Implications of reef ecosystem change for the stability and maintenance of coral reef islands

Abstract Coral reef islands are among the most vulnerable environments on Earth to climate change because they are low lying and largely constructed from unconsolidated sediments that can be readily reworked by waves and currents. These sediments derive entirely from surrounding coral reef and reef flat environments and are thus highly sensitive to ecological transitions that may modify reef community composition and productivity. How such modifications – driven by anthropogenic disturbances and on‐going and projected climatic and environmental change – will impact reef island sediment supply and geomorphic stability remains a critical but poorly resolved question. Here, we review the unique ecological–geomorphological linkages that underpin this question and, using different scenarios of environmental change for which reef sediment production responses can be projected, explore the likely resilience of different island types. In general, sand‐dominated islands are likely to be less resilient than those dominated by rubble grade material. However, because different islands typically have different dominant sediment constituents (usually either coral, benthic foraminifera or H alimeda ) and because these respond differently to individual ecological disturbances, island resilience is likely to be highly variable. Islands composed of coral sands are likely to undergo major morphological change under most near‐future ecological change scenarios, while those dominated by H alimeda may be more resilient. Islands composed predominantly of benthic foraminifera (a common state through the P acific region) are likely to exhibit varying degrees of resilience depending upon the precise combination of ecological disturbances faced. The study demonstrates the critical need for further research bridging the ecological–geomorphological divide to understand: (1) sediment production responses to different ecological and environmental change scenarios; and (2) dependant landform vulnerability.