GCRA  OVERVIEW  FAQ  NEWS  ARTICLES  PHOTOS  REEF ISSUES  RESTORATION  PAPERS  LINKS 

Coral Reef Bleaching In The South Central Pacific During 1994

 Coral Reef Initiative

U.S. Department of State

Washington, D.C.

October 30, 1995

Thomas J. Goreau

President

Global Coral Reef Allianceand

Raymond L. Hayes

Professor

Department of Anatomy

Howard University

College of Medicine

Washington, DC 20059

 NOTE: This report presents detailed descriptions of coral reef bleaching at 19 sites across the South Pacific during 1997. The analysis of 15 environmental variables at each site shows that while many environmental factors that are linked to human population density all have strongly negative effects on live corals, none are related to the amount of bleaching. Only high temperatures are correlated with bleaching. Analysis of satellite sea surface temperatures at sites in the Pacific, Indian Ocean, and Atlantic show clear relationships between high temperatures and coral bleaching. The implications of these findings for environmental policy to protect reefs at national and global scales are outlined. This book, based on a study undertaken for the Coral Reef Initiative of the US State Department, was published by the Global Coral Reef Alliance in 1995. Copies containing large numbers of color photographs and graphs can be ordered from GCRA.

TABLE OF CONTENTS

Foreword

Summary

Introduction

Methods

Results:

A. Site descriptions and reef assessments

1. French Polynesia

a. Tahiti 1

b. Moorea

c. Rangiroa

2. Cook Islands

a. Aitutaki

b. Rarotonga

3. American Samoa

B. Interviews with Other Observers

Statistical Analysis and Discussion

A. Bleaching and local environmental factors

B. Bleaching, sea surface temperature and anomalies

Conclusions

Recommendations

Glossary of terms

Literature cited

Acknowledgments 

FOREWORD

The coral reef bleaching episode which appeared in the central south Pacific region in early 1994 evoked considerable concern among local marine scientists. In response to that concern, this project was conceived. A field survey was necessary to attempt to document the impacts of this event upon the reef communities at several accessible sites within that region.

The International Coral Reef Initiative of the U.S. Department of State approved the travel of two US-based scientists with several years of field experience working on coral reef bleaching in the Caribbean. Their objective in this study was to survey directly several representative coral reefs in two French Polynesian archipelagos, in the southern Cook Islands, and in Samoa. In their travels, however, they were to interview local residents whose experiences over time and into other areas of the region might allow reliable data to be added to the study beyond areas actually visited. This effort would allow them to assemble records from a broad geographical area, to trace back in time the local history of these reefs, and to develop a comparative database on reef vitality. This study was designed to document both acute and long term influences of stresses upon the region's coral reef ecosystems.

Our report is offered as a working resource to marine laboratories, scientific libraries, and individual researchers with interests in regional characterizations of coral reef and associated ecosystems. We expect that the observations and information documented herein will be of value to other marine scientists working in the central south Pacific region as well as in large scale regional studies within the tropical near shore marine environment. Therefore, we are pleased to contribute this report as a practical reference and as a complement to further coral reef investigations.

Thomas J. Goreau and Raymond L. Hayes

December 1994

SUMMARY

A mass coral reef bleaching event was first reported across the South Central Pacific Ocean during March, 1994. Because bleaching may be related to climate change and because of the importance of coral reef protection for sustainable development of tropical marine resources, the Bureau of Oceans and International Environmental and Scientific Affairs of the U. S. Department of State sponsored a rapid field survey of the bleaching event as one of the first action programs funded under the International Coral Reef Initiative.

Nineteen reef sites were surveyed between August 3 and 22, 1994. This study covered sites within an area of 2,500 by 1,000 km and included a variety of reef types in lagoons, bays, and open water conditions in French Polynesia (Tahiti, Moorea and Rangiroa), the Cook Islands (Aitutaki and Rarotonga), and American Samoa (Tutuila). This was the first attempt to compare bleaching intensity and history across the area affected. Previous field investigations by the same team of researchers have been limited to one island or country, so that local impacts could not be readily distinguished from regional stresses.

The types of hard and soft corals present, percent living coral cover, the degree of bleaching, recent coral mortality, calcareous and fleshy algal covers, and effects of hurricanes, storms, sedimentation, eutrophication, fishing practices, crown of thorns starfish, and extreme climatic variability were assessed by visual and photographic surveys. In order to determine the history of changes in the reef ecosystem, interviews were conducted with locally-based professional divers, government officials, environmental agencies, non-governmental organizations, journalists, and scientists. These individuals had first-hand knowledge of the extent, duration, intensity, and start of bleaching, and provided information about other areas which were not examined directly by the authors.

Evidence of coral bleaching was found at all sites examined in August. Since the 1994 event had begun six months previously, bleaching to whiteness was relatively rare; most bleached corals were recovering or had already recovered, but many branching corals were recently dead. Water temperature, salinity, dissolved oxygen, and pH were measured at these sites. All values were in the normal range for the time of year, the cool season in the southern hemisphere. While there were significant differences from place to place in these parameters, the pattern was different for each variable. Coral mortality was seen at all sites, but these patterns were site-specific and did not correlate with bleaching. Reefs in American Samoa were most damaged, followed by lagoonal reefs in Tahiti and Moorea. The highest living coral cover was observed in the southern Cook Islands of Rarotonga and Aitutaki.

The only common factor at all sites had been exposure to abnormally high sea surface temperatures (more than 1 °C above the average value in the warmest months of the year) which affected the region between February and March of 1994. Corals in lagoonal habitats near human populations were less subject to bleaching than those from outer reef slopes. Bleached corals on outer slopes in American Samoa that had been subjected to sediment and pollution stresses were the slowest to recover. The results of this survey indicate that reefs in the South Pacific are highly vulnerable to a 1 °C elevation above average sea surface temperature. Similar patterns also appear to have been present during coral bleaching episodes which took place off the coasts of East Africa and Brazil in 1994. This suggests that corals around the world are near their upper temperature tolerance limits.

MAP 1. The Pacific Ocean. Relative locations of the Society Islands, Tuamotu Islands, Cook Islands and American Samoa (map adapted from Stanley, 1993) 

INTRODUCTION

Coral reef bleaching (please see the Glossary at the end of the volume for detailed explanations of this and other scientific terms used in this report) is caused when environmental stress causes corals to lose the symbiotic inter-cellular algae from which corals derive their color, most of their food, and their ability to rapidly grow their skeletons and form the coral reef structure. Without their microscopic algae, coral tissues turn transparent, revealing the white limestone skeleton beneath. This apparent whitening is called 'bleaching" because it superficially resembles dead corals killed by chemical bleaching agents, but in fact the tissue is still present and the coral is alive. However it is starving, unable to grow its skeleton, and under increased risk of succumbing to other stresses such as algae overgrowth or excessive sedimentation. Death of the corals ultimately results in the erosion the growing wave-resistant coral reef structure on which all other reef creatures rely on for shelter and for food.

Prior to the 1980s all known cases of bleaching were geographically limited in extent and caused by clear local stresses such as restricted circulation (Mayor, 1918; Yonge & Nicholls, 1931; Jaap, 1979) or hurricane floods (Goreau, 1964a). Large scale coral bleaching, affecting huge areas of reef both near to and far away from known stresses, was unknown before the 1980s, but has since become more frequent and intense (Williams et al., 1987; Williams & Bunkeley Williams, 1990; Glynn, 1991, 1993; Jokiel & Coles, 1992; Goreau et al., 1993; Goreau & Hayes, 1994). While most field researchers have attributed mass bleaching since the 1980s to unusually high ocean temperatures, there has been controversy over whether coral reefs ecosystems as a whole are undergoing climate-related stress because a number of local stresses are also potentially capable of contributing (Salvat, 1992).

In early April 1994, Thomas Goreau noticed, while scanning the Climate Diagnosis Bulletin from the National Oceanic and Atmospheric Administration (NOAA), that a warm water mass exceeding 30 °C was displayed on sea surface temperature (SST) maps of the South Central Pacific (Goreau et al., submitted). Having just completed an analysis of SSTs and coral reef bleaching in that region during 1991 (Goreau and Hayes, 1994), he immediately alerted several colleagues by facsimile, including Bernard Salvat, Director of the E.P.H.E. Iaboratory in Moorea, French Polynesia. A response from Salvat indicated that coral reef bleaching had begun in late March in both the Society and Tuamotu Islands of French Polynesia. Dr. Goreau also alerted several academic colleagues, including Raymond Hayes of Howard University, and the Coral Reef Initiative of the U.S. Department of State about this event. The Coral Reef Initiative of the U.S. Department of State is designed to foster partnerships among federal government agencies, state and local governments, and non-governmental organizations. The specific objectives of these partnerships are to promote resource conservation and sustainable development in coral reefs and associated ecosystems as well as to strengthen management of these sensitive and fragile coastal marine environments (U.S. Department of State, 1994). Through domestic and international networking, the Coral Reef Initiative has focused upon research, assessment and monitoring of coral reefs on a global scale (Crosby et al., 1995).

Deputy Assistant Secretary Rafe Pomerance encouraged Hayes and Goreau to submit a research proposal to the Coordinator of the Coral Reef Initiative. In June, that proposal was submitted, recommending that they survey the areas influenced by this event. After review and identification of funds, this project was approved by the State Department and the two scientists went into the field to survey the coral reef bleaching episode from August 3-22, 1994. The major objectives of this research trip were to evaluate the reliability of the NOM satellite sea surface temperature data as a predictor of coral reef bleaching, to determine the current bleaching status of the reefs, and to assess the role of local factors influencing reef health.

METHODS

Only non-destructive methods were used in this underwater survey, mainly still photography, visual estimates of reef cover and stresses, and direct measurements of water quality. Interviews were also conducted with long-term residents familiar with the local near shore marine environment. To maximize coverage of the reef, which is necessary because of its high spatial variability, rapid visual estimates were made throughout the course of as long a swim as possible given constraints of compressed air supplies or wave conditions

Both shore diving and boat diving were utilized, with breath-hold or SCUBA technique depending on depth and accessibility. Line transects were not used because they would have severely limited the area of reef covered (Carleton and Done, 1995). Live coral and algae cover estimates refer to the average value seen on each dive on hard bottom only, and do not include sandy or muddy bottom. Comparisons of our visual cover estimates were made with local divers. Al1 estimated values were either identical or differed by no more than 10%. We are, therefore, confident that our estimates of cover and percentage bleaching are valid within a 10% range of error.

All photographs were made using a Nikonos camera (V or IVB) with a 15mm lens or a 35mm lens with a close-up attachment for macro-photography. Fields were selected for photography to represent the reef profiles or to represent the appearance of bleached coral colonies observed on the reefs. Images with the 15 mm lens were shot from a foreground distance of 5-8 feet; close-up images were taken at a constant distance of 12 inches from the camera lens. Strobe lighting was provided for all photographs using a Nikon 103 unit. All original photographs were shot with Fuji 100 ASA slide film. Unless otherwise indicated in the figure legends as photos by Pete Ely, all pictures used to document this survey were taken by Hayes.

Bleached coral tissue was distinguished from corals which appeared white because of recent predation by Acanthaster planci (Crown-of-Thorns starfish) and from dead coral skeleton, and live tissue was differentiated from filamentous algae overgrowth, which can be similar in color. This was done by close visual examination and whether coral tissue was bleached and intact or dying was determined by fanning the water by hand just above the coral surface to see if it was covered by filamentous algae or had streamers of necrotic tissue after predation.

Timing of coral mortality was estimated from the sequence of algae succession and overgrowth of dead skeleton. Coral which had just died within days had white skeleton without intact tissue. This was apparently colonized by filamentous algae within days to weeks. This either grew into filamentous turf algae communities or cyanobacteria, or was succeeded by fleshy algae. After several months encrusting calcareous red algae, particularly Porolithon, had colonized the projecting tips of coral skeletons which had died early in the 1994 bleaching event. Coral which had died in the 1991 bleaching event were structurally intact, but completely covered by Porolithon, or by fleshy algae in eutrophic lagoonal habitats. The skeletons of coral which had died in earlier bleaching events were significantly bio-eroded by boring clams and worms and scrapers of calcareous algae such as parrotfish, snails, and sea urchins.

Stresses to reefs were estimated using relative values, except for Acanthaster, which was based on counts of the number of individuals seen per dive. Water temperature, salinity, pH, and dissolved oxygen were measured using a calibrated submersible electronic probe. Six small samples (2-3 cc per sample) of coral mucus were taken from the water column as well as from bleached, unbleached and necrotic portions of coral colonies by gentle suction into clean syringes held near the coral tissue. These samples were analyzed by Garriett Smith, a microbial ecologist at the University of South Carolina, for bacterial taxonomic content based upon carbon utilization assays (Jindal et al., 1994, 1995).

Many corals and algae were identified to the species level and many to only the family, or genus level, using primarily Veron's "Corals of the Indo-Pacific", and Magruder and Hunt's "Seaweeds of Hawaii". The very large number of species in certain genera, and the lack of good species level guides for the area, makes identification of species very difficult in genera like Acropora, Montipora, and families like the Faviidae and Fungiidae. Since samples of coral and algae were not taken in this study, and since many species in certain families can only be identified by microscopic examination of dead skeletons by a handful of taxonomists, we believe that identification to genus was the best compromise to allow extensive coverage in a very brief time without destructive sampling.

Some of the sites studied have been collected in much more detail by taxonomic experts, and their reports should be regarded as definitive with regard to species distributions. Most genera are effectively monotypic (only one species occurring in a particular habitat or region), and most genera with many species tended to be bleached in a similar way (our observations and Paulay, personal communication). However a few genera showed a wide range in the bleaching response of different species, and it is clear that more detailed work is required by experts in both coral and symbiotic algae taxonomy before these differences can be fully understood.

Direct field measurements of conductivity (salinity), water temperature, pH, and dissolved oxygen concentration were taken at each site surveyed using a Daigger portable water analysis meter. This battery powered unit was outfitted with a multiple probe assembly suspended from an eight meter waterproof insulated cable. Although this unit had been factory calibrated, it was re-calibrated and field validated against other instrumentation before each application. Digital readings were made at 1 meter and 7 meter depths for each site. As these values were very similar, they were averaged. All data were recorded manually, including date and time of day. These data were also logged automatically in memory for eventual downloading directly into a computer for tabular retrieval and analysis. Routine statistical methods were applied in subsequent analysis of numerical data.

Ocean sea surface temperatures and their anomalies were based on NOAA data largely derived from global satellite advanced very high resolution radiometer (AVHRR) measurements as reported in the Oceanographic Monthly Summaries for 1994, based on the climatology of Reynolds et al. (1993, 1994).

At the 19 South Pacific sites where field work was conducted in August 1994, 15 environmental parameters were evaluated Other environmental factors were also identified at different sites which were not included in this comparative assessment because they usually affected only one or a few sites. These include factors such as dredging, damage to corals from fish collectors, use of dynamite or poisons, factory effluents, diver damage, overfishing or heavy grazing by herbivorous fish. Those factors are discussed in the individual site descriptions section. Physical damage from hurricanes is important in structuring reef communities (Rogers, 1993), but it was not possible to get good before and after assessments of hurricane damage to reefs at many sites from local observers. Although sediment stress to corals (Rogers, 1990) was evident at several sites we did not estimate water turbidity directly because it is so extremely variable that long time series of measurements are needed to establish significant differences. The factors which were assessed statistically for all of the sites are as follows:

1. Habitats were classified according to degree of protection from storm waves. A value of (1 ) was assigned to all exposed fore reef sites, (2) to semi enclosed bays, (3) to lagoon and reef crest habitats. This parameter is a proxy for average wave energy.

2. Water temperature measurements were made with a recording meter during field work in August. This period corresponded to near the coolest time of year at all sites.

3. Salinity was measured by a recording conductance sensor as total density anomaly in parts per thousand. The period of measurement corresponded to the dry season at all sites. Tutuila had the highest rainfall and river runoff into the coastal zone, followed by Rarotonga and Tahiti.

4. Dissolved oxygen was measured by calibrated recording meter. Values would also be expected to be seasonal, dependent on temperature, algae photosynthesis, reef respiration, and the rate at which oxygen consuming pollution and sewage contributes to the coastal zone.

5. pH was measured by calibrated recording meter.

6. Live coral cover was estimated visually to the nearest 10%. These estimates were confirmed by review of wide-angle lens photographs. All bottom cover estimates refer only to hard bottom, and do not include sand and mud bottom.

7. Proportion of corals bleached was estimated visually over the course of each dive. These were confirmed through wide angle photographs.

8. Fleshy algae cover at each site was estimated visually.

9. Calcareous algae cover was estimated visually.

10. Human population density in the watersheds inland from each site was ranked on a relative scale. The lowest values refer to areas with no shore population, and the highest densities are for the capital cities of Papeete, Tahiti, and Pago Pago, Tutuila. This parameter serves as a proxy for census data, which was not available at all sites at the required scale.

11. Mud cover of corals was estimated visually by the two scientific divers and through wide-angle photographs.

12. Acanthaster planci (Crown-of-Thorns starfish) seen were counted on each dive. At only one site were there too many to count, and the value was estimated.

13. Years since last cyclone or major storm was based on interviews with local divers, scientists or government officials.

14. Previous bleaching events were based on local interviews.

15. Pollution was ranked on a relative comparative scale, based on the amount of debris (bottles, paper and plastic items, etc.) and artifacts (junked construction materials, munitions, etc.) seen.

Statistical regressions were run between all pairs of variables. Factors were compared using non-parametric statistics. Some measurements were measures of the status of the entire ecosystem (such as the bottom cover), but other factors vary strongly on a seasonal basis and were probably different at the time when bleaching began (such as measurements of water temperature, salinity, and oxygen content) than when they were measured in August. Some factors were estimated only on a comparative, rather than a quantitative basis (for example habitat was represented by relative wave exposure rather than wave height measurements, and population by relative density ranks rather than from census figures). Since many assessments were based on comparative rankings we have contrasted them using qualitative rank order statistics rather than parametrically. The Spearman correlation coefficients were calculated between all pairs of variables. This examines the degree to which the 15 factors are ranked by their relative magnitudes in the same order at all 19 sites. This method has been previously used to analyze 13 environmental factors at 11 coral reef sites in Jamaica (Goreau 1992).

Monthly mean sea surface temperatures and their anomalies were taken from maps published in the Oceanographic Monthly Summary by the National Oceanic and Atmospheric Administration, the National Ocean Service, the National Weather Service, and the National Environmental Satellite, Data, and Information Service, Ocean Products Branch, of the United States Department of Commerce. This data set provides a consistent data base with complete global coverage, based on a combination of satellite-derived measurements and those from oceanographic vessels.

