the logistic demands of the concurrent trap sampling. As soon as all six lines had been deployed between the traps, the first deployed was pulled in, rebaited and redeployed, followed immediately by the redeployment of the second line and so on. Hauling and redeploying was thus a continuous activity for the duration of a fishing session.
