Fig. 5 Foraging circle refinement on the Great Barrier Reef, Australia. Panel a) shows foraging circles and refined foraging circles (over more suitable predicted habitat), mapped for each species using their GBR breeding sites. Panel b) plots increasing simulated foraging circle refinement at each GBR breeding site, and where the thresholds used for refined foraging circles mapped in a) lie in relation to the trade-off between size and refinement confidence. If GBR colonies have local tracking data then colony foraging circles are refined using GBR colony-specific models (yellow); otherwise global foraging circles are refined using multi-colony models (blue). Unrefined foraging circles are mapped and their areas shown at habitat suitability percentiles of 0. Refined foraging circles from transferability-supported refinement are shown in green. Refined foraging circles from area-limited refinement are shown in red for colonies that need further refinement to get below a 100,000 km2 limit (purple dotted line). Greater foraging circle refinement reduces the confidence that refined foraging areas include known foraging areas, as demonstrated for a red-footed booby (RFBO) colony (see supporting information for breakdown of predicted known foraging area inclusion at all GBR breeding sites).
Table 5 Summary of foraging area of the Great Barrier Reef breeding seabird community. Total foraging area is estimated from unrefined global foraging circles and refined foraging circles from transferability-supported refinement and area-limited refinement (<100,000km2). The foraging areas generated by both refinement methods are supported by associated refinement confidence (predicted probability that refined foraging circles contain known foraging areas). All species totals represent total foraging area of all colonies, with overlapping areas dissolved. For a breakdown of results per colony see supporting information.