In this report we have analyzed all available data from January, 1993, through September, 1994. Twenty-two sites were evaluated in the Pacific Ocean, six in the Indian Ocean, and five in the South Atlantic (see Maps 8 and 9; also see Table 7A and 7B for geographical coordinates). Values at each site were determined by preparing an overlay transparency on which all sites were marked, superimposing it on the published maps and reading the values right below marks. Values were interpolated to the nearest 0.1 degrees Celsius from the slope of marked contours. This generally assumes that slopes are linear between contours, and it is estimated that non-linear contour slopes could provide inaccuracies of the order of plus or minus 0.1 degrees Celsius in most cases. However, where there are very sharp spatial temperature gradients, contours are close together and precision errors could be larger, up to 0.3 degrees Celsius in extreme cases. Sites near near warm or cold currents and western boundary current areas are especially likely to be affected. A different source of uncertainty occurs where temperature gradients are flat in the core of the warmest regions. The maximum values are not given, and thermal contours within these areas have to be estimated from those outside. Generally conservative assumptions were made, so that the values given are likely to be under-estimates of the true values. This problem affected only a few sites very near the equator.

MAP 2. Tahiti, Society Islands.

The Aquarium (lagoon) to the north and The Gap (outer slope) to the south were the two sites surveyed (map adapted from Stanley , 1993)

SITE DESCRIPTIONS AND REEF ASSESSMENTS:

TAHITI, SOCIETY ISLANDS, FRENCH POLYNESIA

Background: Bleaching was first reported in March, 1994 (Henri Poliquen, personal communication). Almost all corals were affected, but the majority had either recovered, as indicated by tissue pigmentation, or were dead by August. Mortality of all lagoon corals had occurred after bleaching in 1984, which eliminated all dominant large plate-shaped Acroporas. Some bleaching took place in 1987, but before that time there had been recovery from the previous event. High mortality and tissue fluorescence of small branching Acroporas were seen after bleaching in 1991. A very minor bleaching event took place in 1993 (Drollet et al., 1994). Tahiti was in the direct path of several hurricanes in the 1980s, with the most recent in 1987.

1. Lagoon, "The Aquarium", northwest Tahiti off-shore from the end of the Faa'a airport runway (August 6 1994)

a. Physical setting: Patch reefs surrounded by sand in depths of 3 to 5 meters.

b. Corals and bleaching: The majority of corals were dead, and those live corals (Porifes lobata, Porites lutea, Pocillopora verrucosa, Pocillopora eydouxi, Pocillopora damicornis, Acroporas, brown and purple encrusting Montiporas, and Pavona) were physically overgrown by algae. Also, several bleached and recovering anemones (Heteractis) were observed (Figures 1-3). Synarea (Porites rus) colonies up to 10 meters across were in good condition. Large thicket-like colonies of one branching Acropora species (possibly A. robusta) up to 10 meters across were recovering from bleaching and were growing at the rate of 2-3 cm per month (Poliquen, personal communication).

c. Algae: Turbinaria ornata, Sargassum, and 1 cm high branching brown turf algae were growing on the tops of dead coral heads. Turf algae predominate, including Sphacelaria. Some Chaetomorpha were seen, but all were cryptic due to grazing. Algal cover was approximately four times that of stony coral cover.

d. Local reef stresses: High density of the grazing sea urchin Diadema, but they appear to have little impact on algae because no bare grazing halo is visible around the coral heads where they are found in dense swarms. High eutrophication was indicated by overgrowth of corals by fleshy algae (Figure 2). There was evidence of recently suspended sediment on living coral. The lagoon is subjected to high human population-dependent stresses, including shipping activity, industrial waste and human sewage, in the Fa'aa area and the western suburbs of the capital, Papeete (population of around 75,000).

Figure 1. Site #1. "The Aquarium" in lagoon of Tahiti. Extensive patch of new growth of Acropora in recovery from recent bleaching (1994). The living coral is emerging from algal overgrowth on the dead coral frame. Depth 2-3 meters.

Figure 2. Site #1. 'The Aquarium'' in lagoon of Tahiti. Algal overgrowth on bleached and dying (center) and old dead (left) Acropora in lagoon. Depth 3-5 meters.

Figure 3. Site #1. ''The Aquarium'' in lagoon of Tahiti. Bleached anemones (Heteractis) with clownfishes in shallow lagoon site. Depth 3-5 meters.

2. Outer reef slope, "The Gap", near Tuputu Pass, off Punaauia, west Tahiti (August 6 1994)

a. Physical setting: Steep, nearly vertical wall with a slope 75° to 90° was examined from 3 to 26 meters depth.

b. Corals and bleaching: Live coral cover averages 70%, with values up to 80-90% in deeper water below (16 meters) and less than 50% in the wave breaking zone (upper 5 meters). Pocillopora was dominant, followed by Acropora, and Montipora. All corals are small except for some shelf-like Montiporas on edges of canyons, up to 1-2 meters across. Most corals appear to be only a few years old, and may have settled since the 1991 bleaching event. Around 10% of live corals showed bleaching, with very few colonies bleached to complete whiteness. Almost all bleached corals were partially bleached and in a recovery phase. Bleaching was observed in Acropora, Pocillopora, Fungia, Pavona, Montipora, and Heteractis anemones (with clownfish). Coral fluorescence, when observed, was confined to the oral disk of normally colored Acropora. Around 10% of corals were recently dead and covered with filamentous algae; almost all dead corals were of the genera Acropora and Pocillopora (Figures 4-5)

c. Algae: Encrusting red algae (mostly Porolithon) cover about 20%, with parrotfish grazing bite marks. Branching algae ranged between 1% and 5%. these algae were primarily wave resistant Turbinaria ornata (shallower than 6 meters). There are few Halimeda, and all are cryptic (in crevices protected from parrotfish grazing). Blue-green algal growth (probably Lyngbya) and Dictyota grew on recently dead and physically undamaged corals which had died after bleaching during the previous months. Corals which had died from bleaching in 1991 were covered by Porolithon.

d. Local reef stresses: Intense grazing by abundant surgeonfish and damselfish was observed. The area does not appear to be affected by pollution or sediments and the steep slopes keep it well exposed to open ocean waters. The major cause of coral mortality since hurricanes in the mid-1980s appears to be bleaching, since corals are dead and overgrown in place, without physical damage. Diving activities nearby use moorings and are carefully controlled with no evidence of damage due to anchors or diver contacts.

MAP 3. Moorea, Society Islands. Lagoon and outer slope survey sites on the north coast near Tiahura motu (map adapted from Stanley, 1993).

MOOREA, SOCIETY ISLANDS, FRENCH POLYNESIA

Background: Previous mass coral reef bleaching had occurred in 1984 and 1987. In 1991, bleaching occurred again in Moorea and was studied by Salvat (1991), Gleason (1993), and Goreau and Hayes (unpublished observations). That episode was remarkable because of the fluorescence of the bleached corals (Jardin et al., in preparation). A few cases of bleaching were seen in 1993.

Bleaching began in March, 1994, according to Bernard Salvat, Catherine Jardin, and Yannick Chancerelle. The reports of Houegh-Guldberg (1994) and Houegh-Guldberg and Salvat (1995) are based upon transect analyses during April, 1994. Most corals were affected, but they had already either recovered or died by August. Hurricane history for the area is identical to the neighboring island of Tahiti, as are ocean temperatures. However Moorea has a far smaller population.

Figure 4. Site #2. ''The Gap'' on outer slope of Tahiti. One completely bleached colony of Acropora surrounded by recovered assortment of Pocillopora and Acropora corals. Depth 20 meters.

Figure 5. Site #2. "The Gap" on outer slope of Tahiti. Partially bleached Acropora and Pocillopora corals in recovery from bleaching of 1994. Depth 20 meters.

3. Lagoon, "Barrier reef", on inside edge of reef flat, 200 meters southwest of Tiahura pass, near site of Salvat thermograph (August 7 1994)

a. Physical setting: Patch reefs were interspersed with sand bottom cover at about 2 meters of water depth. Long-term surveys of species composition and bleaching have been undertaken near this site by Salvat since 1991, along with water temperature measurements.

b. Corals and bleaching: The area is dominated by 2 to 3 meter diameter heads of Porites lobata, P. Iutea, and P. rus (Synarea), with about 20% Pocillopora verrucosa and very few Acropora. Very little bleaching was seen, with only a few small pale patches on Porites heads (Figures 6-73. The soft coral, Sarcophyton, has become very abundant, covering between a fourth to a half as much bottom as hard corals. Soft corals were virtually absent in 1991, with only one colony being seen on the back reef and two on the fore reef.

c. Algae: Algal cover and species diversity have increased greatly in the last three years. In 1991 only Turbinaria ornata and Caulerpa racemosa were common

in the back reef, with Chaetomorpha and blue-green algal mats confined to the area immediately in front of Club Med where sewage was soaking through the beach sand. In 1994 there was extensive algal overgrowth on the top of most large Porites and Synarea heads in the lagoon. Algae cover as much of the hard bottom as corals. Common algal species seen in 1994 were Turbinaria ornata, Sargassum (vast increase, almost absent in 1991), Chlorodesmis, Enteromorpha cryptic but common, Padina, Dictyota, large mats of blue green algae of various colors, fine and coarse grained Halimeda and possibly Siphonocladus. No Caulerpa was seen in 1994.

d. Local reef stresses: A rapid increase of fleshy algae and soft corals has taken place since 1991, indicating that the north lagoon is undergoing rapid eutrophication, although not yet as bad as in the Tahiti lagoon. This is thought due to rising nutrients associated with increasing sewage inputs from rapidly expanding tourism and shore hotels. Additional stress from resuspended sediments are due to high motorboat traffic, largely tourism-related. There is no industry on Moorea except a pineapple juice canning plant.

4. Outer reef slope, north Moorea, 200 meters east of Tiahura pass, near site previously studied in 1991 (August 7 1994)

a. Physical setting: Reef slopes gently at about 30° from the surface to 20 meters. It was examined between 10 and 20 meter depths. Low buttresses (around 1 meter high) exist with sand channels in between. This site is near one which was examined by Hoegh-Guldberg in April, one which has been studied continuously by Salvat since the previous bleaching event in 1991, and one at which Acropora survivorship and fluorescence were studied by Jardin et al. in 1991 (in prep.).

b. Corals and bleaching: Live coral cover was about 60%. Pocillopora was dominant, about twice as common as Acropora species. P. verrucosa was roughly 2 to 3 times more common than P. eydouxi. About 20-30% of all live coral was bleached, almost all partially, with few colonies showing full bleaching. About 80% of live Pocillopora was normal, about 20% was partially bleached, and an additional 10% was dead. About 20% of Acropora were normal in color, about 60% were partially bleached and were recovering from their bases upwards to their tips. About 20% were completely bleached to yellow or white, and about 20% were dead. Mortality and algal overgrowth were most pronounced on bushy species (poss. A. subglabra) and on very thin knobby yellow species (pose. A. aculeus). Most Acropora were in the A. humilis, A. aspera, A. nobilis, or A. formosa groups and were at various stages of recovering. More than 80% of Fungias and Montastrea curta were partially bleached. About 20% of Montipora were partially bleached. Some partial bleaching was seen in green encrusting Plesiastrea. Very little bleaching was evident in Porites. Overall hard coral bleaching was roughly 2 to 3 times more abundant than on the fore reef in Tahiti. This feature might be due to the shallower slope and higher light intensity on the bottom. Almost all coral colonies were small, less than 50 cm diameter, and probably had settled since the 1991 bleaching event. Dead coral cover ranged from 10-20%, made up mostly of Acropora and Pocillopora. Bare rock and rubble constituted about 30%, mostly covered by Porolithon.

c. Algae: Porolithon was encrusting around 20% of bottom. Around 10% fleshy algae, largely Turbinaria, Sargassum, and fuzzy brown turf algae are present. Algal overgrowth on corals is mostly fine Dictyota. Two kinds of Halimeda (fine and large plates) are seen, but all are cryptic due to grazing, with exposed portions grazed flat.

d. Local reef stresses: Coral cover was fairly low, and algal cover and algae species diversity were higher than in 1991. Only Turbinaria ornata was common in 1991. There is a high incidence of dead coral and rubble. Corals which died in 1991 are covered with Porolithon; in contrast, dead corals from 1994 are covered with fleshy algae and filamentous algae turf. The last hurricane hit the area in 1987. There is a high level of anchor and diver damage here, due to the large number of sport divers from hotels on the northwest of the island.

MAP 4. Rangiroa, Tuamotu Islands. Sites surveyed were Avatoru Pass, Tivaru (outer slope), and a patch reef within the lagoon (map adapted from Stanley, 1993).

RANGIROA, TUAMOTU ISLANDS, FRENCH POLYNESIA

Background: Bleaching began in March 1994 (Yves Lefevre, personal communication) and this was the worst episode seen here in ten years. Bleaching occurred in 1991; mild bleaching occurred in both 1992 and 1993. Many coral species have been affected, with the order of severity being greatest in Acropora, intermediate in Pocillopora, and lower in Porites, which bleached only in 1994. Heferactis anemones have been bleached during past bleaching episodes (Lefevre, personal communication). The last hurricane took place in 1983, but very strong wave activity and coral damage have occurred along exposed shores following rough weather in 1991 and 1993. Mortality from bleaching is most obvious in areas which suffered heavy breakage from waves in the 1993 storms.

Figure 6. Site # 3. Lagoon on north side of Moorea. Bleached Pocillopora showing recovery from bleaching. Depth 3-5 meters.

Figure 7. Site # 3. Lagoon on north side of Moorea. Close-up of Porites with recovering, bleached and dying zones. Dead coral is overgrown by algae (right). Depth 3-5 meters.

5. Interior of northwest quadrant of Rangiroa lagoon (August 9 1994)

a. Physical setting: This patch reef of 100 meters across is surrounded by deep water in the interior of the lagoon and was several kilometers removed from uninhabited atoll islets. High wave energy was indicated by pounding surf and abundant rubble, which was cemented together by encrusting red algae and beach rock below 7-10 meter depths. Rangiroa is the world's second largest atoll, and is so large that the lagoon is extremely well mixed by wind and waves, with conditions inside similar to the open ocean. Well developed reefs fringe both the inside and outside of the lagoon, and coral species found are similar on both sides

b. Corals and bleaching: Living coral cover was about 50%, with a lot of rubble. Acropora is the most common genus, containing three species (one with rounded heads, one with bushy heads, one with flat heads) up to 1 meter across. Pocillopora and Porites are subdominant and roughly equal in numbers, with the latter 1 to 2 meters in diameter. All Acroporas showed partial bleaching, but only a minority of Pocillopora and Porifes. Brain corals (Leptoria) were uncommon and also partially bleached. Less than 10% of Acropora colonies were recently dead or dying. (Figure 8)

c. Algae: Porolithon covered most hard ground and rubble. Halimeda was present, but all were cryptic and grazed. No fleshy algae were seen.

d. Local reef stresses: Physical damage from breaking waves was observed.

6. Outer reef slope, near Tivaru Pass, west end of Rangiroa atoll (August 9 1994)

a. Physical setting: The reefs face open ocean along the outer slope of the atoll. The slope was about 15° out to 10-12 meter depth. There are no inhabitants within several hours by boat, but the area is occasionally visited by fishermen. Tivaru Pass is entirely filled by sand and rubble, and water flows through it only during hurricanes.

b. Corals and bleaching: Live coral cover was about 50%. Pocillopora made up half of all coral colonies, Acropora 30%, and others (mainly Porites, Montipora, Montastrea curta, Favia, and Millepora) make up 20%. About 90% of all Acropora colonies were partially bleached. This included all of one bushy species, which were recovering brown color from the bases. Two of the least common species (one corymbose, one with large rounded polyp tips) were entirely unaffected. Less than 10% were recently dead. About 50% of Montipora colonies were partially bleached, and others were completely bleached. 10 to 20% of Pocillopora verrucosa were bleached, most partially bleached to pale colors, but about 1% bleached white. Coral cover was highest in shallow fringing reefs from 1 to 5 meters depth, and below this depth largely consisted of coral rubble and small flattened colonies. (Figures 9-10)

c. Algae: Porolithon encrusted limestone cover was roughly 20%. Halimeda of 2 or 3 species covered about 10% of the bottom, forming 5 cm high bushes which did not appear to be grazed. Fleshy algae were not seen.

d. Local reef stresses: Wave damage is evident from previous hurricanes and storms. Parroffish herbivory is clearly very low based on lack of Halimeda grazing and bite marks on Porites. This could be due to low fishing pressure and high abundance of predatory fish. Both blacktip and whitetip sharks were seen, and sharks (and manta rays) are so abundant in Rangiroa to be a major attraction to sport divers. Small black encrusting sponges were a unique feature of the site.

Figure 8. Site #4. Patch reef in lagoon of Rangiroa. Bleached Acropora on small patch reef in middle of lagoon of the atoll. Depth 4-7 meters.

Figure 9. Site #5. Tivaru pass, outer reef slope of Rangiroa. Bleached Acropora surrounded by clusters of calcareous algae, Halimeda. Depth 10 meters.

Figure 10. Site #5. Tivaru pass, outer reef slope of Rangiroa. Bleached Acropora and soft coral on a spur formation in outer reef slope. Depth 10 meters.

Figure 11. Site # 6. Avatoru pass in Rangiroa. Bleached Pocillopora and Acropora attached to coral rock along edge of pass. Depth 4-5 meters.

Figure 12. Site # 6. Avatoru pass in Rangiroa. Bleached and recovering zoanthid soft coral, Palythoa in pass. Depth 5 meters.

7. West side of Avatoru Pass, north Rangiroa (August 9 1994)

a. Physical setting: A strong current flows northward out of the atoll in response to the prevailing trade winds, which had been blowing with unusual force from the southeast for three days before, eroding up to 6 meters from beaches along the interior of the atoll. These conditions, known as "maurua", occur around 10 days a year, in the coolest season. High wave energy was also indicated by complete absence of sand in the cobble to boulder-sized outer beaches, primarily composed of hurricane tossed Pocillopora corals, and protected by beach rock. The wall of the pass was examined between the surface and the bottom at 7-10 meters along most of its length.

b. Corals and bleaching: Live hard coral cover was about 50%. The predominant species was Pocillopora verrucosa up to half a meter to a meter in diameter, with roughly equivalent area occupied by large Porites heads up to 1 or 2 meters in diameter with a very high density of parrotfish bite marks. Neither species was bleached. Only one Acropora was seen, and it was partially bleached. Alcyonarian soft corals (Lobophytum or Sinularia) up to 1 meter across were abundant, up to 20% cover. Around 10% was made up of the zoanthid soft coral Palythoa, which was partially bleached and fully contracted. (Figures 11-12)

c. Algae: Porolithon covered most limestone buttresses and coral rubble. Halimeda was present in small amounts, but only in cryptic habitats with projecting growth bitten flat by parrotfish. No fleshy algae or turf algae were seen.

d. Local reef stresses: High currents and high wave energy are obvious at this site. Possible oil pollution by ships and boats using the pass to get to the atoll's main dock constitute additional stress to the reef.

MAP 5. Aitutaki, Cook Islands. Two outer slope sites on west side of the atoll were surveyed: Shipwreck and Fish Spot near Maina motu. Lagoon site near the airport was also surveyed (map adapted from Stanley, 1993).

