In the 90/91 fish breeding season, four cross-shelf cruises (October, November, December and January), each lasting 10 days were undertaken in the central Great Barrier Reef. Each cruise was scheduled between the third and first quarter of the moon, with the exception of December, when Cyclone Joy curtailed operations.Four light-traps, three at the surface and one at 20m, were anchored immediately downstream of four coral reefs (Keeper Reef, Helix Reef, Faraday Reef and Myrmidon Reef) and sampled daily. These reefs, chosen for their similar size, isolation and position on the southern side of the Magnetic passage, span the full width of the reef matrix off Townsville. Each daily sample taken from a trap represented the integrated catch from three periods of fishing during the previous night: 2100-2200, 0000-0100, 0300-0400 hrs EST.This near reef sampling was complemented by sampling of a cross shelf transect, which consisted of drifting light-traps at five open-water stations in either the GBR Lagoon (Stations CS0-CS4), the Magnetic Passage (CS5-CS9) or the western Coral Sea (CS10-CS14). Depending on the weather, each nightly transect (and hence the entire transect of 160 km) was completed three times in a cruise. At each of the 15 stations, two buoy strings were released 200-300 m apart; each string supported one light-trap at the surface and another at 20m. All traps were recovered after one hour of fishing and their contents were processed as the ship steamed to the next station.Oceanographic data was also collected from a series of moored current meters, located on both sides of the reef matrix with a third in the Magnetic Passage. This coverage was considered sufficient to model hydrodynamic flows through the matrix and around individual reefs.In the 91/92 breeding season the effort near each of the four reefs was boosted by an extra trap at 20 m. Traps had been upgraded and the depth of the deep drifting traps was increased so that they sampled within 5 m of the bottom, to a maximum of 100 m. Effort was reduced in the Coral Sea due to the extra difficulty of working off the continental shelf at night and the consistently low return from such sampling. Sampling off the shelf was reduced to a single night per cruise but compensated for whenever possible by additional effort at the inshore end of the transect, which was the most productive area. Sampling was carried out between the third and first quarter moons of September - January. A further 6 sites were established across Palm Passage during this season (PP14-PP18) and sampled in October, November and December.In 92/93, three nights of sampling in the GBR Lagoon (CS0-CS4) was undertaken on each of five cruises during that summer (new moons of October-February).In 91/92 and 92/93, profiles of water temperature, salinity and light-transmission (depending on availability of a transmissometer) were captured with a Seabird (TM) CTD at each of the open water stations.All samples were preserved in 70% alcohol to preserve the otoliths in the heads of these fishes, so that a portion of them could be extracted and measured for age and growth histories by interpreting daily growth rings in their microstructure.