Degree Name

Bachelor of Science (Honours)


School of Earth & Environmental Sciences


Sarah Hamylton


The aim of this honors project was to map changes in the sea floor communities at the regional scale within the Capricorn Bunker Group between the years of 2001 and 2014. The Capricorn Bunker Group is located at the southern end of the Great Barrier Reef near 30⁰ south and 150⁰ east and consists of 16 islands with 22 reefs, with 8 of these islands being vegetated and little in the way of topographic relief (Queensland Government, 2014).

This project accomplished its aim through a three stage method after the second set of benthic community ground coverage data had been gathered on the 2014 Joint Benthic Field & Remote Sensing Survey as under taken by Sarah Hamylton et al (2014). Initially bathymetry DEM’s were created according to the method proposed in Stumpf & Holderied’s 2003 paper with minor adjustments based upon the source of the Satellite image. Then the reefs were mapped on the basis of their geomorphic zones as found via satellite images overlaid on top of bathymetry DEM’s in ARC Scene. The mapping allowed us o subsequently subsample the data on the basis of the desired geomorphic zone. Finally the change detection analysis was undertaken which uncovered a variety of results through the comparison of video records of benthic community ground cover. A wide variety of data was used in the project and this included bathymetry layers generated from satellite imagery, depth measurements photographs and video transects. A wide variety of community components were made available for study through the video surveys which included substrate type, the presence of hard aglae and differing coral morphologies. Of these it was decided to focus on the hard coral branching and hard coral massive classes for their role in the reef building process and regional scale community formation. Between the 2001 and 2014 survey results with over 1500 points in the survey datasets with details of the field data collected can be found in Table 2 along with the sort of pre-processing that was undertaken.

In terms of general results most of the survey points experienced no change or some degree of growth over the survey period as seen in Figure 7. However it is to be noted that, that Figure 7 only serves as a relative indicator in terms of a positive or negative change in coral cover without taking into account the amount of change.

Change analysis revealed that the changes to the amount of hard coral cover were heterogeneous in their distribution with regards to each reefs geomorphic zones with the worst coral losses being sustained on the fore reef sections of most of the reefs with the reef flats and lagoons making nominal gains with a few exceptions. Overall 45 % of the points showing a gain in coral cover, 37% of the points showed no change and 18% of the points showed a decline. Additionally in Figure 8 the declines, whilst outnumbered by the points that showed a gain in coral cover, were of a larger magnitude than most of the gains.

The scale of these changes in coral cover adds some additional context to the general change categories. The points experiencing losses typically could expect to experience losses of a greater size than the minor gains made by the other geomorphic zones leading to a potential net loss of coral cover.

On the reefs themselves some smaller scale trends were rendered visible. Whilst there were one or two exceptions the averaged change for the group as a whole ranged from negative four percent to positive four percent. At a broader spatial scale it was uncovered that most of the reefs whose fore reefs were experiencing losses were father away from the coast and could be divided from the fore reefs that were registering growth by a line trending North west to south east. Lagoons also exhibited this trend with the northwest to south east division with the losses being located mostly towards to coast. Additionally it was observed that most of the losses were occurring on smaller hot spots of averaged coral cover.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.