AITUTAKI, COOK ISLANDS

Background: Bleaching began in Aitutaki in March of 1994 (Neil Mitchell, personal communication). The only major bleaching event in the previous 10 years was in 1991. That event was much more severe than 1994, water temperatures reached 31 degrees Celsius, and all bleached corals died. Water conditions were very hot and calm in 1991, and Crown-of-Thorns starfish were not responsible. A major Crown-of-Thorns infestation took place in 1982 and 1983, and swarms were still working their way around the island, leaving patchy damaged trails behind. Other stresses to reefs were coral death in the lagoon in 1987 as a result of exposure and stagnant water due to exceptionally low sea levels, possible minor bleaching in 1986, 1987, and 1989, and wave damage from storms in the early 1990's. The most recent devastating hurricane, Cyclone Sally, hit Aitutaki on December 25 (Christmas Day) of 1986, but no bleaching followed it. Detailed line transect surveys of coral species abundance were conducted around Aitutaki around the time of bleaching by Dr. Gustaf Paulay and the staff of the Cook Islands Conservation Service (in preparation).

8. Lagoon off the northern tip of the island, (August 12 1994)

a. Physical setting: The lagoon was very shallow in this area, around 1 meter deep, and was around 80% Halimeda sand-covered hard limestone bottom, with large coral heads which have grown up to sea level and died on top due to air exposure, but have continued to grow outward to form "micro-atolls". There was an extremely strong current driven by waves pounding over the reef flat by the prevailing winds. Large numbers of blue Linckia starfish and sea cucumbers were in the sandy areas.

Figure 13. Site #9. Shipwreck site and adjacent "Fish Spot" site on extreme southwestern outer slope reef of Aitutaki, Cook Islands. Panoramic reef scene showing coral cover and a single bleached Acropora in foreground. Depth 20 meters.

Figure 14. Site #9. Shipwreck and "Fish Spot" on outer slope of Aitutaki. Edge of shallow reef zone with assorted Acropora showing partial bleaching. Depth 10 meters.

Figure 15. Site #9. Shipwreck and 'Fish Spot" on outer slope of Aitutaki. Upper segment of reef with Acropora of several species, small Pocillopora and Pavona. Depth 10 meters.

Figure 16. Site #9. Shipwreck and "Fish Spot" on outer slope of Aitutaki. Partially bleached Acropora showing recovery from base to tips. Depth 12-15 meters.

Figure 17. Site #9. Shipwreck and ''Fish Spot" on outer slope reef of Aitutaki. Encrusting Psammocora surrounded by Acropora and Pocillopora (Photo by Ely).

Figure 18. Site #10. Mainatai Point on outer slope of Aitutaki. Panoramic profile of reef indicating numerous small Acropora corals amidst coral rubble and a few other coral genera. (Photo by Ely)

b. Corals and bleaching: Large Porites heads formed micro-atolls up to 10 meters across, covering around 20% of the bottom. Many of the larger ones were fused together if sufficiently near. Around 40% consisted of live coral, with flat tops algae covered. A few Acropora species and Millepora dichotoma were found. Less than 5% of corals showed bleaching. However higher levels of bleaching, up to a third of the iive corals, had occurred in Acropora stands in lagoon sites in somewhat deeper and more protected waters around half a kilometer south of this site (Flora, personal communication), and other lagoon sites had complete mortality of Millepora and 80% mortality of Acropora (Paulay, personal communication). More detailed results will be published from his transect data at sites around the island (Paulay, in preparation). Coral diversity in the lagoon was observed by Mitchell to have been much greater before exposure and stagnant conditions in 1987 and before coral mortality from bleach in 1991.

c. Algae: Algal cover was high, consisting around 50% Halimeda opuntia and around 10% Turbinaria ornata on the dead areas covering the top of microatoll coral heads.

d. Local reef stresses: This area is protected from all but hurricanes by the reef flat, but the last hurricane had broken an enormous amount of coral rubble which formed a steep rubble beach. The shallow reef lagoon is extensively walked on by Aitutakians during reef gleaning activities. Those observed included group gill netting using sticks to beat the water to drive fish and use of crowbars to turn over rocks and shovels to dig clams out of rubble and sand. Fish poisons, derived from traditional plant materials, are also used, especially "ore papua" derived from the Barringtonia asiatica tree seed pods, Tephrosia purpurea, or Derris trifoliata (Whistler, 1992)

Giant clams have been overharvested, and the Cook Islands Fisheries Service has instituted a seed farm to restock the area. From 1970 until the market collapsed in the mid 1980's, most of Aitutaki was cultivated with banana for export to New Zealand, and the lagoon may have been subjected to higher levels of fertilizer, chemicals, and soil erosion at that time.

There is an old tradition of quarrying live Porites heads from the lagoon to be burned for quicklime, and the oldest houses are made from coral cemented with lime. However such houses have not been built for a long time since concrete became available, and the old communal lime kilns closed by the 1950's or 1960's. Only small scale lime production now takes place for whitewashing houses.

9. Outer reef west of southernmost point of Aitutaki island, near the shipwreck at the mid point of the outer atoll fringing reef (August 11 1 994)

a. Physical setting: Reef flat slopes moderately steeply, 45° to 60° with strong vertical zonation, and has numerous canyons and buttresses.

b. Corals and bleaching: The upper 10-20 feet have about 50% live coral cover with large amounts of red calcareous algae-encrusted bare rock and dead coral. Corals in this area are all small, largely less than 50 cm across, and appear to be recent recruits which settled after the last hurricane had made a near complete sweep of previous corals. Live coral cover increases sharply with depth and is 60-70% between 40 to 80 feet and deeper. The species diversity and evenness were very high, but Acropora was the most common genus, with around twice as many species as French Polynesia. Full bleaching was relatively uncommon, less than 5% of corals, but around 20% were partially bleached. Species observed to be partially bleached were many Acropora species including A. humilis, digitifolia, verweyi, tenuis, hyacinthus, and numerous others; Pocillopora verrucosa, eydouxi, and meandrina; perhaps 6 Montipora species, Astreopora, Fungias, Lobophyllia, Favias, Goniastrea, Leptoria, Montastrea, Merulina, and others. See Figures 8-13. Detailed species surveys were made nearby by Paulay and the staff of the Cook Island Conservation Service. (Figures 13-15)

c. Algae: The upper 3-7 meters are about 50% composed of lilac and pink Porolithon covered reef framework and dead corals, and below this it covers about 30%. Halimeda was common, but everywhere cryptic. Other algae were very rare, and some blue green algae were seen.

d. Local reef stresses: There has been severe coral mortality in the past from coral bleaching and from wave stress by hurricanes and tropical storms. The are just north of the site had very heavy ongoing mortality from an Acanthaster planci swarm, and around 50 Crown-of-thorns starfish were seen on the dive. Acanthaster predation is easily distinguished from bleaching because it leaves corals with large rounded white patches, having a sharp edge between dead skeleton with wisps of necrotic tissue and the completely normal-colored coral tissue beyond range of the starfish stomach which is extruded through the mouth. In contrast, bleached corals have intact tissue with a gradation in colors. Bottom cover estimates at this site refer to the area which had not yet been reached by the southward-migrating Acanthaster swarm, which did not appear to discriminate between bleached and normal corals.

10. Outer reef slope at the southwest corner of the Atoll (August 11 1994), Mainatai Point.

a. Physical setting: The habitat is very similar to the previous site, sloping around 45° to 60°, but is more remote from land. Abundance of fish was very high, especially around canyons and crevices in the reef.

b. Corals and bleaching: Coral species found were very similar to the previous site, as was the vertical zonation: young corals predominated in shallower areas with larger corals deeper down. However while the live coral was also 50% in shallow water, it became even higher with depth, being more than 80% from 10-30 meters and deeper. While full bleaching was less than 5%, only around 10% of the corals were partially bleached. Not only was bleaching less marked at this site than the previous one, most Acroporas at this site showed no signs of bleaching, and were distinguished by extremely healthy colors, with exceptionally strong natural light fluorescence. See Figures 14-31. Other sites along the western coast of the atoll are pictured in Figures 32-44. These resemble in composition the reef as described at Mainatai. (Figures 16-24)

c. Algae: Algae were similar in type to the previous site but less abundant as a result of the higher coral cover. Porolithon decreased to around 10 to 20% below 10 meters. It appears that this site had been a little more protected from the hurricane which had reduced coral cover in shallower areas.

d. Local reef stresses: This reef, while subject to similar general historical conditions as the previous site, appeared to be in better condition in that coral cover was higher and bleaching lower. The area is near the corner of the semiatoll most remote from human influence. While the previous site was infested with crown of thorns, only one small juvenile was seen at this site.

Figure 19. Site #10. Mainatai Point on outer slope of Aitutaki. Panorama showing dead and bleached Acropora, Favites and Millepora. Depth 10 meters.

Figure 20. Site #10. Mainatai Point on outer slope of Aitutaki. Dead and partially bleached Acropora. Depth 15 meters.

Figure 21. Site #10. Mainatai Point on outer slope of Aitutaki. Normal Pocillopora beside cluster of bleached, partially bleached and dead Acropora. Lobophyllia, Montipora and Pavona also included in field. Depth 20 meters.

Figure 22. Site #10. Mainatai Point on outer slope of Aitutaki. Acanthaster damage to Acropora, and assorted corals showing partial bleaching. Depth 20 meters.

Figure 23. Site #10. Mainatai Point on outer slope of Aitutaki. Bleached soft coral species (center) surrounded by assortment of small corals showing normal to bleached color patterns. Depth 15 meters.

Figure 24. Site #10. Mainatai Point on outer slope of Aitutaki. Dead and partially bleached Acropora surrounded by Lobophyllia and other corals. Depth 15 meters.

Figure 25. Site #10. Mainatai Point on outer slope of Aitutaki. View of outer reef slope showing bleached and dead Acropora (center) with bleached Pavona and partially bleached Pachyseris. Depth 15 m.

Figure 26. Site #10. Mainatai Point on outer slope of Aitutaki. Edge of outer slope reef showing bleached Montipora and Acropora. Depth 10 meters.

Figure 27. Site #10. Mainatai Point on outer slope of Aitutaki. Assortment of corals of various genera at base of reef, showing a wide range of colors. Dead Pocillopora overgrown by red calcareous algae also seen. (Photo by Ely).

Figure 28. Site #10. Mainatai Point on outer slope of Aitutaki. Denuded skeleton in wake of Acanthaster (Crown-of-Thorns starfish). (Photo bv EIy).

Figure 29. "The Ridge" on outer slope of Aitutaki. Colony of Pocillopora eydouxi on reef slope. (Photo by Ely).

Figure 30. 'The Ridge'' on outer slope of Aitutaki. Acanthaster grazing over coral with encrusting red algae and coralline algal orange overgrowth. (Photo by Ely).

Figure 31. Atuatane outer stope on Aitutaki. Dead and algal-encrusted corals showing the calcareous algae Halimeda (Photo by Ely).

Figure 32. Maungapu outer slope on Aitutaki. Brown algal growth on dead coral rock. Colony of Coeloseris (Photo by Ely).

MAP 6. Rarotonga, Cook Islands. Black Rock and Pinnacle sites on the outer slope along the west side of the island and the lagoon at Black Rock were surveyed (map adapted from Stanley, 1993).

RAROTONGA, COOK ISLANDS

Background: Coral reef bleaching began in March of 1994 (Greg Wilson, personal communication) and was most noticeable in Pocillopora. This bleaching was not related to the presence of Acanthasfer, physical damage or pollution. Bleaching was highest on the outer slope northeast Rarotonga reefs. Bleaching first was seen in 1991, and had not occurred in the previous 20 years. Massive bleaching in 1991 followed long hot calm periods with low sea level, and only affected corals in the lagoon, not on the outside reefs. Yellow Milleporas were first to bleach, but recovered within weeks. Almost all corals were affected, and mortality was very high. Hot doldrum conditions with no wind, low tides, and poor circulation caused the whole lagoon and reef flat to stink from dead coral. Very good lagoon reef recovery afterwards resulted in higher diversity of corals. 1994 bleaching was much milder, and mainly affected outer slope reefs. A major crown of thorns proliferation took place in the 1970's, causing the government to force everyone to collect them all around the island. On one day huge numbers were killed and buried in pits, and the problem had been minor since. The last devastating hurricane, cyclone Sally, hit Rarotonga from the north on New Years Day, 1987, after passing over Aitutaki, but with less force.

11. Outer reef slope off Black Rock at the northwest of Rarotonga (August 13 1994)

a. Physical setting: Reefs have a medium steep slope, 45° to 60°, but considerable topographic complexity, with many pinnacles and canyons fringed with shelf corals.

b. Corals and bleaching: Coral cover was very high, around 90%, down to more than 30 meters, but with some dead Acropora. No full bleaching was seen, but 20% of colonies were partially bleached. Coral species cover, species diversity, and evenness were the greatest seen, higher than Aitutaki. The same genera were seen, but more species seemed present. Partial bleaching was most noticeable in Pocillopora verrucosa, with around half of the colonies affected. Colonies were large, up to 1 to 2 meters in diameter, but many colonies had segments covering a quarter to half the colony where all branches were pale or nearly white. Very few Acroporas were bleached, in sharp contrast to the relative rankings seen in French Polynesia. Almost all Montastrea curta were bleached very pale to nearly white. Porites was rare, but bleached in spots. Partial bleaching was seen in Favia, Favifes, Leptoria, Acanthastrea, and Pavona species. A Turbinaria coral species with a yellow rim was common, and never seen bleached (Figures 31-33)

c. Algae: Calcareous encrusting red algae, primarily Porolithon, and Halimeda covered about 10% of the bottom. Dictyota were present, but were rare (no more than 1%).

d. Local reef stresses: Reefs appeared to be affected primarily by natural stresses. Cyclone Sally passed directly over Rarotonga on January 1, 1987 (New Year's Day), after hitting Aitutaki several days earlier. Reefs on the north coast were most affected. Acanthaster damage to the reefs has been rare, but a major outbreak did occur in the 1970's.

Current degradation of the reefs of Rarotonga are attributable to commercial aquarium fish collectors. Most large Pocillopora corals have been partially or completely smashed open by aquarium trade tropical fish collectors who routinely break the coral heads to drive flame angel fish out from between the branches into nets. These collectors operate with permits to collect by netting from government, although their presence is actively opposed by some local sport divers. Their activities degrading the coral reef are nearly impossible to document and enforce. There is considerable spear fishing by collectors, sometimes using SCUBA gear and in areas where the fish have been tamed through repetitive feeding. Dynamite use in fishing also has been reported. Ciguatera is a problem with some species of surgeonfish and triggerfish.

Some areas of Porolithon, which were initially thought to be overgrown by a yellowish calcareous algae, were later reported to be affected by Coralline Lethal Orange Disease or "CLOD" (Littler and Littler, 1995). This algal growth was observed on less than 5% of the calcareous algae (Figure 30).

12. Outer reef near "The Pinnacle", off the southwest corner of Rarotonga (August 14 1994)

a. Physical setting: Physical setting consists of mounds and hills on a gradually sloping (30°) coral plain. The reefs here are quite a bit further offshore than those in the northwest, in generally rougher waters because of greater exposure to the trade winds, and are consequently less affected by divers or fishermen. The area was relatively unaffected by the last hurricane

b. Corals and bleaching: Coral cover was very high, around 90%. Acropora was the most common genus, with perhaps a dozen species with a wide range of morphologies. Species diversity and colony size were high, and similar to the previous site. Bleaching was very rare at this site, and only about 1% of colonies were affected. All were Pocillopora verrucosa with bleaching in some branches. (Figures 34-37)

c. Algae: Similar to previous site, except that fleshy algae were not seen.

d. Local reef stresses: Wave stresses are evident. Acanthasfer was rare, only one was seen. Many large Pocil/opora heads had been broken by fish collectors. Other branching corals nearby were undamaged, indicating that wave stress was not responsible.

Figure 33. Site # 11. "Black Rock" on outer slope of Rarotonga. Several corals in panoramic view of reef Pavona. Acropora, and Goniastrea. Depth 10 meters.

Figure 34. Site # 11. ''Black Rock'' on outer slope of Rarotonga. Assortment of corals covering reef: Acropora, Pocillopora, Pavona, Lobophyllia. Depth 20 meters.

Figure 35 Site #12. "The Pinnacle'' on outer slope of Rarotonga. Partially bleached coral showing overgrowth by filamentous algae at skeletal tips. Depth 20 meters.

Figure 36. Site #12. "The Pinnacle" on outer slope of Rarotonga. Two partially bleached colonies of Acropora and bleached Montastrea. Depth 20 meters.

Figure 37. Site #12. "The Pinnacle'' on outer slope of Rarotonga. Corals indicating normal colors and growth forms: Acropora, Pavona, Pocillopora. Depth 15 meters.

Figure 38. Site #12. "The Pinnacle" on outer slope of Rarotonga. Bleached, dead and algal encrusted Pocillopora with Porites, Pavona, Acropora. Depth 20 meters.

Figure 39. Site #12. "The Pinnacle" on outer slope of Rarotonga. Colonies of Pocillopora verrucosa broken apart to collect tropical aquarium fish. Because surrounding corals indicate no evidence of physical damage, this bashing of Pocillopora is not attributed to high wave energy, but to specifically targetted and deliberately destructive human activity. Depth 20 meters.

Figure 40. Site #12. "The Pinnacle" on outer slope of Rarotonga. Close-up of Acropora showing bleaching at base of colony. Depth 20 meters.

Figure 41. Site #12. ''The Pinnacle" on outer slope of Rarotonga Close up of Acropora showing green fluorescence Depth 3-5 meters.

13. Northwest Rarotonga Lagoon, off Black Rock (August 14 1994)

a. Physical setting: Lagoon was very shallow, about 2 meters deep. Very large coral heads were abundant, but most of these were dead.

b. Corals and bleaching: Coral heads were largely made up of Porites microatolls, with some young Acropora of around 4 species. Live coral cover was around 30% of reef surface, the rest being dead and algae overgrown, especially on top of heads. No coral bleaching was seen.

c. Algae: Calcareous algae covered about 60% of reef surfaces. This community was very abundant and diverse, including Halimeda opuntia, Galaxuara rugosa, Galaxuara acunimata, Galaxaura fastigiata, Amphiroa fragilissima, and Jania, which were overgrowing live coral and Porolithon. Large mats of the green fleshy bubble algae Valonia aegagropila up to 30 cm. across were overgrowing coral bases and made up about 10% of hard bottom. Fuzzy brown turf algae were found on intertidal rocks. Fragments of the fleshy algae Turbinaria ornata, Sargassum, Dictyota, and Enteromorpha were found washed up on the beach, but were not seen growing in the lagoon.

d. Local reef stresses: Algae were actively overgrowing corals, indicating eutrophication. The fact that branching calcareous algae dominate over fleshy ones indicates that eutrophication is in the early moderate stages and can still be reversed if prompt action is taken to improve water quality by reducing nutrient loading from inadequately treated sewage. The lagoon here is lined with hotels and homes. This area had been badly affected several years ago by being downcurrent from where dredging of sand was done to extend the airport runway. Runoff from juice factory wastes has affected other areas of the Rarotonga lagoon. The sea urchin Echinometra is extremely abundant, and causes high grazing pressure and active erosion of dead coral heads. Reef walking by tourists and reef gleaning by local residents add further stress. Rarotonga has very high rainfall, and erosion of soil causes turbidity after rains.

