Coral reefs occur in a variety of fringing, barrier, and atoll settings throughout the tropical and subtropical world. Coral reefs constitute important and productive sources of biodiversity; they harbor more than 25% of all known marine fish (Bryant et al., 1998), as well as a total species diversity containing more phyla than rainforests (Sale, 1999). Reefs also represent a significant source of food for many coastal communities (Wilkinson et al., 1999). Coral reefs serve important functions as atoll island foundations, coastal protection structures, and sources of beach sand; they have economic value for tourism (which is increasingly important for many national economies) and support emerging opportunities in biotechnology. Moberg and Folke (1999) have published a comprehensive list of goods and ecological services provided by coral reef ecosystems.
The total areal extent of living coral reefs has been estimated at about 255,000 km2 (Spalding and Grenfell, 1997). As much as 58% (rising locally to >80% in southeast Asia) are considered at risk from human activities, such as industrial development and pollution, tourism and urbanization, agricultural runoff, sewage pollution, increased sedimentation, overfishing, coral mining, and land reclamation (Bryant et al., 1998), as well as predation and disease (e.g., Antonius, 1995; Richardson et al., 1998). In the past these local factors, together with episodic natural events such as storms, were regarded as the primary cause of degradation of coral reefs. Now, Brown (1997), Hoegh-Guldberg (1999), and Wilkinson (1999), for instance, invoke global factors, including global climate change, as a cause of coral reef degradation.
Previous IPCC assessments have concluded that the threat of sea-level rise to coral reefs (as opposed to reef islands) is minor (Bijlsma et al., 1996; Nurse et al., 1998). This conclusion is based on projected rates of global sea-level rise from Warrick et al. (1996) on the order of 2-9 mm yr-1 over the next 100 years. Reef accretion at these rates has not been widely documented, largely because most reefs have been growing horizontally under stable or falling sea levels in recent years (Wilkinson and Buddemeier, 1994). Schlager (1999) reports an approximate upper limit of vertical reef growth during the Holocene of 10 mm yr-1, suggesting that healthy reef flats are able to keep pace with projected sea-level rise. The situation is less clear for the large numbers of degraded reefs in densely populated regions of south and southeast Asia, eastern Africa, and the Caribbean (Bryant et al., 1998), as well as those close to population centers in the Pacific (Zann, 1994).
Positive trends of SST have been recorded in much of the tropical ocean over the past several decades, and SST is projected to rise by 1-2°C by 2100. Many coral reefs occur at or close to temperature tolerance thresholds (Goreau, 1992; Hanaki et al., 1998), and Brown (1997) has argued that steadily rising SST will create progressively more hostile conditions for many reefs. This effect, along with decreased CaCO3 saturation state (as CO2 levels rise), represent two of the most serious threats to reefs in the 21st century (Hoegh-Guldberg, 1999; Kleypas et al., 1999).
Several authors regard an increase in coral bleaching as a likely result of global warming. However, Kushmaro et al. (1998) cite references that indicate it is not yet possible to determine conclusively that bleaching episodes and the consequent damage to reefs are caused by global climate change. Corals bleach (i.e., pale in color) because of physiological shock in response to abrupt changes in temperature, salinity, and turbidity. This paling represents a loss of symbiotic algae, which make essential contributions to coral nutrition and clarification. Bleaching often may be temporary, with corals regaining color once stressful environmental conditions ameliorate. Brown et al. (2000) indicate that some corals in the Indian Ocean, Pacific Ocean, and Caribbean Sea are known to bleach on an annual basis in response to seasonal variations in temperature and irradiance. Major bleaching events can occur when SSTs exceed seasonal maximums by >1°C (Brown et al., 1996). Mortality for small excursions of temperature is variable and, in some cases, apparently depth-related (Phongsuwan, 1998); surviving coral has reduced growth and reproductive capacity. More extensive mortality accompanies temperature anomalies of 3°C or more over several months (Brown and Suharsono, 1990). Hoegh-Guldberg (1999) found that major episodes of coral bleaching over the past 20 years were associated with major El Niño events, when seasonal maximum temperatures were exceeded by at least 1°C.
Corals weakened by other stresses may be more susceptible to bleaching (Glynn, 1996; Brown, 1997), although Goreau (1992) found in Jamaica that anthropogenically stressed areas had lower bleaching frequencies. More frequent and extensive bleaching decreases live coral cover, leading to reduced species diversity (Goreau, 1992; Edinger et al., 1998) and greater susceptibility to other threats (e.g., pathogens and emergent diseases as addressed by Kushmaro et al., 1996, 1998; Aronson et al., 2000). In the short term, this bleaching may set back reef communities to early successional stages characterized by noncalcifying benthos such as algae, soft corals, and sponges (Done, 1999). Reefs affected by coral bleaching may become dominated by physically resilient hemispherical corals because branching corals are more susceptible to elevated SST, leading to a decrease in coral and habitat diversity (Brown and Suharsono, 1990). Differential susceptibility to bleaching among coral taxa has been reported during the large-scale event in 1998 on the Great Barrier Reef (Marshall and Baird, 2000).
The 1998 bleaching event was unprecedented in severity over large areas of the world, especially the Indian Ocean. This event is interpreted by Wilkinson et al. (1999) as ENSO-related and could provide a valuable indicator of the potential effects of global climate change. However, the 1998 intense warming in the western Indian Ocean has been associated with shifts in the Indian dipole rather than ENSO.
Attempts to predict bleaching have met with variable success. Winter et al. (1998) compared a 30-year record of SST at La Parguera, Puerto Rico, with coral bleaching events at the same location but could not forecast coral bleaching frequency from the temperature record. On the other hand, analyses of recent sea temperature anomalies, based on satellite data, have been used to predict the mass coral bleaching extent during 1997-1998 (Hoegh-Guldberg, 1999).
Recently it has been suggested that a doubling of CO2 levels could reduce reef calcification, but this effect is very difficult to predict (Gattuso et al., 1999). Kleypas et al. (1999) argue that such effects could be noticed by 2100 because of the decreased availability of CaCO3 to corals. In combination with potentially more frequent bleaching episodes, reduced calcification could impede a reef's ability to grow vertically in pace with sea-level rise.
The implications for reef-bound coasts in terms of sediment supply, shore protection, and living resources may be complex, either positive or negative, and are difficult to predict at a global scale. However, there have been suggestions that fishing yields will be reduced as reef viability decreases, leading to reduced yields of protein for dependent human populations, and that the effects of reducing the productivity of reef ecosystems on birds and marine mammals are expected to be substantial (Hoegh-Guldberg, 1999).
Other reports in this collection