Surveys of benthic communities were carried out at Davies Reef, a mid-shelf reef, and Myrmidon Reef, an outer-shelf reef in the central region of the Great Barrier Reef. On each reef, 6 comparable sites were surveyed along contiguous 10 m line transects that followed a depth contour. Sites were located at the reef crest, 5 m, 10 m, 15 m and 20 m depths on the windward slope and one reef flat site running parallel to and about 50 m leeward of the reef edge. The depths were measured using the reef crest lip as datum.At Davies Reef, 12 replicate transects were established per site, except for the reef flat, which had 18 transects (78 transects). At Myrmidon Reef, 10 replicate transects were surveyed per site (60 transects). On each transect the reef substrata were categorized using 19 morphological descriptors, some of which were taxonomically based. A plastic fibre tape marked in centimetres was laid along the depth contour and the intercepts of each substrate category were recorded as a series to the nearest centimetre. Instances of one benthic type visibly influencing its near neighbours on the transect (dead patches visible along the boundaries of one or both the neighbours) were recorded.These data were used for both spatial and structural analyses of the benthos along each transect. The same list of categories was used for each analysis, except that non-living categories such as sand/rubble and water were not included in the structural analyses. Spatial analyses involved calculation of the percent cover of each category averaged over the replicate transects. Coral/coral interactions and interactions between corals and other biota were analysed for Davies Reef only. The lengths of the gaps between hard corals, hard corals and soft corals and hard corals and sponges were measured. Structural analyses were based on the mean number of occurrences of each category per transect and the total length of each category per transect.