MAP 7. Tutuila Island, American Samoa. Two sites on the north coast, Masefau Bay and Fagasa Bay as well as two sites on the south coast, Afao Bay and Faga'alu Bay were surveyed (map adapted from Stanley, 1993).

TUTUILA ISLAND, AMERICAN SAMOA

Background: Reef bleaching began in March of 1994 (Peter Craig, personal communication) and up to 70-80% of corals became completely white. Bleaching in 1994 was much more severe than the previous event in 1991. Chuck Brugman (personal communication), who dived Samoan waters since 1975 saw coral reef bleaching for the first time in 1989 and has noted it regularly since. He described the bleaching as always rare and isolated until after the hurricanes of late 1991, when it became frequent in shallow water. Samoan reefs, which had been in excellent condition had been damaged by Cyclone Ofa in 1990 and then devastated by hurricane Val in 1991 (Craig, 1994). This cyclone was very unusual, because after the eye had passed over Samoa, it turned around and passed over it a second time from the opposite direction, staying in the area for six days. Shallow reefs all around Samoa were almost completely destroyed by the hurricane to depths of 20 meters. 1994 was far worse than any previous bleaching, and the only time the Heferactis anemones bleached. Bleaching also took place on the Manu'a islands and offshore bank reefs down to depths of 65-70 meters, but was less pronounced below 40-50 meters where the water becomes cooler by 1 to 2 °C. Crown of thorns starfish began increasing in the early 1970's and reached a peak in early 1979, doing great damage (Craig et al., 1993).

14. Masefau Bay northeast of Tutuila, northwest slope (August 16 1994)

a. Physical setting: Masefau Bay has an axis oriented to the northeast, and fringing reef grows along both sides of the bay. The reef on the northwest side grows on a nearly vertical drop off from a reef flat at less than a meter deep to a depth of 30 meters, where there is a sloping mud and sand plain.

b. Corals and bleaching: Live coral cover is very low, 20% overall, varying between 0% to 40% on various outcrops. There was a large amount of dead coral, both broken fragments and dead corals which are not physically damaged but are covered with mud over part or all of their surfaces. Large amounts of hurricane rubble were piled up everywhere. Coral diversity was fairly high, but there were very few Pocillopora orAcropora colonies. About 50% of all colonies were bleached, and around a third were bleached fully white over all or much of their surfaces. White bleached corals appeared to especially susceptible to sediment accumulation, apparently being able to put less energy into removing sediment. Corals observed with fully white bleached patches included Pachyseris (the most heavily affected species), Porites, Montastrea curta (all uniformly bleached), Pavona clavus, plate-like Pavonas, Favia, Favites, and possibly Leptoseris. Minor or partial bleaching was seen in Diploastrea, (and possibly Plesiastrea and other Favids), Stylonociella, many species of Montipora, Milleporas, Synarea, Goniopora, and Goniastrea. Little bleaching was seen in Astreopora, Lobophyllia, Psammocora, Hydnophora, Coscinarea, Cyphastrea, Gardinoseris, Halomitra, or Oxypora. (Figures 40-44)

c. Algae: Algae dominated the reef framework, with about 50% calcareous algae cover and 30% fleshy algae. Porolithon and encrusting red algae cover the 1991 hurricane rubble. Branching calcareous algae, especially Halimeda, was extremely abundant, and did not appear grazed. Halimeda made up most of the sand component in the sediment, mixed with terrigenous mud. Also in large clumps were Amphiroa fragilissima, Jania, Galaxaura acunimata, Galaxaura fastigiata, Galaxaura rugosa, and Liagora species, plus several kinds of Porolithons, red and violet Lithophyllums, Peysonnelia rubrum and other encrusting calcareous reds. Fleshy algae included large mats of blue-greens similar to Lyngbya, Centroceras, Chlorodesmis, and minor amounts of Dictyota.

d. Local reef stresses: The area was severely damaged by the hurricane, and was subject to high sediment stress from mud and soil which was eroded by the numerous rivers that flowed down steep mountain slopes into the Bay. It was undergoing eutrophication, and algae are overgrowing corals. High suspended matter concentrations appear to support filter feeding and bacterial feeding organisms, indicated by large numbers of small yellow and white vase shaped sponges. Bleaching was seen in anemones, alcyonarians, in one encrusting sponge, and in encrusting Didemnid ascidians.

Figure 42. Site #14. Masefau Bay on outer slope of American Samoa. Panorama indicating severe bleaching and sedimentation on Pachyseris corals. Dead reef frame also evident. Depth 10-15 meters.

Figure 43. Site #14. Masefau Bay on outer slope of American Samoa. Bleached Merulina colonies. Depth 20 meters.

Figure 44. Site #14. Masefau Bay on outer slope of American Samoa. Colony of Pachyseris showing mottled appearance of partial recovery from bleaching. Small corals appear in periphery with some background sediment. Depth 20 meters.

Flgum 45. Site #14. Masefau Bay on outer slope of American Samoa. Bleached coral showing sediment damage and partial recovery at edges. Depth 20 meters.

Figure 46. Site #14. Masefau Bay on outer slope of American Samoa. Bleached Pachyeris indicating partial recovery and some dead corals. Depth 20 meters.

15. Masefau Bay northeast of Tutuila, southeast slope (August 16 1 994)

a. Physical setting: The reef slopes at about 30-45° from 1 to 20 meters, where the mud and sand plain began.

b. Corals and bleaching: Coral cover is about 20%, but is very variable from place to place. There is much less mud stress to corals than on the other side of the bay, apparently as a result of circulation patterns in the Bay. About 30% of corals were bleached, most partially. Species were generally similar to the other side of the bay but Acropora and Pocillopora were more common and there was much less Pachyseris. In addition Galaxea and Fungias (including bluish, pale, and yellow-rimmed species) were common. There was very little bleaching in Acropora. There were several Acropora species not previously encountered, similar to A. humilis, A. samoensis, A. kirstyae, A. cytherae or A. Iatistella (in deeper water at the base of the sediment plain, forming anastomosing colonies), and A. echinata or A. subglabra. Pocillopora verrucosa was much more affected than P. eydouxi. About half of all Pocillopora colonies were recently dead, with a layer of fine filamentous green and brown algae, and Porolithon colonizing only the tips of the knobby verrucae. Older dead corals, as from the hurricane, are completely covered with Porolithon.

c. Algae: Calcareous algae contribute to about 60% of bottom cover. Large amounts of coral rubble are overgrown with Porolithon. Fleshy algae cover is about 20%. Algae were similar to the previous site, except that fleshy algae were less abundant (especially blue-green algae), and the branching calcareous reds and greens were more abundant, especially Halimeda and Galaxaura. Peyssonellia was very common on deeper and steeper areas between 15 to 20 meters, especially under overhangs.

d. Local reef stresses: Sediment stress is apparently lower than on the other side of the Bay. Large amounts of garbage were seen as bottles, plastic, diapers, and food scraps. The more gentle slope of the reef on this side has apparently allowed much more trash to be trapped. Residents of the village at the head of the Bay make a practice of shoveling their garbage into the river during heavy rains as a quick way of disposal. The site is undergoing eutrophication.

16. East side of Fagasa Bay, northwest Tutuila (August 16 1994)

a. Physical setting: Topographically very complex, as the bottom consists of huge rounded basalt boulders down to about 15 meters, where the mud and sand plain starts. Sediment, population, and wave stresses all appear to be lower at this site than the previous bay.

b. Corals and bleaching: Live coral cover was very low, at less than 10%, and around 20% of the colonies were partially bleached. Coral types were similar to the previous site, with Galaxea and Fungias especially common. (Figure 47)

c. Algae: Calcareous algae covers 80% of the bottom. Basalt boulders are covered with Porolithon crusts. Fleshy algae make up around 10%. Calcareous algae were similar to the previous site, but with greater abundance of encrusting calcareous red algae, especially Porolithon, which completely covered basalt surfaces.

d. Local reef stresses: This area consists of a coral community on basalt boulders without a reef framework. Terrigenous sediment was relatively low, but large amounts of village garbage, including plastic, bottles, diapers, and sanitary napkins were seen. The site is undergoing eutrophication, but at a lower level than the previous two sites.

Figure 47. Site #16. Fagasa Bay on outer slope of American Samoa. Pocillopora verrucosa, partially bleached Fungia, and rubble at base of reef. Depth 10 meters.

17. Outer slope off Atauloma Bay or Afao Bay, southwest Tutuila (August 17 1994)

a. Physical setting: Very well developed reef formations facing open ocean on a slope of 15 to 30 degrees. Constructional coral buttresses up to 6 meters high, with Halimeda sand channels between. Only occasional terrigenous sediment enters the bay as there are only small streams, and the reef flat traps sediment close to the source.

b. Corals and bleaching: Live coral cover down to 20 meters was about 50%. There were large piles of coral rubble everywhere. There were large numbers of young corals 5 to 15 cm across which had settled since the hurricanes. The fore reef has extremely well developed constructional spur and groove morphology. Shallow areas are dominated by Acropora, and deeper areas by Porites. A remarkable feature is huge, very smooth surfaced, ancient Porites heads, up to 3 meters high and 5 meters across. These had some dead areas, but were clearly the major survivors of the hurricane, since almost all other corals were small and young. 40 to 50% of the corals had some bleaching. Porites was heavily bleached, with large pure white areas, or white spots ranging in size from 1-2 meters down to less than a centimeter. These appeared to be recovering from bleaching by filling in from the pigmented surroundings, and all appeared to be fully bleached in photographs of the area taken by Craig in March and in July. Alcyonarians were very abundant, including blue and green encrusting types, mostly in water shallower than 3 meters, and large stalked Sarcophyton colonies in deeper water. All had areas bleached white, especially Sarcophyton, but appeared to be recovering. Anemones were also largely bleached. Bleaching was also seen in Astreopora (bleached patches common), Goniopora (some pale patches) Alveopora (few bleached patches), Coscinarea (white patches common), Leptoseris (very white areas running at right angles to valleys), Pachyseris (bright white patches), Gardinoseris (white patches common), Galaxea (especially on horizontal bases and vertical calyx sides), Diploastrea (bleached in radial spokes within each polyp), Favia, Favites, Montipora, Hydnophora, Merulina, Turbinaria (but only rarely, most were very darkly pigmented with a bright yellow rim), Montastrea curta, Pavona, Psammocora (bleaching on top), Platygyra, Goniastrea, Oulophyllia (white patches), Leptoria (white patches), Millepora dichotoma (spotting), and various Favids which could include Oulastrea, Plesiastrea, Diploastrea, Lepatastrea, and Cyphastrea. Lobophyllia were few, and not bleached (Figures 48-52).

c. Algae: Calcareous red algae encrusted all the rubble and bare reef surfaces, covering around 45-50% of the bottom. They included several kinds of Porolithon, Mesophyllum in shaded underhangs, Hydrolithon, and Peyssonellia. Live Halimeda and branching reds were very rare, but the sediment was almost entirely made up of whole Halimeda opuntia sand. Fleshy algae were present in low abundance, less than 5%, mainly Bryopsis, Chlorodesmis, Lobophora, Ralfsia, and blue greens growing on recently dead corals, especially Pocillopora, in which Porolithon had colonized all the verrucae tips. Dictyota and Laurencia were uncommon.

d. Local reef stresses: Devastating damage was done by the hurricane of 1991, which killed most branching and platy corals down to at least 20 meters, and created huge piles of rubble. Sediment stress was not evident. Algal overgrowth was not apparent. Overfishing is reported by spear fishermen on SCUBA, who sometimes use fish poisons, sodium hypochlorite, or dynamite. There is also collecting of fish for export for the aquarium trade. This reef site had the highest living coral cover and the least terrigenous sediment seen in American Samoa, and was recovering from hurricane damage at a rapid rate, but needs to be protected from further anthropogenic stress.

Figure 48. Site # 17. Afao Bay onouter slope of American Samoa. Pocillopora, Acropora, Montipora and Leptoseris at edge of reef . Depth 10 meters.

Figure 49. Site # 17. Afao Bay on outer slope of American Samoa. Partially bleached and dead hard corals and white-bleached soft corals. Depth 20 meters.

Figure 50. Site # 17. Afao Bay on outer slope of American Samoa. Bleached soft coral growing on surface of dead coral rock. Depth 15 meters.

Figure 51. Site # 17. Afao Bay on outer slope of American Samoa. Almost completely bleached and mottled coral of Porites. Depth 20 meters.

Figure 52. Site # 17. Afao Bay on outer slope of American Samoa. Porites colony showing distribution and incidence of residual bleaching zones. Depth 20 meters.

18. Outer slope off Faga'alu Bay, south Tutuila (August 17 1994)

a. Physical setting: The reef flat here is quite narrow, only around 10 to 20 meters wide, and drops off nearly vertically to a depth of 20 meters, where a mud plain abuts the base of the reef.

b. Corals and bleaching: The reef flat was composed mostly of dead rubble, and had large amounts of muddy sediment, garbage, steel drums, and scrap metal. The vertical wall had very low coral cover, 10% or less. There was large amounts of coral rubble, but also many intact dead sediment-covered corals and portions of living coral colonies which were dead (Figure 54). Species diversity was less than the previous site, and seemed to be a sediment stress-adapted community. About half of all corals were bleached, and intense white bleaching was seen in Porites, Montipora, Diploastrea, Acropora, Pocillopora, Merulina, Pachyseris (very white, and high sediment mortality), Gardinoseris, Astreopora (heavy bleaching), Echinophyllia (heavy bleaching), and in Alcyonarian soft corals. Diploastrea showed a curious pattern of recovery of pigmentation, in which individual mesenteries were normally pigmented, with the rest white, resulting in a bicycle-spoke or pie-slice coloring pattern in individual polyps. (Figures 53-54)

c. Algae: About 70% of the bottom was covered with Porolithon and other encrusting red algae, but there were few Halimeda or branching calcareous red algae. About 20% was covered with reddish blue-green algae mats, and Sphacelaria turf, but no fleshy macrophyte algae were seen.

d. Local reef stresses: Algal turf overgrowth of corals is apparent. Sediment stress is also very high. There were large amounts of garbage ranging in age from recent plastic and bottles to massive amounts of World War II trucks, boats, construction debris, and portions of concrete bunkers which had all apparently been bulldozed over the reef flat in the late 1940s and early 1950s. There was a very high abundance of sponges, especially an orange branching species and small white vase shaped sponges. This may indicate high levels of bacteria in the water, which they filter out and feed on (Reiswig, 1971). This reef was in the most degraded condition of any seen in American Samoa, with many dead and dying corals and very few small ones. The area is greatly affected by sediment and pollution from Pago Pago Harbor which is just upcurrent, including waste from two tuna canneries which are released into the Harbor through a pipeline. Pago Pago has became severely eutrophic from cannery wastes released from shore into the head of the Bay, which has very slow circulation. About 5 or 6 years ago the sludge began to be separated and transported by barges to dump sites at sea 3 or 4 miles offshore. For the past three years or so, liquid wastes are piped to discharge sites in the center of Pago Pago Bay rather than being released directly from shore as was done for many preceding years (Craig, 1994).

19. Reef flat, Atauloma Bay, southwest Tutuila (August 17 1994)

a. Physical setting: Extensive fringing reef flat about 100 meters across, essentially at sea level, with a small stream entering from the North. Terrigenous sediment blankets the shore and the innermost reef flat, but is trapped near where it enters and most of the reef flat has a thin cover of white limestone reef sand in crevices and tide pools. The ecological discussion refers to the community on the seaward side of the fringing reef flat.

b. Corals and bleaching: No bleaching was seen on the reef flat. The corals appeared to be highly adapted to sediment, freshwater, and mechanical stress, and composed a very distinct community from the fore reef. They also appear to be highly resistant to temperature stress as well. Coral cover was less than 10% near the shore where freshwater and mud enter but rose to 50% towards the outer edge of the reef flat. The nearly flat bottom is dominated by at least 8 Acropora species, including flat, corymbose, table-like, encrusting, massive, and horizontally as well as vertically branching species. Encrusting Montipora species were also present, and some Pocillopora, many of which were dead.

c. Algae: Porolithon covered dead corals and bare rock between coral colonies was around 30% of the bottom. About 20% was composed of fleshy algae. Dictyosphaeria versluysli clumps 1 to 5 cm across were scattered all across the reef flat. Bright green Chlorodesmis mats were up to 10 to 20 cm across. A very large variety of blue green algae formed dense mats in crevices and over the bases of dead Acropora and Pocillopora. These colors included from red, blue, aqua, and green.

Figure 53. Site #18. Faga'alu Bay on outer slope of American Samoa. Bleached Acropora with algal encrusted rock and heavy sediment loading. Depth 20 meters.

Figure 54. Site #18. Faga'alu Bay on outer slope of American Samoa. Sediment loading and bleaching of a living Porites colony. Depth 20 meters.

d. Local reef stresses: Heavy use by reef walkers who use metal bars to turn over rocks and corals to get at clams, and who stand on the edge of the reef to line- fish. The entire surface is likely to be periodically stepped on by humans (Enright, 1993). Sediment and freshwater stress appears to be limited to a only small area near the river mouth, except during periods of exceptionally heavy rain. The whole area was devastated by hurricanes in 1991, overturning and killing most corals. Eutrophication is evident from high levels of the sewage indicator algae Dictyosphaeria and blue-green algae (cyanobacteria).

INTERVIEWS WITH OTHER OBSERVERS ABOUT SITE DESCRIPTIONS AND ASSESSMENTS

FRENCH POLYNESIA

Henri Poliquen reported SCUBA diving in the Austral islands and southern Tuamotos in July of 1994. He saw coral reef bleaching in Tubuai and Rurutu, comparable to or less than that in Tahiti, and slight coral bleaching was observed in Mururoa. Francis Rougerie of ORSTOM reported bleaching in the Gambier Islands (Mangareva) in a telephone interview in Tahiti.

Yannick Chancerelle provided us a preliminary report on coral reef bleaching based upon his observations in Moorea, in the leeward Society Islands, and throughout the Tuamoto archipelago. He noted that all lagoon coral species in the lagoon around Moorea had bleached at the peak of the episode. Intense fluorescence of bleached corals was observed only in May and June. He noted total bleaching in Pavona varians and Pachyseris at depths of 40 to 60 meters. His observations (Chancerelle, in preparation) from outer reef habitats in the following locations are tabled below

SITE COAST MONTH % BLEACHED

Society Islands:

Moorea North May 20-25

Tahiti Northwest May 20-25

Tetiaroa South May 10-20

Raiatea West May 10-20

Tikehau Northwest June <10

Mataiva West July <10

Tuamotu Islands:

Rangiroa North July <10

Takapoto South July <10

Nengo-nengo North Juiy <10

Nengo-nengo Lagoon July 10-20

Marutea West, South August <10

Yves Lefevre dives extensively across the Northern Tuamotos. He confirmed coral reef bleaching similar to that seen in Rangiroa at Tikehau, Manihi, Fakarava, Toau, Kauehi, and 6 other atolls.

