Search

High‐precision uranium‐thorium (U‐Th) dating of dead branching corals from Luhuitou reef, Sanya, northern South China Sea indicates that the reef framework grew episodically over the past 7,000 years. Episodes of coral reef growth (“switch‐on”phases) occurred in response to regional warming during the mid‐Holocene Climate Optimum, Medieval Warm Period, and Current Warm Period, when the East Asian summer monsoon was strong and the East Asian winter monsoon was weak. In contrast,episodes when reef growth dramatically slowed or ceased (“switch‐off”phases) occurred during comparatively cold periods (e.g., Dark Age Cold Period and Little Ice Age) and are linked to abrupt weakening of the East Asian summer monsoon and concurrent strengthening of the East Asian winter monsoon. In the context of global warming, the northern South China Sea may become an important refugium for coral reef growth of up to 2 °C above present, but only if local anthropogenic pressures are reduced.

Reconstructing coral reef histories provides a window of understanding into reef response to changing environmental and climatic conditions over various temporal scales. Here we present the results of 117 U-Th dates from emergent reef flat and slope cores and surface death assemblages, combined with previously published fossil microatoll data, to capture the entire sequence of reef growth at Mazie Bay, inshore Southern Great Barrier Reef (GBR). Coral U-Th dates indicate that Mazie Bay reef initiated ~6,900 years before present (yr. BP) quickly filling accommodation space. While rates of vertical reef accretion (5.3 ± 1.0 mm year⁻¹) were comparable to the GBR average during the mid-Holocene (~5.0 mm year⁻¹), reef flat progradation occurred at a rate 1.5- to 6-fold previous GBR rates until 5,100 yr. BP (~70.4 cm year⁻¹). Average progradation slowed to ~7.1 cm year⁻¹ in the subsequent ~4,000 years and reef slope cores indicate this reef had largely "turned-off" by 400 yr. BP, with modern coral communities existing as a veneer over the largely senescent framework. Death assemblage dates highlight coral disturbance and recovery regimes in response to increased cyclone activity 1960-1985 AD and recent extreme sea surface temperature and flood events post 2000 AD. U-Th dating of mid-Holocene to modern coral deposits from Mazie Bay reef provides a unique insight into past reef development, response to recent disturbance regimes, and potential for future reef growth. In the case of Mazie Bay, our data suggest limited accommodation space and increased occurrence of sea surface temperature extremes will restrict future reef growth at this site.

The tropical and subtropical oceans of the Southern Hemisphere are poorly represented in present-day climate models, necessitating an increased number of paleoclimate records from this key region to both understand the Earth's climate system and help constrain model simulations. Here we present a site-specific calibration of live collected massive Porites Sr/Ca records against concomitant in situ instrumental water temperature data from the fore-reef slope of Heron Reef, southern Great Barrier Reef (GBR). The resultant calibration, and a previously published Acropora calibration from the same site, was applied to subfossil coral material to investigate Holocene water temperatures at Heron Reef. U-Th-dated samples of massive Porites suggest cooler water temperatures with reduced seasonal amplitude at ~5.2 ka (2.76–1.31i?½C cooler than present) and ~7 ka (1.26i?½C cooler than present) at Heron Reef. These results contrast the previous suggestion of a mid-Holocene Thermal Maximum in the central GBR around 5.35 ka and 4.48 ka, yet may be explained by differences in temperature of the shallow ponded reef flat (central GBR) and the deeper reef slope waters (this study) and potential large reservoir correction errors associated with early radiocarbon dates. Combining coral-based water temperature anomaly reconstructions from the tropical and subtropical western Pacific indicates a coherent temperature response across the meridional gradient from Indonesia and Papua New Guinea down to the southern GBR. This similarity in reconstructed temperature anomalies suggests a high probability of an earlier expression of a mid-Holocene Thermal Maximum on the GBR between ~6.8 and 6.0 ka.

This study was funded by the National Environmental Research Programme (NERP) Tropical Ecosystems Hub Project 1.3 "Characterising the cumulative impacts of global, regional and local stressors on the present and past bio-diversity of the GBR" to J-xZ, JMP, SGS, TRC, Y-xF, and others and Australian Research Council LIEF grant (LE0989067 for the purchase and installation of the MC-ICP-MS essential for this study) to JxZ, JMP, Y-xF, and others, as well as an Australian Postgraduate Award (APA) to NDL. We are grateful to Eric Matson from the Australian Institute of Marine Science (AIMS) for technical assistance in obtaining the fossil coral cores and to Hard Rock Earth Works (Rod Morelli and associates) in Townsville for allowing access to the corals. We also thank St. Vincent's Private Hospital, Brisbane for CT scans, and Queensland Diagnostic Imaging, Indooroopilly for X-rays. We thank an anonymous reviewer for helpful comments on an earlier manuscript. The data used in this study are available through the cited references, figures, and tables as well as in the supporting information to this manuscript.