Linear breakwater reefs of the greater Caribbean: Classification, distribution & morphology.

Geomorphic differences among Caribbean reefs have long been noted. These differences are considered to reflect the presence of reefs in different stages of development, following an incomplete recovery from rapid deglacial sea-level rise. But the possibility that these reflect real developmental differences caused by variation in wind, wave, and climate regime, has never been fully considered. Here, for the first time, we quantify the geomorphology and distribution of Greater Caribbean reefs using satellite images in Google Earth and public-domain bathymetry. To do this, we first standardise their classification based on shallow geomorphology, substrate depth, and physiographic setting, and then count and categorise the total number of reefs. These data show a total of 1023 linear breakwater reefs with a combined length of 2237 km. Of this total length, 80% are fringing reefs, 16% are barriers and 4% are faros and atolls. In terms of categories, there are 16 reef subtypes present, but only 9 are common. Their distribution, however, is not uniform. In particular, flat-subtypes form 60% of breakwater reefs in southern regions, but are less common in northern regions where crest-subtypes dominate (80%). To distinguish the geomorphology of these common reef subtypes, we collect size- and length-related morphometric data from their main reef zones. These data reveal that flat and crest subtypes also have morphometric differences: flat subtypes have well-constrained morphologies with statistically consistent unimodal morphometrics characterised by large back-reef zones, smaller and steeper reef fronts, and more sinuous and persistent crestlines. Crest subtypes, by contrast, have multimodal morphometrics suggesting less consistent morphologies (or unresolved subtypes), and are characterised by crestlines with lower sinuosity, more variable back-reef and reef-front areas, and slopes. These differences in geomorphology and distribution imply that flat- and crest-subtypes are neither successional stages of a single reef type, nor a genetically related sequence of types, but distinct reefal geoforms with different modes of development. In subsequent work we will explore what controls these differences.

Early recovery dynamics of turbid coral reefs after recurring bleaching events.

The worlds' coral reefs are declining due to the combined effects of natural disturbances and anthropogenic pressures including thermal coral bleaching associated with global climate change. Nearshore corals are receiving increased anthropogenic stress from coastal development and nutrient run-off. Considering forecast increases in global temperatures, greater understanding of drivers of recovery on nearshore coral reefs following widespread bleaching events is required to inform management of local stressors. The west Pilbara coral reefs, with cross-shelf turbidity gradients coupled with a large nearby dredging program and recent history of repeated coral bleaching due to heat stress, represent an opportune location to study recovery from multiple disturbances. Mean coral cover at west Pilbara reefs was monitored from 2009 to 2018 and declined from 45% in 2009 to 5% in 2014 following three heat waves. Recruitment and juvenile abundance of corals were monitored from 2014 to 2018 and were combined with biological and physical data to identify which variables enhanced or hindered early-stage coral recovery of all hard corals and separately for the acroporids, the genera principally responsible for recovery in the short-term (<7 years). From 2014 to 2018, coral cover increased from 5 to 10% but recovery varied widely among sites (0-13%). Hard coral cover typically recovered most at shallower sites that had higher abundance of herbivorous fish, less macroalgae, and lower turbidity. Similarly, acroporid corals recovered most at sites with lower turbidity and macroalgal cover. Juvenile acroporid densities were a good indicator of recovery at least two years after they were recorded. However, recruitment to settlement tiles was not a good predictor of total coral or acroporid recovery. This study shows that coral recovery can be slower in areas of high turbidity and the rate may be reduced by local pressures, such as dredging. Management should focus on improving or maintaining local water quality to increase the likelihood of coral recovery under climate stress. Further, in turbid environments, juvenile coral density predicts early coral recovery better than recruits on tiles and may be a more cost-effective technique for monitoring recovery potential.

Self-generated morphology in lagoon reefs.

The three-dimensional form of a coral reef develops through interactions and feedbacks between its constituent organisms and their environment. Reef morphology therefore contains a potential wealth of ecological information, accessible if the relationships between morphology and ecology can be decoded. Traditionally, reef morphology has been attributed to external controls such as substrate topography or hydrodynamic influences. Little is known about inherent reef morphology in the absence of external control. Here we use reef growth simulations, based on observations in the cellular reefs of Western Australia's Houtman Abrolhos Islands, to show that reef morphology is fundamentally determined by the mechanical behaviour of the reef-building organisms themselves-specifically their tendency to either remain in place or to collapse. Reef-building organisms that tend to remain in place, such as massive and encrusting corals or coralline algae, produce nodular reefs, whereas those that tend to collapse, such as branching Acropora, produce cellular reefs. The purest reef growth forms arise in sheltered lagoons dominated by a single type of reef builder, as in the branching Acropora-dominated lagoons of the Abrolhos. In these situations reef morphology can be considered a phenotype of the predominant reef building organism. The capacity to infer coral type from reef morphology can potentially be used to identify and map specific coral habitat in remotely sensed images. More generally, identifying ecological mechanisms underlying other examples of self-generated reef morphology can potentially improve our understanding of present-day reef ecology, because any ecological process capable of shaping a reef will almost invariably be an important process in real time on the living reef.

Coral colonisation of an artificial reef in a turbid nearshore environment, Dampier Harbour, western Australia.

A 0.6 hectare artificial reef of local rock and recycled concrete sleepers was constructed in December 2006 at Parker Point in the industrial port of Dampier, western Australia, with the aim of providing an environmental offset for a nearshore coral community lost to land reclamation. Corals successfully colonised the artificial reef, despite the relatively harsh environmental conditions at the site (annual water temperature range 18-32°C, intermittent high turbidity, frequent cyclones, frequent nearby ship movements). Coral settlement to the artificial reef was examined by terracotta tile deployments, and later stages of coral community development were examined by in-situ visual surveys within fixed 25 x 25 cm quadrats on the rock and concrete substrates. Mean coral density on the tiles varied from 113 ± 17 SE to 909 ± 85 SE per m(2) over five deployments, whereas mean coral density in the quadrats was only 6.0 ± 1.0 SE per m(2) at eight months post construction, increasing to 24.0 ± 2.1 SE per m(2) at 62 months post construction. Coral taxa colonising the artificial reef were a subset of those on the surrounding natural reef, but occurred in different proportions--Pseudosiderastrea tayami, Mycedium elephantotus and Leptastrea purpurea being disproportionately abundant on the artificial reef. Coral cover increased rapidly in the later stages of the study, reaching 2.3 ± 0.7 SE % at 62 months post construction. This study indicates that simple materials of opportunity can provide a suitable substrate for coral recruitment in Dampier Harbour, and that natural colonisation at the study site remains sufficient to initiate a coral community on artificial substrate despite ongoing natural and anthropogenic perturbations.