NORTHERN COOK ISLANDS

Mild bleaching was reported in Manihiki, Northern Cooks, in August 1994 and in Mauke and Mangaia in the Southern Cooks in July 1994 (Flora and Flora, 1 994).

WESTERN SAMOA

Coral bleaching was reported to be comparable to that described for American Samoa (Craig, personal communication from SPREP).

NIUE

Coral reef bleaching is reported by Terry Coe to have occurred during February, 1994. This episode lasted for about one month and was followed subsequently by recovery. At first this event was thought to be attributable to grazing of A. planci or to coral spawning, but these possibilities were later discounted as responsible factors. This is the first time such bleaching has been observed in the period since 1967 as reported by the owner of the only professional SCUBA diving operation in the area since 1981 (Terry Coe, personal communication).

KlRlBATI

No bleaching has been seen in the Gilbert Islands which have been under high clouds and El Nino conditions with persistent and heavy rain (Tererei Abete, personal communication). Also, no bleaching was observed in Tarawa during June by Flora. However, the "hot spot" which caused bleaching in the South Central Pacific stalled over the area in July and August.

No bleaching is reported from the Phoenix Islands, although bleaching may have occurred based on elevated satellite SST anomalies. Only six families of coconut harvesters and no fisheries officials live there, so observers were not available to confirm changes in the local reefs.

MARSHALL ISLANDS

Severe coral reef bleaching was observed in the lagoon around Majuro during November 1992, leading to the death of 50% of the Acropora colonies (Paulay, personal communication). Recently, no bleaching was seen in Majuro during July 1994 (Flora, personal communication). Bleaching in early 1994 was not expected due to the absence of thermal anomalies. However, anomalous water temperatures have occurred since August, 1994, and bleaching could have developed thereafter.

TOKELAU, WALLIS AND FORTUNA

No reports of coral bleaching are available, but bleaching may be likely based on satellite SST anomalies.

FIJI, TONGA, TUVALU AND VANUATU

No reports have been received. Coral reef bleaching is not expected due to fairly normal seasonal satellite SSTs.

HAWAII

No bleaching was seen in Maui in June by Flora. Bleaching not expected due to seasonally normal satellite SSTs.

STATISTICAL ANALYSIS AND DISCUSSION

BLEACHING AND LOCAL ENVIROMENTAL FACTORS

The values of the fifteen environmental parameters measured or estimated at all nineteen sites were examined for statistically significant relationships between variables by simply determining the degree to which their ranks co-varied at all sites. Table 1 presents the values of the variables at each site and table 2 gives the values of the correlation coefficients and their statistical significance. The Spearman non-parametric rank order correlation coefficients were used. One asterisk (*) indicates a statistically significant correlation with a P value of 0.05 or less (i.e. 95% probability that this relationship could not have been produced by chance), two asterisks (**) indicates a strongly significant correlation with a P value of 0.01 or less (i.e. 99% probability), and three asterisks (***) indicates a very strongly significant correlation with a P value of 0.001 or less (or a 99.9% probability of confidence that this correlation could not be expected to have arisen randomly).

Most environmental variables were not significantly correlated, as shown by the correlation matrices in Tables 2A and 2B. Some of the variables, however, were significantly correlated, either positively (Table 3) or negatively (Table 4). Graphs 1 and 2 display these significant positive and negative relationships, respectively. The interactions between all environmental and ecological variables are evaluated below, followed discussion of likely mechanisms by which these factors interact. In almost all cases intuitively obvious hypotheses for these relationships are developed. In only a few cases are unexpected relationships identified, which either reflect unknown linkages between variables or simply statistical artifacts.

Table 1A. Environmental data assessed at survey sites.

Site Location Habitat Temperature Salinity Oxygen pH

1 Tahiti lagoon 3 NA NA NA NA

2 Tahiti fore reef 1 NA NA NA NA

3 Moorea lagoon 3 26.2 27.9 NA 8.56

4 Moorea fore reef 1 25.6 27.7 NA 8.43

5 Rangiroa lagoon 1 25.8 28.4 96 8.57

6 Rangiroa fore reef 1 27.6 28.6 92 8.77

7 Rangiroa channel 2 25.7 28.1 96 8.52

8 Aitutaki lagoon 3 23.9 28.1 100 8.42

9 Aitutaki fore reef 1 1 24.7 28.1 98 8.63

1 0 Aitutaki fore reef 2 1 25.3 28.2 97 8.62

11 Rarotonga fore reef 1 1 23.9 27.4 100 8.56

1 2 Rarotonga fore reef 2 1 NA NA NA NA

13 Rarotonga lagoon 3 NA NA NA NA

14 Samoa bay 1 2 27.0 27.0 94 8.49

1 5 Samoa bay 2 2 NA NA NA NA

1 6 Samoa bay 3 3 27.6 28.3 94 8.43

17 Samoa fore reef 1 1 27.6 27.7 92 8.69

18 Samoa fore reef 2 2 28.0 27.5 92 8.64

1 9 Samoa reef flat 3 NA NA NA NA

NA = not available

Table 1 B. Environmental data assessed at survey sites.

Site Live Coral Bleached Fleshy algae Hard algae

1 20 10 50 30

2 70 10 5 25

3 30 1 50 20

4 60 30 10 30

5 50 10 0 50

6 50 50 0 50

7 50 5 0 50

8 40 5 10 50

9 70 20 0 30

10 80 10 0 20

11 90 20 1 9

12 90 1 0 10

1 3 30 0 10 60

14 20 50 30 50

15 20 30 20 60

16 10 20 10 80

17 50 40 5 45

18 10 50 20 70

19 50 0 20 30

Table 1 C. Environmental data assessed at survey sites.

Site Population Mud Acanthaster Yrs cyclone Prev. Bl. Pollution

1 5 2 0 7 3 4

2 4 1 0 7 3 2

3 3 2 0 7 3 3

4 2 1 0 7 3 2

5 1 1 0 2 1 1

6 1 1 0 . 2 1 1

7 3 1 0 2 1 2

8 2 2 0 7 1 2

9 1 1 50 7 1 1

10 1 1 1 7 1 1

11 2 1 1 7 1 1

12 2 1 1 7 1 1

·13 3 2 0 7 1 3

14 3 4 0 3 1 4

15 3 3 2 3 1 4

16 3 3 0 3 1 4

17 3 2 0 3 1 2

18 5 4 0 3 1 5

19 4 2 0 3 1 3

Table 2A. Correlations between environmental variables.

Factors Hab Temp. Sal. Oxy. pH L.Cor. Bl. Cor. Fl. alg

Habitat - .1363 -.1684 .0129 -.451 -.6974 -.436 .7434

Temperature NS - .0334 -.9720 .329 -.6826 .554 .3120

Salinity NS NS - .839 .157 .0632 -.311 -.6068

Oxygen NS (***) NS - .476 .5332 -.707 -.2526

p H NS N S NS NS - .2084 .344 -.4076

Live coral (***) (**) NS NS NS - .21 -.7206

Bleached NS * NS (*) NS NS - .0208

Fleshy algae *** NS (*) NS NS (***) NS -

Hard algae NS * NS NS NS (***) NS NS

Human Population ** NS (*) NS NS (*) NS ***

Mud cover ** * NS NS NS (***) NS ***

Acanthaster NS (*) NS * NS * NS NS

Yrs. typhoon NS (*) NS * NS NS NS NS

Prev. bleach NS NS NS *** NS NS NS NS

Pollution *** * NS NS NS (***) NS ***

NS = not significant

Table 2B. Correlations between environmental variables.

Factors Hd. algae Pop. Mud A. planci Yrs.T Prev.BI Poll

Habitat .3504 .5911 .6623 -.4015 .0418 .102 .715

Temperature .6050 .5375 .5670 -.6076 -.6275 -.057 .565

Salinity .2583 -.5724 -.4215 -.0338 -.3539 -.201 -.448

Oxygen -.5330 -.4469 -.4533 .6133 .6509 1.000 -.466

p H -.1108 -.2199 -.1661 .2353 -.2665 -.343 -.374

Live coral -.7702 -.5655 -.8800 .4694 .3499 .000 -.881

Bleached .3218 -.0817 .2184 .0441 -.3617 -.119 .128

Fleshy algae .2127 .7183 .7983 -.3675 .1342 .410 .871

Hard algae - .2237 .5977 -.3691 -.5983 -.395 .551

Human population NS - .6244 -.4433 -.0431 .330 .824

Mud cover ** ** - -.2854 -.1929 -.127 .897

Acanthaster NS NS NS - .3250 -.305 -.403

Yrs. typhoon (**) NS NS NS - .463 -.138

Prev. bleach NS NS NS NS * - .182

Pollution * *** *** NS NS NS -

NS = not significant

Table 3. Positive relations between variables based on Spearman rank correlation coefficients. Graph 1 depicts these relationships in graphic format.

VERY STRONGLY SIGNIFICANTLY POSITIVE (P < 0.001)***

Habitat: Fleshy algae

Habitat: Pollution

Human population density: Fleshy algae

Mud cover: Fleshy algae

Pollution: Fleshy algae

Pollution: Mud cover

Pollution: Human population density

Previous bleaching: Dissolved oxygen

STRONGLY SIGNIFICANTLY POSITIVE (P < 0.01)**

Habitat: Human population density

Habitat: Mud cover

Human population density: Mud cover

Mud cover: Calcareous algae cover

SIGNIFICANTLY POSITIVE (P < 0.05)*

Proportion of bleached coral: Temperature

Calcareous algae cover: Temperature

Mud cover: Temperature

Pollution: Temperature

Acanthaster: Dissolved oxygen

Years since hurricane: Dissolved oxygen

Live coral cover: Acanthaster

Pollution: Calcareous algae cover

Previous bleaching: Years since hurricane

Graph 1. Significant positive correlations between variables based on Spearman rank coefficients as shown in Tables 2A, 2B and 3. The denser the arrow connecting variables, the more statistically significant the relationship: P = <0.001 as represented by the thickest arrows; P = <0.01, by the thinner arrows; and P = <0.05 by the thinnest arrows.

Table 4. Negative relations between variables based on Spearman rank correlation coefficients. Graph 2 depicts these relationships in graphic format

VERY STRONGLY SIGNIFICANTLY NEGATIVE (P < 0.001 )***

Live coral cover: Habitat

Live coral cover: Fleshy algae cover

Live coral cover: Calcareous algae cover

Live coral cover: Mud cover

Live coral cover: Pollution

Dissolved oxygen: Temperature

STRONGLY SIGNIFICANTLY NEGATIVE (P < 0.01)**

Live coral cover: Temperature

Calcareous algae cover: Years since hurricane

SIGNIFICANTLY NEGATIVE (P < 0.05)*

Bleached coral cover: Dissolved oxygen

Live coral cover: Human population density

Acanthaster: Temperature

Years since hurricane: Temperature

Salinity: Fleshy algae cover

Salinity: Human population density

Graph 2. Significant negative correlations between variables based on Spearman rank coefficients as presented in Tables 2A, 2B and 4. The denser the arrow connecting variables, the more statistically significant the relationship: P = <0.001 is represented by the thickest arrows; P = <0.01, by the thinner arrows; and P = <0.05 by the thinnest arrows.

The habitat parameter was very strongly positively correlated with both pollution and with fleshy algae cover. It was strongly positively correlated with mud cover and population density, and very strongly negatively correlated only with coral cover. The habitat parameter is a measure of exposure to open ocean wave energy, with higher values indicating sheltered and protected reefs. As sheltered coasts are the homes of the bulk of the human population, they have higher rates of soil erosion and sewage nutrient inputs to the coastal reef zones, both of which damage corals.

Water temperature, although actually measured in the field in August near the coolest time of year, was positively correlated with proportion of bleached corals, with calcareous algae cover, with mud cover, and with pollution. It was negatively correlated with years since the last cyclone and with Acanthaster, strongly negatively correlated with live coral cover, and very strongly negatively correlated with dissolved oxygen. Some of these relationships result from climatic gradients or historic accidents, since recent storm damage has been worst at the northern sites, which were also the warmest, and closer to the main hurricane tracks. Cyclones damage coral, removing the food for Acanthaster and allowing encrusting calcareous algae to spread over dead reef rubble and framework

Salinity was negatively correlated with population and with fleshy algae cover. This reflects the fact that human populations, and the sewage nutrients they release to fertilize algae, are most abundant near freshwater sources.

Dissolved oxygen was very strongly correlated with previous coral reef bleaching events, positively correlated with years since the last hurricane and with Acanthaster, negatively correlated with proportion of corals bleached, and very strongly negatively correlated with temperature. Oxygen is much less soluble at high water temperatures. The other correlations may result from the much lower dissolved oxygen levels measured in Samoa, the warmest site, which also had low Acanthaster abundance and recent storm damage. The negative relationship found between bleaching and dissolved oxygen is unexpected, and could be important if it indicates that stress from oxygen depletion near polluted sites makes bleaching worse or delays coral recovery.

pH was not correlated with any other variable. Its variations were small, and apparently without ecological significance.

Live coral cover (Graph 3) was positively correlated only with Acanthaster, but it was negatively correlated with population, strongly negatively correlated with temperature, and very strongly negatively correlated with habitat, fleshy algae cover, calcareous algae cover, mud cover, and pollution. All statistically significant interactions negatively affect live coral cover except only for Acanthaster, which is attracted to and eats live corals. The negative relationship with temperature reflects the historic site differences between cyclone-damaged, low-live coral cover Samoa, and Rarotonga, the coolest site with the highest coral cover. As noted in the site descriptions, there were additional strongly negative stresses to live corals, such as dredging, diver damage, factory effluents, overfishing, etc. at certain sites, especially the lagoonal ones, which are not included in the statistical analysis because they were important at only a few sites.

The proportion of corals bleached (Graph 4) was negatively correlated only with dissolved oxygen, and positively correlated only with temperature. High temperatures are linked to bleaching, even though our data were measured at the season of minimum temperature. Maximum temperatures should be much more highly correlated. Low oxygen levels could make bleaching worse or slow coral recovery. Comparison of bleaching to seasonal maximum temperatures and their anomalies are contained in another section.

Fleshy algae cover (Graph 3) was very strongly positively correlated with habitat, population, mud cover, and pollution, negatively correlated with salinity, and very strongly negatively correlated with live coral cover. This is because the abundance of fleshy algae is largely caused by excess nutrients from land-based sewage sources, and the over-fertilized algae overgrow, smother, and kill corals in eutrophic waters.

Calcareous algae cover (Graph 3) was strongly positively correlated with mud cover, positively correlated with temperature and pollution, strongly negatively correlated with years since the last hurricane, and very strongly negatively correlated with live coral cover. Calcareous algae appear to be elevated near high population, but not as much as fleshy algae. The association with mud may reflect their greater tolerance for lower light levels than corals. After cyclones damage corals, encrusting calcareous algae are able to take over and cover the reef rock and rubble, but after a few years corals settle on and overgrow them under healthy reef conditions, causing their cover to decline.

Human population density was very strongly positively correlated with fleshy algae and with pollution, strongly positively correlated with habitat and with mud cover, and negatively correlated with salinity and with live coral cover. Human activities cause greatly elevated land-based sources of eroded soil, sewage nutrients, and garbage, and elevated nutrients cause algae to overgrow and kill corals. The negative relationship with salinity is because less people live in highly saline areas which tend to have little available fresh water.

Mud cover was very strongly positively correlated with pollution and with fleshy algae cover, strongly positively correlated with habitat, with population, and with calcareous algae cover, positively correlated with temperature, and very strongly negatively correlated with live coral cover. Erosion is strongly dependent on dense populations clearing hillside forests for agriculture and housing, and it depends strongly on local rainfall, temperature, geology, and topography. Its impacts on corals are greatest in protected inshore habitats.

Acanthaster planci, the poisonous "Crown of Thorns" starfish, has swarmed over and killed reefs across the Indo-Pacific on a sporadic and unpredictable basis since the 1960s. At every site in this study local divers at first suspected that Acanthaster was the cause when they saw massive coral bleaching, but they quickly realized from their field observations of the distribution of both that there was no connection between them. Acanthaster was positively correlated with dissolved oxygen and live coral cover, and negatively correlated with temperature. Acanthaster is attracted to live coral, which was most abundant at the cooler sites which had less recent hurricane damage. Very few Acanthaster were seen at all sites but one, so these relationships could be fortuitous results of inadequate statistical sampling of rare events.

Graph 3. Coral and algae bottom cover at all sites.

Graph 4. Percentage of corals bleaching at all sites.

The number of years since the last cyclone was positively correlated with previous bleaching events, and with dissolved oxygen, was negatively correlated with temperature, and was strongly negatively correlated with calcareous algae cover. Tropical storms are the major natural factor which damages live coral and provides substrate for calcareous encrusting algae to overgrow reef limestone. Such storms are more frequent in the warmer areas, which also have lower oxygen levels.

Previous bleaching events were very strongly positively correlated with dissolved oxygen, and positively correlated with years since the last cyclone. These correlations are the only ones for which we are unable to offer a logical mechanism to explain the associations, and we suspect therefore, that these correlations may be a purely fortuitous statistical artifact which results from the small number of previous bleaching events at most sites.

Pollution was very strongly positively correlated with habitat, with fleshy algae cover, with population density, and with mud cover, positively correlated with temperature and calcareous algae cover, and very strongly negatively correlated with live coral cover. Pollution is very closely linked to high human population densities and to the proliferation of algae in response to high nutrient loading of inshore protected habitats.

These statistical analyses reveal many interesting environmental patterns. Although each reef had a unique set of environmental stresses, a large group of environmental variables correlated very closely with each other, in particular habitat type, human population density, mud cover, fleshy algae cover, and pollution. These variables are interrelated because human population densities are highest along protected shore lines, and they result in high levels of erosion, pollution, and excessive coastal zone nutrient levels as the result of runoff from land-based, population-dependent sewage releases

Human population densities on most Pacific, Caribbean, and Indian Ocean island nations are sufficiently high that most are probably undergoing general nutrient overloading of the coastal zone, and the rest are likely to be undergoing eutrophication on a local basis around population centers (Goreau and Thacker, 1994). Tables 5 and 6 show island nations where bleaching has occurred and their population densities. These data underestimate coral reef bleaching because the 1994 bleaching events are not included. Were those data available for inclusion, our conclusion that the distribution of bleaching has little relationship to human population density would be strengthened. Twenty eight island nations have human population densities above 200 inhabitants per square kilometer, and 33 have densities less than 200. Bleaching was thoroughly documented in 50% of high density islands and 33% of low density islands. The category of probable bleaching is based on the recognition of temperature anomalies from OMS data which are similar to those anomalies which preceded documented events in neighboring island nations. Probable bleaching has taken place in 36% of the high density island nations and in 48% of the low density ones. Island nations where bleaching is not considered to have occurred because of insufficient temperature anomalies include only 14% of the high density islands and 18% of the low ones. Bleaching reports are less likely from low density islands, which are often remote and inaccessible, deficient in fresh water, poorly populated, and lacking in marine laboratories or divers trained to recognize coral reef bleaching. These tables indicate that most coral reefs have undergone mass bleaching in recent years, and that this bleaching is independent of population-dependent stresses, such as from eutrophication that has affected many of these reefs.

