Degree Name

Doctor of Philosophy


School of Earth and Environmental Sciences


The world’s coral reefs provide valuable socio-economic services to millions of people and are host to a biodiverse array of marine life. Yet they face an uncertain future given increasingly heightened exposure to natural and anthropogenic disturbances, which can interact and / or occur at such high frequencies and intensities that coral reefs are unable to recover sufficiently between events to maintain a stable state dominated by hard-corals. The primary goal of current coral reef conservation strategies is to protect a reef’s capacity to recover following a disturbance or disturbances (i.e. resilience), and thus remain in a hardcoral dominated state. To do so effectively requires an understanding of not only of how reefs are affected by individual disturbances, but also how multiple disturbances can and will likely interact to produce unexpected ecological consequences (ecological surprises). It is well recognised that disturbances frequently interact synergistically to affect coral reefs more severely than would be expected simply from the sum of the individual events. However, recent work in the Caribbean found that tropical cyclones (TC) can also interact antagonistically with thermal stress, such that the latter affects reefs less severely than would otherwise be expected. The turbulent ocean mixing, disruption of large scale circulation patterns and heavy cloud coverage associated with TCs can alter the biophysical properties of reefs and their waters that control bleaching. During the 2005 thermal stress event in the Caribbean, TCs were shown to cool the sea surface enough to mitigate coral bleaching over a larger area than the likely span of wave damage that TCs typically generate. Given that sea surface temperature (SST) has been rising at unprecedented rates in many regions throughout the tropics, a key question to ask from a conservation perspective that has not yet been investigated is therefore: do many coral reefs stand to benefit from TCinduced cooling in an era of increasing thermal stress? This thesis addressed this question by demonstrating that: (1) TC cooling can lower SST sufficiently to either reduce thermal stress levels below meaningful thresholds or slow / prevent thermal stress build-up over a season (e.g., enough to benefit reefs) across broad regions as well as isolated areas; (2) A given TC can benefit (reduce and / or prevent the build-up of thermal stress) a greater area of reef than it harms (via physical damage from TC waves) to produce a ‘net cooling benefit’; (3) Thermal stress relief from TC cooling occurs frequently enough at some reefs to provide a meaningful, though intermittent, refuge from bleaching and the associated sub-lethal effects; and (4) The distribution and timing of such relief varies across the world’s reefs sufficiently such that it provides a basis upon which to rank reefs by conservation priority.

As an initial step, a global analysis over the recent past (1985-2009) across the tropics (extended to 40°N to 40°S) found thermal stress (measured by degree heating weeks - DHW) to be negatively correlated to tropical cyclone activity (measured by frequency of high winds -TC gale days) at reef areas across the entire Caribbean (Chapter 2). While other regions did not show this effect over such a broad spatial (basinwide) and temporal (25 years) scale, increasing the resolution of the analysis yielded similar negative correlations in regions such as Japan, Western Australia and the Great Barrier Reef (GBR).

To explore the potential for a TC to result in a ‘net cooling benefit’ a case study of severe TCs on Australia’s GBR showed that two out of four produced a net cooling benefit, even under the most severe conditions during which the extent of wave damage is maximised (Chapter 3). Thus, further effort to examine patterns of TC cooling at broader spatial and temporal scales was justified.

To that end, a case study in the Caribbean of two extreme seasons, where both thermal stress and TC activity was high, demonstrated that beneficial cooling can occur across broad regions and that the absence of this cooling would have likely resulted in thermal stress of greater severity and duration (Chapter 4). Thermal stress levels were examined in reef areas and, for each season, thermal stress intensity (hotspots) and duration (DHW) was notably lower in areas where accumulated seasonal TC cooling was at least 1°C. This confirmed the findings of the initial global assessment, which used a TC’s footprint (composite of gale radii) as a proxy for cooling, and showed that a global analysis using actual SST data to reconstruct cool wakes is indeed warranted.

Finally, an examination of TC cooling over the same 25 year period as the initial analysis revealed a number of reef regions where cooling likely played a role in reducing the severity and duration of past thermal stress events (Chapter 5). Based on the spatial and temporal patterns of TC-induced cooling, it is clear that some reefs experience beneficial cooling more often than others, providing a basis for differentiating between reefs to target conservation priorities.

This thesis provides the first evidence that TC-induced wake zones may provide refuge from thermal stress to coral reefs over broad spatial and temporal scales, which can ultimately contribute to the overall resilience of coral communities in an era of climate change. The methodology and results presented here can be used as a basis for incorporating the dynamics of TC-induced thermal stress relief into coral reef conservation planning. To that end, it is recommended that future efforts examine these dynamics at finer spatial and temporal scales.

FoR codes (2008)

0405 OCEANOGRAPHY, 040604 Natural Hazards, 0501 ECOLOGICAL APPLICATIONS, 050202 Conservation and Biodiversity, 069902 Global Change Biology, 090903 Geospatial Information Systems



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.