Fringing reef growth on the central Great Barrier Reef: signatures of sea-level change, storms and sedimentation

Coral reefs globally are impacted by a range of natural and anthropogenic stressors. Inshore reefs on Australia's Great Barrier Reef (GBR) are widely argued as degraded following a decline in coral cover and diversity over the past few decades. Inshore reefs are located between the 20 m isobath and the coast, where they are exposed to a variety of natural stressors, including high turbidity and/or sedimentation, as well as episodic cyclones and associated freshwater flood events. The impacts of these natural stressors may have been amplified by anthropogenic factors since European settlement of coastal catchments along the GBR, both directly as a result of activities such as modified land use within the catchments, and indirectly via the effects of human-induced global climate change. Anthropogenic factors are commonly implicated as drivers of recent ecological changes on inshore reefs, though isolating the various effects of natural and anthropogenic stressors remains difficult and poorly understood, partly due to a paucity of long-term data on baseline reef condition and variability prior to European settlement. Long-term records from coral reefs have great potential to address this deficit, and to detail past variability in reef growth and ecology to improve our understanding of present and future reef condition. However, such records are rare. The main aim of this research was to investigate in detail the Holocene development of fringing reefs over a cross-shelf transect in the central GBR, as baseline context for understanding present reef condition. In particular, the objectives of this research were to: 1) determine reef initiation ages and antecedent substrates; 2) reconstruct the chronostratigraphy of the fringing reefs along this transect to establish past rates and styles of reef development and any variability over time, including detailed examinations of the palaeo-ecological coral community compositions; 3) investigate the influence of Holocene sea-level change and cyclones on reef development; 4) describe and quantify the contemporary ecological community composition and structure and determine whether this has changed since European settlement; and 5) investigate Holocene reef development and present reef condition across the shelf, to identify variability and similarities across this gradient, and to examine how such patterns reflect the influence of key environmental parameters. This research focused on fringing reefs at four sites in the central GBR near Bowen that extend across a 40 km gradient from the mainland coast to the mid-shelf. The fringing reefs were located at: a) the mainland-attached Bramston Reef; b) Stone Island, ~3 km further offshore; c) Middle Island, ~10 km offshore; and d) Holbourne Island, ~40 km north of Bramston Reef. These sites were chosen because they provide a unique opportunity to examine Holocene fringing reef development across a mainland to mid-shelf transect within the central GBR where valuable historical records of reef condition are also available that extend back to the end of the 19th Century. In total, 42 reef cores were collected across the four sites. Sedimentological and palaeoecological analyses, coupled with uranium-thorium (U-Th) dating were used to develop chronostratigraphic records of reef growth. In addition, the ages and elevations (measured very precisely using a Real Time Kinematic Global Positioning System) of un-moated fossil Porites microatolls were used as a proxy for past sea level and documented the minimum age for reef flat development. Contemporary reef geomorphology and ecological community structure were quantified using a variety of techniques, including precise topographic surveying, underwater videography and photo quadrat surveys. The chronostratigraphic records of reef growth revealed that all of the reefs examined in this study initiated in the early- to mid-Holocene, between 5,396 ± 51 yBP (Bramston Reef) and 7,873 ± 17 yBP (Middle Island). Generally, initiation occurred earlier at the further offshore sites, probably as a result of the pre-reefal foundations being flooded earlier during the postglacial transgression. The reefs established over a variety of substrates, including unconsolidated transgressive sands and gravels (Bramston Reef), compacted regolith (Middle Island), and last interglacial reef (Holbourne Island). The mode of reef development varied subtly between sites and was affected by the shape of the underlying pre-Holocene surface, variations in sedimentation, the degree of exposure to cyclones, and Holocene sea-level instability. All reefs rapidly accreted vertically and began to form a reef flat at sea level within ~1,000 years of initiation, regardless of the reef start-up time. Average rates of vertical reef accretion were highest at the inshore locations (up to 9.5 mm/yr), where the non-framework reef matrix sediments include a high proportion of mud (up to 53.8 ± 17.4% on average). The high mud content contributed to rapid net reef accretion by burying the reef framework, enhancing coral framework preservation, limiting the impacts of bioerosion and contributing to reef structure volume. Reef flat development began in the mid-Holocene when sea levels were up to 1.0 m higher than present, as recorded by back reef fossil microatolls at Stone Island and Bramston Reef, and accretion continued as sea-level fell and stabilised at the present level. This late-Holocene sea-level fall is reflected in the slowing of reef accretion at most sites after the majority of the reef flat was emplaced. At Stone Island, the reef flat was entirely emplaced by ~5,000 yBP and little reef accretion has occurred since. Similar patterns were observed at other locations where relatively little reef growth occurred over the past 5,000 years (Middle Island and Holbourne Island) to 2,000 years (Bramston Reef). Palaeo-ecological analyses of the coral framework within the cores revealed the inshore sites were comprised of a diverse coral assemblage that persisted throughout the Holocene. Twenty-five and 28 coral genera were firmly identified in the cores from Bramston Reef and Stone Island, respectively, while 15 and 10 genera were identified in the cores from Middle and Holbourne Island, respectively. These estimates of diversity are likely to be conservative, as a considerable proportion of the material in most cores, particularly from Middle and Holbourne Islands, was comprised of detrital material, which was hard to confidently identify to genera. The geomorphological impacts of cyclones on past reef development were most evident at the two locations furthest offshore (Holbourne and Middle Islands). Chronological gaps in the internal reef structure during the mid-Holocene of 3,500 or 5,000 years at Holbourne Island and Middle Island, respectively, are attributed to stripping of the outer reef flat and upper reef slope by cyclones. Geomorphological features on the reef flats and islands at these locations, including shingle ridges and basset edges, provide complementary depositional evidence to support the hypothesis that the outer reef framework was stripped by cyclones. Radiometric dating of fossil microatolls on the reef flat at these two sites indicates that the fossil microatolls are relatively young (