 TABLE 5. Coral reef bleaching reported from high population density island nations (after Goreau and Thacker, 1994). Human population density for each island nation is expressed as population/sq. km. Only islands exceeding 200 are listed. Bleaching is categorized as severe (XX), moderate (X), probable (+) or unknown. (?).

ISLAND POPULATION DENSITY BLEACHING

SINGAPORE 4527.5 ?

BERMUDA 1094.0 X

OKINAWA 944.0 XX

MALDIVES 785.2 XX

BARBADOS 604.7 X

TAIWAN 562.0 ?

MAURITIUS 543.6 +

TUVALU 500.0 +

NAURU 476.2 X

ZANZIBAR 381.0 +

PUERTO RICO 366.7 X

ARUBA 321.2 +

U.S. VIRGIN ISLANDS 312.9 X

MARTINIQUE 302.0 +

SAINT VINCENT 283.5 +

MARSHALL ISLANDS 281.8 +

COMORO ISLANDS 271.6 XX

GRENADA 267.4 X

SRI LANKA 257.0 ?

GUAM 255.0 ?

TRINIDAD AND TOBAGO 249.3 +

HAITI 248.4 +

NETHERLANDS ANTILLES 238.8 X

JAMAICA 227.0 XX

SAINT LUCIA 223.0 X

PHILLIPINES 216.0 +

AMERICAN SAMOA 206.0 XX

REUNION 204.0 X

TABLE 6. Coral reef bleaching reported from low population density island nations (after Goreau and Thacker, 1994). Human population density is expressed as population/sq. km . Only islands with densities less than 200 are listed. Bleaching is categorized as severe (XX), moderate (X), probable (+) or unknown(?).

ISLAND POPULATION DENSITY BLEACHING

GUADELOUPE 197.8 +

FED. STATES OF MICRONESIA 162.4 +

SAINT KITTS AND NEVIS 160.9 +

SEYCHELLES 158.2 +

DOMINICAN REPUBLIC 156.4 +

ANTIGUA AND BARBUDA 152.3 +

SAO TOME E PRINCIPE 131.7 ?

TONGA 131.2 X

SAINT MARTIN 114.0 +

ANGUILLA 105.0 ?

KIRIBATI 103.3 X

CUBA 98.4 X

INDONESIA 98.0 XX

CABO VERDE 97.9 ?

DOMINICA 95.9 +

MONTSERRAT 94.0 +

CAYMAN ISLANDS 92.0 XX

MARIANA ISLANDS 85.4 ?

BRITISH VIRGIN ISLANDS 80.0 X

COOK ISLANDS 72.0 +

WESTERN SAMOA 55.8 XX

FRENCH POLYNESIA 50.0 XX

FIJI 40.9 X

BELAU 34.9 +

BAHAMAS 19.3 X

MADAGASCAR 19.0 ?

VANUATU 13.2 +

SOLOMON ISLANDS 12.3 +

PAPUA NEW GUINEA 9.0 +

NEW CALEDONIA 8.0 ?

NIUE 7 7 +

TURKS AND CAICOS 6.0 +

AUSTRALIA 2.0 X

Live coral cover is very strongly negatively correlated to all the strongly population-linked variables, and also to calcareous algae cover. These inverse relationships are because mud and fleshy algae kill live corals near high human population densities. Encrusting calcareous algae proliferation is most common after tropical storms when they overgrow freshly made dead coral rock and rubble surfaces. The abundance of branching calcareous algae is positively related to land-based, population-dependent nutrients. The relationship with nutrients is not as strong as it is for fleshy algae, whose growth kills calcareous algae as well as corals under high nutrient conditions.

All correlations with live coral cover, except only one, are negative. This means that almost all environmental stress factors reduce living coral cover. None favor reef cover by corals, with the exception of Acanthaster, which is positively correlated with coral cover. This curious and unexpected relationship (Acanthaster eats and kills corals) could be because the starfish are attracted to high coral cover. However, this association could also be due to inadequate statistical sampling of rare events because the starfish were absent or rarely seen at all sites but one, onto which they had apparently migrated only recently.

Coral reef bleaching was positively correlated with temperature, and negatively correlated with dissolved oxygen. This result could be a statistical anomaly, since oxygen levels were distinctly lowest in American Samoa. Both the most polluted and the warmest sites were also identified there. Pollution and elevated temperature operate independently to reduce oxygen, so the combination of the two factors at the same site would further amplify low oxygen stress. A single very strong statistical correlation, between oxygen and previous bleaching events, could be a coincidence, or it might indicate that low oxygen levels prolong the recovery of corals to bleaching. Further experiments are needed to test this relationship.

The rapid reef assessment methodology used here relied on direct field estimation of a large range of variables, using ranked values for stresses, and comparing them by non-parametric statistical analysis. This allowed a wide range of stresses to be evaluated even though there was little opportunity to make quantitative measurements (for example for pollution, nutrients, wave energy, seasonal variations, etc.). Since every reef faces its own unique environmental conditions, the study would have gained greatly from the opportunity to evaluate more sites. Nevertheless the number of sites examined over so large an area in such a short time appears adequate to have delineated many significant environmental trends. This rapid reef assessment approach provides more data per unit time and per unit cost than retrievable through conventional line transect analysis. While the use of line transects are fully complementary to the methods used here, it would have been impossible to have acquired as much information in this study had they been used. The research approach selected for this survey can be used to quickly and efficiently assess factors important in affecting the status of coral reefs over large areas, and to identify those environmental factors which require more detailed quantitative long-term monitoring. Further improvements could be made by improved quantitative estimates of many stresses such as habitat wave energy, population density, pollution, turbidity, etc.

The correlations found here strongly indicate both the influence of climatic anomalies on bleaching and of population-density dependent stresses on damaging coral reefs across the South Pacific. The major conclusions we draw from this analysis are that while many population-dependent and natural stresses are now acting to reduce living coral cover and causing reef deterioration, they are not related to mass coral bleaching. Each site examined had a unique complex of potential sources of stress. However, these individual factors appeared to bear little relationship to the causality of coral bleaching, except for having high temperature anomalies in the preceding warm season (next section).

BLEACHING, SEA SURFACE TEMPERATURE, AND ANOMALIES

In this section we analyze NOAA satellite-derived sea surface temperature data with regard to the global distribution of mass coral reef bleaching during 1993 and 1994.

Temperature readings from satellite data near continental shore lines or near high islands are often somewhat too low, probably because of the high cloudiness caused by rising air near the land-sea boundary. The small clouds which develop, especially under hot conditions, contribute a signal to the satellite sensor which actually reflects cloud top temperatures, rather than sea level values. Where clouds are smaller than the 4 km minimal resolution of the satellite pixels (and typical of tropical thermal convection) it is very difficult to remove sub-pixel scale contamination of the sea surface temperature signal since clouds are too small to be seen (Clarke, personal communication). The OMS data set is effectively smoothed over an even greater scale, up to several hundred kilometers by the data processing algorithm (Reynolds and Marisco, 1993; Reynolds and Smith, 1994). The low-resolution OMS temperature data set underestimated Caribbean sea surface temperatures when compared to the high resolution (4 km.) satellite images and in situ water measurements, especially under the warmest conditions (Goreau and Hayes, 1994, 1995). This is thought to be due to the highly complex spatial pattern of current flows and cloudiness around high Caribbean islands. In contrast to the Caribbean, the low resolution data appeared to work quite well for identifying warm season thermal anomalies greater than 1 °C (hot spots), which were found to be coincident with every major mass bleaching episode since 1983 in the much larger Pacific and Indian Oceans, where conditions are generally more spatially uniform (Goreau and Hayes, 1994). Hot spots are areas where the ocean surface becomes unusually warm due to local cloud and wind conditions that allow extra heating to occur.

 The reported satellite values must be understood to reflect open ocean surface temperatures offshore from the reef sites actually being evaluated. These values are representative of temperatures on outer slope reefs exposed to well circulating open ocean water, but they can be up to a few degrees different from water temperatures in reefs whose waters are poorly circulating. Shallow inshore areas and lagoons can get either considerably hotter or cooler than open ocean waters when winds, waves, currents, tides, sea level, or cloud cover are extremely high or low. For this reason accurate in situ water temperature measurements are definitive for assessing the temperatures actually experienced by reef corals under varying weather and circulation conditions (Gleeson, 1994; Gieeson and Strong, 1994). Such data are currently available at very few sites.

NOAA is considering stopping publication of the Oceanographic Monthly Summary for budgetary reasons, and stopped publishing the high resolution data used identify anomalies in the Caribbean in October, 1995. This publication has proven invaluable for following sea surface temperatures and their anomalies in coral reef regions around the globe. Without the OMS publications this study would have been far more difficult and time consuming, and indeed probably impossible because of lack of funds to acquire data sets stored in alternative media. We advise strongly against discontinuing distribution of the data without cost to the scientific community. Loss of access to these data will interfere with efforts to document coral bleaching and will inhibit development of early alert networks to monitor changes in the marine environment related to sea surface temperatures.

MAP 8.

World map indicating Atlantic and Indian Ocean sites whre NOAA sea surface temperatures were assessed for 1993 and 1994. The area of the insert is detailed in Map 9.

MAP 9.

The insert area from Map 8 showing the Pacific Ocean sites where detailed satellite sea surface temperatures were assessed during this study. Coral reef bleaching was observed directly during this study at sites marked by squares. Eyewitness accounts were reported at sites marked by diamonds. Although inquiries were made during this study, no reports of bleaching were received from sites marked by circles.

TABLE 7A. Geographic site coordinates to the nearest half degree.

LOCATION LATITUDE LONGITUDE

PACIFIC OCEAN SITES:

Mangareva, Gambier Islands 23.5S 135W

Mururoa, Tuamoto Islands 22S 139W

Rangiroa, Tuamoto Islands 15S 148W

Tahiti-Moorea, Society Islands 18S 149W

Tubuai, Austral Islands 23S 149W

Rurutu, Austral Islands 22S 151W

Rarotonga, Cook Islands 22S 161W

Aitutaki, Cook Islands 18S 161W

Manihiki, Cook Islands 11 S 162W

Niue 19S 170W

Tutuila, Samoa 14.5S 171W

Funafuti, Tuvalu 8.5S 179E

Tokelau 9S 172W

Uvea, Wallis & Futuna 13.5S 176.5E

Vavau, Tonga 19S 174W

Lau, Fiji 18S 179W

Canton, Kiribati 1 S 172.5W

Christmas, Kiribati 3N 156W

Tarawa, Kiribati 2.5N 172E

Majuro, Marshall Islands 7N 171E

Okinawa, Ryukyu Islands 26N 128E

Maui, Hawaii 21N 156W

TABLE 7B. Geographic site coordinates to the nearest half degree.

LOCATION LATITUDE LONGITUDE

INDIAN OCEAN SITES:

Malindi, Kenya 3S 40E

Zanzibar, Tanzania 6S 40E

Mayotte, Comoros 12.5S 45E

Mauritius 20.5S 58E

Mahe, Seychelles 5S 55.5E

Male, Maldives 4N 7.RF

ATLANTIC OCEAN SITES:

liha do Fernando de Noronha, Brazil 3S 33W

Recife, Pernambuco, Brazil 8S 35W

lihas dos Abrolhos, Bahia, Brazil 1 8S 38W

Arraial do Cabo, Rio de Janeiro, Brazil 23S 41W

liha do Sao Sebastiao, Sao Paulo, Brazil 24S 46W

The values of monthly mean sea surface temperatures and of the SST anomaly are provided in the figures corresponding to the following 33 sites shown in maps 8 and 9 and listed in Table 7. These sites include locations where bleaching was documented, where it was known not to have taken place, and where it is not known whether it took place. Temperatures and anomalies are first discussed individually for each site, and the complete results are summarized in a table and graphs, which are then statistically analyzed. Individual site Graphs 8A-B through 40A-B are grouped together at the end of the section for ease of comparison.

Mangareva, Gambier Islands

Water temperatures were higher and more prolonged in 1994 than 1993. An anomaly of up to +1.3 degrees occurred during the warm season in 1994. Bleaching was reported in 1994 (Graphs 8A and 8B).

Mururoa, Tuamoto Islands

Water temperatures reached higher values in 1994 than in 1993, and occurred during an anomaly of +1.1 degrees. Bleaching was reported in 1994 (Graphs 9A and 9B).

Rangiroa, Tuamoto Islands

Water temperatures reached 29.3 during a +1.0 degree anomaly in 1994, and reached 29.2 in 1993 during a +1.1 degree anomaly. Bleaching was reported in both years, with 1994 being worse. Temperatures stayed hotter longer in 1994. Bleaching was also seen throughout all the western and central Tuamotos. Some in situ temperature data are available for the lagoon (Graphs 10A and 10B).

Tahiti-Moorea, Society Islands

Water temperatures were considerably warmer in 1994 than 1993, and the maximum occurred at the end of a two month +1.1 degree anomaly. Bleaching took place during 1994 in Tahiti, Moorea, and all the Society Islands. In situ temperature data is available in fore reef and lagoon sites (Graphs 11 A and 11 B).

Tubuai, Austral Islands

Water temperatures were much warmer in 1994 than 1993, during a two month +1.1 degree anomaly. Bleaching was reported in 1994 (Graphs 12A and 12B).

Rurutu, Austral Islands

Water temperatures were much warmer in 1994 than 1993, during a +1.2 degree anomaly. Bleaching was reported in 1994 (Graphs 13A and 13B).

Rarotonga, Cook Islands

Water temperatures were much warmer in 1994 than 1993, during a three month +1.1 degree anomaly. Bleaching was reported in 1994. The remainder of the period was colder than normal (Graphs 14A and 14B).

Aitutaki, Cook Islands

Water temperatures were much warmer in 1994 than 1993, during a +1.1 degree anomaly. Bleaching was reported in 1994 (Graphs 15A and 15B).

Manihiki, Cook Islands

Water temperatures have a very small annual range. The 1994 warm period was hotter and longer than 1993. The maximum temperature, 30.1 degrees, came during a +0.9 degree anomaly, but the hot spot was unusually long lasting because a +1.1 degree anomaly and a +1.0 degree anomaly occurred just before and just after it. Bleaching was reported in 1994. Some in situ water temperature data are available only in the nearly enclosed lagoon (Graphs 16A and 16B).

Niue

Temperatures were considerably warmer in 1994 than 1993. Anomalies were up to +1.1 degrees during the summer of 1994, while the rest of the period generally had negative temperature anomalies. Bleaching was reported only in 1994 (Graphs 17A and 17B).

Tutuila, American Samoa

Water temperatures were much warmer during 1994 than in 1993. The 1994 maximum occurred during a two month long +0.8 degree anomaly. The 1993 maximum occurred during a +1.1 degree anomaly, but it occurred too early in the year to cause exceptionally high absolute water temperatures, while the 1994 anomaly occurred during what would normally have been the hottest months. In situ temperature measurements are available for some coastal bays which suggest satellite values are underestimates (Graphs 18A and 18B).

Funafuti, Tuvalu

Water temperatures were very constant in most of 1993, but had large warm swings in 1994. The warmest occurred during an anomaly of +1.1 degrees, and other temperature peaks occurred during +1.0, +1.0, and +1.2 degrees anomalies after it. It is not known if bleaching occurred in Tuvalu (Graphs 19A and 19B).

Tokelau

Water temperatures were the most constant of any site, but reached considerably higher values in 1994 during a +0.9 degree anomaly. Two +1.1 degree anomalies occurred during cooler times of both years. It is not known whether bleaching occurred in either year (Graphs 20A and 20B).

Uvea, Wallis & Futuna

Maximum water temperatures were considerably higher in 1994 than in 1993, and the highest values occurred during a +1.0 degree anomaly. It is not known if bleaching has occurred in 1994 (Graphs 21A and 21 B).

Vavau, Tonga

Maximum water temperatures were distinctly higher in 1994 than in 1993, and the 1994 maximum occured during a +0.8 degree anomaly. Anomalies were generally negative except for this period. It is not known if bleaching occurred in 1994 (Graphs 22A and 22B).

Lau, Fiji

Maximum water temperatures were distinctly higher in 1994 than in 1993, and the 1994 maximum occurred during a +0.5 degree anomaly. Anomalies were generally negative except for this period. It is not known if bleaching occurred in 1994 (Graphs 23A and 23B).

Canton, Kiribati

Temperatures reached higher levels in 1994 than in 1993. The 1993 maximum occurred during an anomaly of +1.4 degrees, and the 1994 maximum during a +1.8 degree anomaly. It is not known whether bleaching occurred in either year (Graphs 24A and 24B).

Christmas, Kiribati

The temperatures were generally higher in 1994 than in 1993. The 1993 maximum occurred during an anomaly of +1.1 degrees in the last month 1993, and the 1994 maximum in February during an anomaly of +0.9. Maximum temperatures occurred near the end of the year, and anomalies were very high in September. It is not known weather bleaching occurred in either year. In situ temperature measurements may be available (Graphs 25A and 25B).

Tarawa, Kiribati

Water temperatures reached higher values in 1994 than in 1993, during a two month anomaly of +1.2 and +1.3 degrees. A +1.1 degree anomaly took place during the cool season before. There have been no reports of bleaching in Tarawa (Graphs 26A and 26B).

Majuro, Marshall Islands

Maximum water temperatures were much warmer in 1994 than in 1993, coincident with a +1.1 degree anomaly. There have been no reports of bleaching in Majuro (Graph 27A and 27B).

Okinawa, Ryukyu Islands

Maximum water temperatures were considerably higher in 1994 than in 1993. Anomalies were very positive throughout the period, but were lowest in the 1993 warm season. A +2.0 degree anomaly occurred during the warmest month of 1994, and a +2.1 degree anomaly occurred as waters were warming up three months previously. Despite these high temperatures there have been no reports of bleaching from Okinawa (Graph 28A and 28B).

Maui, Hawaii

Water temperatures reached the same level in both years. Anomalies were very small throughout the entire period. No bleaching was reported despite heavy diving activity (Graph 29A and 29B).

Malindi, Kenya

Water temperature maxima were 29.2 degrees in 1993 and 29.1 in 1994. The 1993 warm period was very brief, while that of 1994 was much longer. Thermal anomalies varied greatly over the period, with a value of +0.9 during the warmest month of 1993. The warmest anomalies in 1993 occurred after maximum temperatures had declined. Maximum temperature anomales of +1.1 during November 1993 and February 1994 came just before and after the warmest period, greatly extending it. Bleaching was reported in 1994. The low dip in the middle of the hot season is suspicious because it could possibly be due to cloud cover causing satellite readings to be too low during hot and wet weather. Local climatic data is needed to resolve this (Graph 30A and 30B).

Zanzibar, Tanzania

Maximum water temperatures were 29.2 in April 1993, January 1994 and March 1994. The warm period was much longer in 1994 than in 1993. The January 1994 maximum coincided with the warmest anomaly, +1.2 degrees, versus +0.9 in 1993. Bleaching was reported in 1994 (Graph 31A and 31B).

Mayotte, Comoros

The maximum water temperature was 29.2 in February 1994, and 29.0 in April 1993. The largest anomaly, +1.1 degrees, took place in June 1993, while water temperatures were cool, and the next, +0.8, during the warmest month of 1994. There are no reports of bleaching available (Graph 32A and 32B).

Mauritius

Water temperature maxima were 28.0 in 1994, and 27.8 in 1993. The greatest anomalies, +1.4 and +1.0 both occurred after the maximum temperatures had passed, and so acted to prolong the warm period. There are no reports of bleaching from Mauritius (Graph 33A and 33B).

Mahe, Seychelles

Maximum water temperatures of 30.0 in 1994 were considerably above those of 29.3 in 1993, but the warm period was shorter in 1994. A thermal anomaly of +0.7 coincided with the warmest period recorded in 1994, and a higher anomaly of +0.9 with the maximum in 1993, but this took place at a time when water temperatures are normally declining, and so the absolute temperature was considerably lower despite the higher anomaly. Anomalies of +0.9, and +0.8 took place during cooler periods. There are no bleaching reports from Seychelles (Graph 34A and 34B).

Male, Maldives

Maximum temperatures were 30.1 in 1993 and 29.9 in 1994, both coinciding with anomalies of +0.6 or +0.7 degrees. Two anomalies of +0.9 took place during cool periods. No bleaching has been reported from the Maldives in this period (Graph 35A and 35B).

Ilha do Fernando de Noronha, Brazil

Ocean temperatures reached the same maximum value of 28.2 in both 1993 and 1994, but the warm period in 1994 lasted much longer than in 1993. Anomalies were mostly between 0 and +0.5 degrees, and the maximum anomaly of +0.6 took place in a cool month. Bleaching has not been reported at Fernando de Noronha (Graph 36A and 36B).

Recife, Pernambuco, Brazil

The water temperature maximum of 28.5 in 1994 was well above the value of 28.1 in 1993, and the warm period was much more extended. Temperature anomalies lay between -0.5 and +0.5 degrees, but rose throughout the period. The warmest period had an anomaly of +0.4 degrees. Bleaching in most coral species was reported in 1994 by Elga Mayal of the Universidade Federal do Pernambuco (Graph 37A and 37B).

Ilhas dos Abrolhos, Bahia, Brazil

Water temperature maxima were 28.0 in both 1993 and 1994, but the warm period lasted longer in 1994. Thermal anomalies were generally positive, and reached a maximum value of +1.1 degrees during the warmest period in January 1994. The other warmest period, March 1993, was marked by only a +0.5 degree anomaly. Bleaching therefore followed the highest positive anomaly, which coincided with a maximum water temperature somewhat earlier than normal. Bleaching in most coral species was reported by Clovis Barreira e Castro and colleagues from the Museu Nacional de Historia Natural in Rio de Janeiro (Graph 38A and 38B).

Arraial do Cabo, Rio de Janeiro, Brazil

Water temperature maxima reached the same value, 27.0 in both 1993 and 1994, but the 1993 warm period was more extended. Anomalies were generally positive with maximum values of +0.9 degrees occuring during both warmest months, March 1993 and February 1994. Bleaching was reported in 1994 by Ricardo Coutinho of the Navy School of Oceanography at Cabo Frio. Bleaching was not reported the previous year. Inshore water temperature data are available to see if there was a significant difference in these years (Graph 39A and 39B).

Ilha do Sao Sebastiao, Sao Paulo, Brazil

The maximum temperature recorded was 26.4 in February 1993, and the maximum was 26.0 in 1994. Anomalies were generally positive, with the greatest anomaly, +0.9 during February 1993. A large ocean hot spot lay offshore from southern Brazil for an extended period in 1993, but the 1994 hot spot moved past very quickly. Nevertheless bleaching was reported in 1994 but not in 1993. Adjacent land areas were very hot in early 1994, with record summer temperatures reported in Rio de Janeiro, making it possible that coastal cloudiness masked higher sea temperatures in the 1994 satellite data. In situ temperature measurements by Alvaro Migotto and Eurico Oliveira of the Universidade de Sao Paulo found very hot water temperatures of 30 degrees in Sao Sebastiao Channel. The OMS data presented here are for the ocean on the other side of the island, and are affected by strong offshore temperature gradients, cold coastal currents, and cloudiness. Local weather data is needed to determine if lower winds, waves, and currents could have caused greater coastal warming in 1994 than is indicated by satellite-derived offshore temperatures. In situ temperature data are available in Migotto, 1995 (Graph 40A and 40B).

The areas evaluated in this study include all major reef regions where mass bleaching was reported in 1993 and 1994. In addition mild bleaching events occurred in Jamaica in both years, but only affected 5-10% or less of all corals, i.e. only the most susceptible colonies of the most susceptible species. Water temperatures were measured in situ at Negril Jamaica, and although they reached values of 29.5 to 30.0 degrees Celsius, the critical values for Jamaica, they did not stay hot enough long enough to cause serious bleaching because of a series of frontal systems and tropical storms, including Hurricane Gordon in November 1994. Local bleaching was reported along in-shore areas of the Great Barrier Reef associated with hurricane fresh water flooding from land (DeVantier et al., 1995), and no major hot spot was seen in that area in the NOAA data.

Table 8 summarizes the information for all sites by showing the maximum monthly average water temperature each year, and the maximum anomaly in the warmest period (usually in the warmest month, but sometimes in the months on either side of it). Column M is maximum temperature (in degrees Celsius), A is temperature anomaly (in degrees Celsius), B lists bleaching reports. Reports of bleaching are indicated by (+), years when bleaching was not reported and where it is expected that it would certainly have been reported had it happened are indicated by (-), areas where there are no reports available are indicated by (?), and areas where the thermal anomalies in the warmest months are such that bleaching events which have not yet been reported are regarded as possible according to this analysis are indicated by (P). It is difficult to obtain accurate reports for many remote areas where there are few divers trained to recognize coral bleaching.

TABLE 8A. Maximum monthly average sea surface temperatures and coral bleaching reports for sites during 1993 and 1994, as displayed in Graphs 5 through 36.

SITE LOCATION M93 A93 B93 M94 A94 B94

PACIFIC OCEAN SITES:

Mangareva, Gambier 27.3 0.6 - 27.8 1.3 +

Mururoa, Tuamoto 27.7 0.8 - 28.2 1.1 +

Rangiroa, Tuamoto 29.2 1.1 + 29.3 1.0 +

Tahiti-Moorea, Society 28.8 0.4 - 29.2 1.1 +

Tubuai, Austral 27.7 0.5 - 28.4 1.1 +

Rurutu, Austral 28.1 0.6 - 28.6 1.2 +

Rarotonga, Cook 27.3 0.1 - 28.3 1.1 +

Aitutaki, Cook 28.6 0.4 - 29.2 1.1 +

Manihiki, Cook 29.4 0.4 - 30.1 1.0 +

Niue 28.3 0.3 - 29.7 1.1 +

Tutuila, Samoa 29.3 0.9 - 29.9 0.8 +

Funafuti, Tuvalu 30.1 1.1 ?P 29.9 1.0 7P

Tokelau 29.8 0.8 ? 30.0 0.9 ?

Uvea, Wallis & Futuna 29.2 0.1 ? 30.1 1.0 ?P

Vavau, Tonga 28.0 -0.1 ? 28.9 0.8 ?

Lau, Fiji 28.2 -0.1 ? 28.8 0.5 ?

Canton, Kiribati 29.6 1.3 ?P 29.7 1.8 ?P

Christmas, Kiribati 28.8 0.5 ? 29.0 0.9 ?

Tarawa, Kiribati 29.9 1.0 ?P 30.2 1.3 ?P

Majuro, Marshall 29.4 0.4 - 29.7 1.1 ?P

Okinawa, Ryukyu 28.1 0.6 - 29.1 2.0 ?P

Maui, Hawaii 26.6 0.1 - 26.6 0.3

TABLE 8B. Maximum monthly average sea surface temperatures and coral bleaching reports for sites during 1993 and 1994, as displayed in Graphs 5 through 36.

SITE LOCATION M93 A93 B93 M94 A94 B94

INDIAN OCEAN SITES:

Malindi, Kenya 29.2 0.9 - 29.1 1.1 +

Zanzibar, Tanzania 29.2 0.9 - 29.2 1.2 +

Mayotte, Comoros 29.0 0.4 ? 29.1 0.8 ?

Mauritius 27.8 0.3 ? 28.0 1.0 ?P

Mahe, Seychelles 29.3 0.9 ? 30.0 0.7 ?

Male, Maldives 30.1 0.7 ? 29.9 0.5 ?

ATLANTIC OCEAN SITES:

Fernao de Noronha, Brazil 28.2 0.1 ? 28.2 0.2 ?

Recife, Brazil 28.1 -0.2 - 28.5 0.4 +

Abrolhos, Brazil 28.0 0.5 - 28.0 1.1 +

Arraial do Cabo, Brazil 27.0 0.9 - 27.0 0.9 +

Sao Sebastiao, Brazil 26.4 0.9 - 26.0 0.5 +

Graph 5. Differences between 1993 and 1994 temperatures.

Temperature differences between 1993 and 1994: At 24 of 32 sites (75%) maximum temperatures were higher in 1994 than in 1993, in 4 sites (12.5%) they were the same, and in 4 sites they were lower (Graph 5). Only one site reported significant bleaching in 1993, but there were 17 cases of bleaching reported in 1994. The difference is correlated with an overall rise in maximum temperature between the two years of 0.364 + 0.411 degrees Celsius (mean + standard deviation). Although most sites were notably warmer in 1994, the variability exceeds the mean increase. However the average temperature at all sites is very strongly statistically higher than a population of the same number of sites with no temperature change from year to year (P < 0.001).

Although the population of sites examined was significantly warmer in 1994 than in 1993, no implication can be made from such short term data regarding longer term or global trends because of insufficient sample size. Even though the sites chosen cover large portions of the Pacific, Indian, and Atlantic Oceans, sample sites were predominantly from near tropical regions affected by bleaching and they are not necessarily representative of the world as a whole. Nevertheless the fact that such a large increase in bleaching is correlated with such a small temperature increase suggests that any future global warming could cause more severe bleaching.

Bleaching and maximum water temperatures: Cases of bleaching and non bleaching conditions, when plotted against maximum water temperature reached, overlapped considerably (see Graph 6A). Note that in this figure cases are grouped by 0.5 degree increments of temperature values above that marked, i.e. 25.0 shows all values from 25.0 to 25.5. Cases of no-bleaching were reported for temperature maxima between 26 and 29 degrees, and cases of bleaching were reported for maxima between 26 and 30 degrees. A very large proportion of sites where there are no in-situ bleaching reports available also reached maximum temperatures of above 30 degrees Celsius. This suggests that the sites of inadequate information are disproportionately the hottest sites of all.

Graph 6A. Coral reef bleaching reports and maximum sea surface temperatures.

Graph 6B. Coral reef bleaching reports and maximum SST anomalies.

Sites which did not bleach had a mean maximum temperature of 28.055 + 0.978 degrees, sites which bleached had higher mean maximum temperatures of 28.65 + 1.026 degrees, and samples where no bleaching information was available had still higher mean temperature maxima of 29.236 + 0.758 (Graph 6B). Although higher, the mean temperature at bleaching sites was not statistically significantly different than the non-bleaching sites (P = 0.076). Temperatures at sites with no bleaching information were significantly higher than at bleaching sites (P = 0.031), and they were very strongly significantly higher at sites with no information than at non-bleaching sites (P < 0.001). These results are summarized in Table 9.

Based on detailed study of sea surface temperatures in the Caribbean, it was found that bleaching occurred at hotter temperatures at sites with higher mean annual temperatures than at colder sites, and that at all sites bleaching took place after warm season anomalies of around +1.0 degrees Celsius or more, regardless of the absolute value of the local bleaching threshold temperature (Goreau et al., 1993). This suggested that temperature anomalies in the warm season are better predictors of bleaching than the actual maximum temperature itself, as long as the anomaly takes place at the right season, and has been confirmed using OMS data from the Pacific and Indian Oceans (Goreau and Hayes, 1994).

Bleaching and warm season temperature anomalies: As predicted, a graph of bleaching and non bleaching cases plotted against warm season sea surface temperature anomalies leads to a nearly sharp division between the categories (Graph 7A) which is not seen when bleaching is plotted against absolute maximum temperatures themselves. All cases where no bleaching was reported where it would have been reported had it happened numbered 19, and all of these sites had a warm season anomaly of +0.9 degrees or less. In contrast four cases where bleaching was reported had warm season anomalies of +0.9 degrees or less, and 15 cases had anomalies of +1.0 degrees or more. In general an anomaly of around +1.0 degrees or more in the warm season appears to be a good predictor of bleaching, since the absence of bleaching was never confirmed to have taken place above this temperature anomaly value. However a small minority of cases of bleaching took place below this value, at anomalies of +0.9, +0.8, +0.5, and +0.4. Of 28 cases where it is unknown whether bleaching occurred, 18 had anomalies of +0.9 degrees or less and 10 cases had anomalies between +1.0 and +2n

One possibility is that the four apparently anomalously low bleaching sites have corals with locally lower temperature sensitivities than other sites, another is that bleaching was caused there by local factors other than high temperature. The anomalies in the first two sites are very close to the 1.0 value, and may be cases of local underestimation of sea surface temperature due to local cloudiness. The first site is near Cabo Frio, Brazil, where intense local upwelling, a strong temperature gradient to offshore cold currents, and local cloudiness could all cause satellite values to be too low. The second site is Tutuila, Samoa, a high island which generates intense local cloudiness which would cause large-scale satellite measurements to underestimate the actual values. In-situ measurements taken there at the time by Peter Craig were indeed higher than those seen in satellite data. The final two sites are from the coast of Brazil, and it is very likely that these two are strongly affected by cloudiness due to air rising over steep coastal mountain ranges and by strong offshore thermal gradients due to cold currents. It is worth noting that in-situ measurements at Sao Sebastiao Channel reached 4 degrees higher than the offshore values estimated here from satellite data (Migotto, 1995).

Sites at which bleaching did not occur had mean warm season anomalies of +0.515 + 0.312 degrees, only half as much as at sites where bleaching occurred which had mean warm season anomalies of + 1.011 + 0.232 degrees. Sites where no bleaching information was available had intermediate mean warm season anomalies of +0.768 + 0.507 (Graph 7B). In contrast to the pattern seen with temperatures, known bleaching sites had clearly higher anomalies. While there was no statistically significant difference between sites where no information was available and either sites with observed bleaching (P = 0.065) or with no observed bleaching (P = 0.054), bleaching sites had very strongly significantly higher temperature anomalies than non-bleaching sites (P < 0.001). See Table 9. .

Graph 7A. Coral reef bleaching and maximum monthly average SST. The white bar indicates that reports of bleaching have been received from the sites; the black bar indicates that reports suggest no bleaching a the sites: and the stippled bar indicates that information is unknown from the sites.

Graph 7B. Coall reef bleaching and maximum warm season SST anomalies. The white bar lndicates that reports of bleaching have been received from the sites; the black bar indicates that reports suggest no bleaching at the sites; and the stippled bar indicates that information ls unknown from the sites.

These results confirm that warm season high temperature anomalies, rather than absolute maximum temperatures, are significantly related to the presence or absence of bleaching. In many cases, especially in the Indian Ocean, anomalies larger than 1.0 degrees occurred during cool seasons of the year, but these were not followed by bleaching. The broad range seen in the warm season anomalies at sites where no information is available suggests that this population of sites is made up of two distinct sub-populations, one with small anomalies and one with large anomalies. Based on comparison with sites where information is available, this suggests that the former subpopulation did not have bleaching and the latter did. Using a value of 1.0 degrees as the cut-off, we would expect that 18 of the cases with values of +0.9 degrees or less where there were no field reports available probably did not have bleaching, and 10 cases where the anomaly was +1.0 degrees or more probably did have bleaching which has not yet been reported. These sites included many remote islands which are of extreme environmental importance because they lie in areas which are not regularly monitored but which may be especially sensitive to climatic variations. We are making efforts to contact divers in all of these areas to see if there are observers who can confirm or deny the projections made from this data, however these areas generally have few divers trained to identify coral bleaching.

This analysis strongly confirms that spatial proximity to sea surface temperature anomalies of around +1.0 degrees celsius are a strong predictor of mass coral bleaching only when they happen in the warmest months (Goreau and Hayes, 1994). Positive temperature anomalies ("hot spots") that occurred during times of cold seasonal water temperature were never linked with bleaching, no matter how large they were. Such anomalies were especially common in the Indian Ocean during this period. The results support the suggestion that only anomalies of around +1.0 degrees or more in the warmest months in coral reef areas need to be tracked in real time (Goreau and Hayes, 1994). Such anomalies may increase in coming years due to long term global warming trends and the end of the temporary global cooling caused by the eruption of Mount Pinatubo on June 15, 1991.

Some sites were affected by bleaching at slightly lower anomalies (+0.8 to +0.9 degrees, and even less), unless the estimated anomaly values at those sites are too low. Detailed comparison with in situ records is needed to resolve this. Calibrations of satellite data against in-situ data in the Caribbean showed that the satellite data gave increasing underestimates as temperatures rose to the highest levels (Goreau and Hayes, 1995). This would be expected if high ocean temperatures promoted small scale thermal convection and cumulus cloud formation, leading to increasing systematic temperature underestimates during warmest conditions. In situ calibration to determine possible systematic and site specific biases in the satellite data is therefore an important scientific priority. The results from in situ monitoring need to be compared to satellite data at higher time and space scales (Strong, in preparation). Satellite data should be made available in real time to specialists tracking global bleaching events who are in close contact with divers, researchers, and environmental monitoring agencies around the world.

TABLE 9. SUMMARY OF TEMPERATURE STATISTICS.

Statistical rankings are as follows: A statistically significant correlation is one in which the probability, P, is < 0.05, or 95% probability not random (*); a strongly significant correlation is one in which P < 0.01, or 99% confidence (**); and a very strongly significant correlation is one with P < 0.001 or 99.9% confidence (***). The probability values were computed from Student's t test.

TEMPERATURE CHANGE FROM 1993 TO 1994:

For the population of all sites maximum monthly average sea surface temperatures in 1994 were very strongly significantly warmer (***) than in 1993 (P < 0.001 ).

BLEACHING AND MAXIMUM WATER TEMPERATURES:

Sites where it was unknown whether bleaching had occurred had maximum monthly averages that were very strongly significantly higher (***) than at sites where bleaching did not take place (P < 0.001), and significantly higher (*) than at sites where bleaching did take place (P = 0.031). Sites where bleaching took place had higher mean maximum temperatures than sites where bleaching did not happen, but this was not statistically significant (P = 0.076).

BLEACHING AND WARM SEASON TEMPERATURE ANOMALIES:

Sites where bleaching took place had very strongly significantly higher (***) positive temperature anomalies than sites where bleaching did not take place (P < 0.001). Sites where it was not known whether or not bleaching took place had mean maximum anomalies less than at bleaching sites but more than at non-bleaching sites, but were not statistically significantly different than either (P = 0.065 and 0.054 respectively).

General Introduction to Graphs 8-40

(Pages 134-165)

These graphic displays are derived from Oceanographic Monthly Summary (OMS) data and are presented in our report to demonstrate the monthly variation in sea surface temperature (SST) and anomalies in sea surface temperature (SST Anomalies) at several selected tropical island nation sites during the twenty-one month interval of January, 1993 through September, 1994. These data are from the Advanced Very High Resolution Radiometer (AVHRR) band of the Tiros Orbiting Satellites deployed by the National Oceanographic and Atmospheric Administration of the U.S. Department of Commerce. Applied in conjunction with the site observations and reports we have gathered, associations may be made to establish the correlation in both time and space between thermal characteristics of the environment and the response of bleaching of reef organisms. The consistency of these associations allows us to appreciate the sensitivity of the coral reef to extremely stressful swings in this one environmental variable. Furthermore, the association of SST anomalies in the range of 1.0 degree Celsius or more with bleaching supports the interpretation that there may well be a causal relationship between elevated temperature exposure and coral reef bleaching. These data also should allow us to identify the existence of any acclimation by scleractinian corals which would suggest that continued future exposures to elevated temperature might result in adaptive responses sufficient to preserve the critical symbiotic balance between reef-building corals and dinoflagellate algae.

Graph 8. OMS data for Mangareva, Gambier Islands. A. SST; B. SST Anomaly

Graph 9. OMSdataforMuroroa, Tuamotu Islands. A. SST; B. SST Anomaly

Graph10. OMS data for Rangiroa,Tuamoto Islands. A. SST; B. SST Anomaly

Graph 11. OMS data for Tahiti, Society Islands. A. SST; B. SST Anomaly

Graph 12. OMS data for Tubuai, Austral Islands. A. SST; B. SST Anomaly

Graph 13. OMS data for Rurutu, Austral Islands. A. SST; B. SST Anomaly

Graph 14. OMS data for Rarotonga, Cook Islands. A. SST; B. SST Anomaly

Graph 15. OMS data for Aitutaki, Cook Islands. A. SST; B. SST Anomaly

Graph 16. OMS data for Manihiki, Cook Islands. A. SST; B. SST Anomaly

Graph 17. OMS data for Alofi, Niue. A. SST; B. SST Anomaly

Graph 18. OMS data for Tutuila, American Samoa. A. SST; B. SST Anomaly

Graph19. OMS data for Funafuti,Tuvalu. A. SST; B. SST Anomaly

Graph 20. OMS data for Tokelau. A. SST; B. SST Anomaly

Graph 21. OMS data for Uvea, Wallis and Futuna. A. SST; B. SST Anomaly

Graph 22. OMS data for Vavau, Tonga. A. SST; B. SST Anomaly

Graph 23. OMS data for Lau, Fiji. A. SST; B. SST Anomaly

Graph 24. OMS datafor Canton, Kiribati. A. SST; B. SST Anomaly

Graph 25. OMS data for Christmas Island, Kiribati. A. SST; B. SST Anomaly

Graph 26. OMS data for Tarawa, Kiribati. A. SST; B. SST Anomaly

Graph 27. OMS data for Majuro, Marshall Islands. A. SST; B. SST Anomaly

Graph 28. OMS data for Okinawa, Ryukyu Islands. A. SST; B. SST Anomaly

Graph 29. OMS data for Maul, Hawaii. A. SST; B. SST Anomaly

Graph 30. OMS data for Malindi, Kenya. A. SST; B. SST Anomaly

Graph 31. OMS data for Zanzibar, Tanzania. A. SST; B. SST Anomaly

Graph 32. OMS data for Mayotte, Comoros. A. SST; B. SST Anomaly

Graph 33. OMS data for Mauritius. A. SST; B. SST Anomaly

Graph 34. OMS data for Mahe, Seychelles. A. SST; B. SST Anomaly

Graph 35. OMS data for Male, Maldives. A SST; B SST Anomaly

Graph 36. OMS data for Fernando de Noronha, Brazil. A. SST; B. SST Anomaly

Graph 37. OMS data for Recife, Brazil. A. SST; B. SST Anomaly

Graph 38. OMS data for Abrolhos, Bahia, Brazil. A. SST; B. SST Anomaly

Graph 39. OMS data for Arraial do Cabo, Brazil. A. SST; B. SST Anomaly

Graph 40. OMS data for Sao Sebastiao, Brazil. A. SST; B. SST Anomaly

CONCLUSIONS

Satellite-derived measurements of sea surface temperature anomalies were successfully used in near real time to track the spatial distribution and timing of the 1994 mass bleaching events in the South Pacific (Goreau, Hayes, and Strong, submitted). This was the first major bleaching event whose time and location was specifically predicted from remote sensing data, confirming the validity of the relationship between bleaching and sea surface temperature anomalies derived from study of global records between 1983 and 1991 (Goreau and Hayes, 1994). In addition this data base also revealed separate temperature anomalies which caused mass coral bleaching at the same time along the South Atlantic coast of Brazil and Indian Ocean shorelines of Kenya, Zanzibar, and Tanganylka. As this volume went to press in November 1995 a similar mass bleaching event was being tracked across most of the Caribbean.

Our observations, and information derived by interviews with divers in the South Pacific showed that all areas affected by this anomaly underwent bleaching, and that a substantial proportion of many bleached corals died. No reports of mass bleaching were received from reefs outside of areas affected by hot spot anomalies. Percentage of corals bleaching appears to be closely related to the magnitude and duration of the thermal anomaly more than 1.0 degree Celsius above average in the preceding warm season, but was not related to any other environmental or ecological stress parameter. Where local water temperatures were available, these confirmed the high values seen in NOAA global satellite derived sea surface temperatures.

In general, reefs in American Samoa had the lowest live coral cover, followed by lagoonal reefs in Tahiti and Moorea. The highest coral cover was found on the outer slope reefs in the southern Cook Islands. Reefs in lagoons and protected bays had lower live coral cover than reefs on outer slopes at every island except Rangiroa (where lagoon conditions are very similar to those in the outer slope due to excellent circulation). However, there was much less bleaching visible in lagoons.

Increasing pollution and human population density correlated with decreasing coral cover, increasing algae cover, decreasing dissolved oxygen content, and increasing human population density, but was unrelated to natural damage from storm waves or Acanthaster. Polluted areas appeared to be much less able to recover from natural hurricane and storm physical damage. While natural or human-caused physical damage and pollution was present in every reef, and had strong effects on coral and calcareous algae cover, the stress patterns and histories were unique to each site, and were not related to bleaching.

Experienced local observers interviewed at each site included the first established commercial dive operators on each island. When they first encountered bleaching each at first thought that Acanthaster was responsible, but they subsequently rejected this possibility based on their observations of the distribution of Acanthaster and clear differences in the patterns of white coral caused by bleaching and by predation. Each dive operator independently stated that in their experience no known stress could explain bleaching except for long calm periods of unusually high water temperatures, or "doldrums". While professional divers and local environmental officials at each site regarded coral bleaching as a new and unprecedented phenomenon, there appeared to be insufficient public awareness of its importance with regard to current and potential deterioration of reefs.

Evidence of genetic variability in susceptibility to bleaching was seen in many coral species at the local scale, with some colonies that were unbleached, partially bleached, fully bleached, dying, or recently dead. On the regional scale genetic variability was revealed by some curious patterns in which certain genera appeared to have different relative bleaching susceptibility in different areas. For example, Acropora bleached more readily than Pocillopora in French Polynesia, but the opposite in the Cook Islands. The pattern of recovery of Porites in French Polynesia and the Cook Islands showed pale patches which gradually recovered their color, but in Samoa bleached patches remained pure white and were filled in from the edges by fully pigmented tissue. It is not known how genetic variation in bleaching responses is partitioned between host corals and their symbiotic algae populations.

RECOMMENDATIONS

RAPID DISSEMINATION OF CLIMATE ANOMALY DATA

The close linkage between sea surface temperature anomalies and coral bleaching allows mass bleaching events to be reliably predicted from climatic data (Goreau and Hayes, 1994) once NOAA's global satellite-derived sea surface temperature data base is made available in real time at the highest spatial and temporal resolution. At the present time coral bleaching researchers must use low spatio-temporal resolution data published more than a month afterwards. An alternative is to allow the entire data base to be accessed on the internet, but there is a risk that the data will simply pile up without being analyzed and converted into a predicatively useful format except after the impacts are underway or over. To be most useful it is essential that the data be analyzed and interpreted by coral reef scientists familiar with the phasing, duration, and distribution of thermal anomalies with regard to seasonal temperature cycles in reefs around the world.

We recommend that NOM work in close partnership with field researchers by providing near real-time high resolution data directly to scientists who are tracking global anomalies, are familiar with the location and ecological conditions of reefs worldwide, and with the biological impacts of such anomalies on the entire range of reef organisms. This will allow the data to be rapidly analyzed with regard to the possible impacts on reef ecosystems and the relevant early alert information passed on to observers in affected locations. The Global Coral Reef Alliance would form a suitable partner for NOAA in such an endeavor because of its work on establishing the climatic thresholds for bleaching and links with coral reef protection groups and researchers around the world.

RAPID FIELD RESPONSE TEAM

The biological impact of bleaching events, in particular coral mortality or recovery, are extremely sensitive to the duration and intensity of thermal anomalies, as well as the presence of local stresses to reef ecosystems. Both are profoundly affected by local variations in water circulation over reefs, which may accentuate or diminish the impacts since water temperatures in reefs can be warmer or colder than the open ocean values which are measured by satellites, due to changes in water exchange by local runoff, tides, currents, or sea level variations. The actual biological impact of bleaching on reefs can only be determined by direct inspection of reefs by coral scientists intimately familiar with the ecology of the species affected and able to identify the impacts of all local reef stresses. Most documentation of bleaching events have come only after the events are nearly over, thereby resulting in serious underestimates of the extent of bleaching mortality, or recovery.

In order to adequately document the impact of bleaching events on coral reef ecosystems we recommend that a rapid response field team of marine researchers be formed with the support needed to get rapidly out to affected areas to document the timing, duration, extent, and intensity of bleaching, and identify species affected, both at the first signs of coral reef bleaching and again 6-12 months later. By photographic and video documentation of specific sites it will be possible to determine differential species rates of mortality and recovery, to establish the effects of the duration and intensity of thermal anomalies, and determine whether adaptation is taking place.

We recommend that a worldwide network of coral reef bleaching researchers be established to ensure that all major events are adequately documented in the field, with funding to undertake rapid emergency field work on short notice. While the existing network of diving operations is able to access reefs near populated areas, many reefs which are being affected are remote and cannot be practically reached from such locations. It is therefore especially important to develop mobile, rapidly deployable ship-based observation facilities able to get quickly to the most important areas affected regardless of location. In particular we recommend that the rapid response bleaching team proposed by the Global Coral Reef Alliance (GCRA) be funded to receive real time climate data, coordinate field work, and communicate with local divers, and that this be coupled to the proposal to maintain a continuously available ship-based research platform capable of rapid response to any event, as proposed by the Remedial Eco-toxicological Expedition Fund (REEF).

PUBLIC EDUCATION

The importance of coral reef protection to sustaining marine biodiversity, fisheries, tourism, and shore protection are inadequately understood by coral reef users, coastal populations, policy makers, and the public at large whose actions indirectly or inadvertently place additional stress on reefs. Only an informed public in coral reef countries can protect the biodiversity of reefs in their own "back yards" as a vital and growing resource for their children's children (National Research Council, 1995).

We recommend greatly expanded funding to produce and disseminate written and visual material for public education in schools, at the community level, and through the mass media. We especially recommend development of advanced materials based on the Coral Reef Care Manual for the Philippines produced by Ocean Voice International and the Haribon Foundation, which contains cartoons and information accessible to fishermen and coastal subsistence communities, and the production of documentary films, videos, and CD's on coral bleaching, as well as visual guides to reef stresses showing their symptoms and causes. We also recommend the development of resource handbooks such as the Coral Reef Monitoring Guide published by UNEP (Dahl, 1984) and by the U.S. National Park Service (Rogers, 1994). Where divers, boaters, and tourists are important sources of reef stress, educational materials aimed at teaching "reef etiquette" to visitors need to be widely disseminated. We recommend these be based on materials produced by Reef Relief and other reef protection and diving groups.

LOCALLY-BASED REEF MONITORING

The impacts of bleaching events on coral reef health are best understood where the long term history of reef stresses are well known. We recommend the establishment of coral reef monitoring networks based in the community which is under the water most often, namely the sport dive operators. While detailed scientific monitoring of reefs is needed, there are dozens or hundreds of sport dive operations for each national coastal zone management program or marine lab with systematic long term coral reef research programs. They are capable of far more widespread, frequent, and sustained monitoring efforts. Dive operators should be educated to look out for and document changes in the reefs which they regularly dive on, and linked into an international network to store, analyze, and disseminate such information.

We recommend that a world-wide network be developed, based on the sport dive operators video reef monitoring program established by the Negril Coral Reef Preservation Society in Jamaica. This program works with local marine video producers to obtain copies of footage documenting reef conditions which are taken during their normal operations. No additional time commitment is required on their part except for duplication of footage. The images will be stored on CD/ROM to allow an instantly cross-comparable long term library of reef health. Along with the collection and storage of field data, rapid transfer of information about events in the marine environment should be effected through the establishment of an international computer bulletin board on the Internet World Wide Web.

BASIC AND APPLIED CORAL RESEARCH

There are many fundamental research issues which need to be resolved that have direct implications for reef management. Although current knowledge suggests that corals are unable to adapt to high temperature stresses, as indicated by failure of corals to recover and recolonize areas which have been thermally stressed by power generating or desalinization plant effluents (Jokiel and Coles, 1992), and observations that corals which live in highly stressed habitats require higher temperatures to bleach, but recover more slowly (Goreau, 1992), much more work needs to be done. Ecological observations show that there is a range of susceptibility in most species, and although there is clear evidence that this involves genetic variability in both the corals and their symbiotic algae, however there is inadequate understanding of the genetic, biochemical, or cellular bases of this variation. Resolving these issues is central to understanding whether natural selection or genetic manipulation can produce varieties capable of higher temperature and stress tolerance. This work requires greatly expanded funding for basic and applied research on corals and their symbiotic algae, which requires controlled culture of corals and symbiotic algae under first rate aquarium facilities with extremely clean and rapidly exchanging water, as well as culture collections of the full range of different symbiotic algae species. A large number of different species and genetic strains of these exist (Trench, 1993; Banaszak et al., 1993; McNally et al., 1994; Trench and Long, 1995), with extremely different ranges of temperature tolerance (Iglesias-Prieto et al., 1992). Only a handful of laboratories in the world are capable of maintaining the most sensitive species of corals under viable conditions, and the major collection of symbiotic algae is under serious risk of being lost because of lack of funding for its maintenance (Trench, personal communication).

We recommend that focus be placed on a) the effects of genetic and species variation of corals and symbiotic algae on susceptibility to bleaching, b) the role of multiple stresses on the susceptibility to bleaching and on rates of mortality or recovery, c) the effects of repeated bleaching, d) the potential for adaptation. We therefore recommend that research efforts focus on establishment of facilities for maintaining and growing the widest varieties of coral and symbiotic algae species, involving the highly successful programs at the Oceanographic Museum in Monaco, and that a major program to save and expand the Trench symbiotic algae collections and to fully utilize them in research on the genetic and physiological basis of temperature tolerance be funded. In the current absence of suitable laboratory facilities near a coral reef, we recommend the collection of corals which are relatively resistant to bleaching for detailed laboratory experimentation at facilities like those in Monaco.

LOCAL MANAGEMENT MEASURES

Coral reef countries need to urgently enact coral reef and coastal zone management programs which control damage to reefs caused by human activities in the water, in particular destructive or non-selective fishing practices such as use of dynamite, poisons, chemicals, or smashing corals to extract fish or shellfish, extraction and sale of "live rock" or of reef limestone, overfishing, anchor damage, dredging, and divers touching corals.

We recommend expansion of public education and lobbying for suitable legislation based on the existing work of organizations such as Reef Relief, Project Reef keeper, and numerous coral reef protection and diving organizations. We recommend expanded funding for development of viable low cost mariculture of reef species to replace overfishing of wild stocks and allow the establishment of fish sanctuaries to permit breeding stocks to recover. We recommend enactment of coastal zone management regulations which ban destructive and non-selective fishing methods and funding for their effective enforcement, following the model management of protected reefs in the Cayman Islands and Bonaire, as well as a restriction of the international trade in all reef species. Such efforts at the policy making level should be a major initial focus of the International Coral Reef Initiative.

REGIONAL MANAGEMENT MEASURES

Regional stresses to coral reefs include land and ship based sources of stresses to reefs which are caused by human activities outside the reef waters themselves. Of critical concern are large scale damage to reefs caused by coral overgrowth by weedy algae which are being fertilized by sewage runoff from land as well as fertilizers leached from inefficient usage in agriculture, and turbidity and sedimentation of soils washed away from terrestrial watersheds as the result of deforestation and unsound soil conservation practices. Coral reef countries urgently need to strengthen their existing environmental protection agencies and research institutions to effectively monitor and protect their terrestrial and marine natural resources.

We recommend that the International Coral Reef Initiative focus on a) adoption of coastal zone water quality criteria requiring that waters in coral reef areas be maintained below the critical levels of nitrogen, phosphorus, and chlorophyll scientifically shown to be necessary to prevent coral reefs from being overgrown by nuisance algae, which have recently been established (Bell, 1992; Lapointe and Clark, 1992), b) the use of biological tertiary sewage treatment or use of sealed dry toilets to prevent leaching of sewage derived nutrients into the coastal zone in all coral reef areas with population densities sufficiently high to cause elevation of coastal zone nutrient levels (Goreau and Thacker, 1994), c) linking coral reef conservation efforts to whole-watershed management to control erosion of soil and leaching of fertilizer into surface waters and groundwaters through appropriate management of agriculture and reforestation of hillsides in coastal watersheds, as in proposed community-based whole-watershed and coastal zone management plans developed by the Negril Environmental Protection Trust and the Port Antonio Marine Park and Forest Corridor Project in Jamaica, d) banning discharge of garbage, chemicals, petroleum hydrocarbons, toxic wastes, and radioactive materials on the high seas, and e) obligatory use of shore reception and processing facilities in which these materials are processed and stored for safe land-based disposal in sealed, lined land fills with complete tertiary treatment of liquid effluents.

INTERNATIONAL MANAGEMENT MEASURES

The primary cause of the unprecedented mass bleaching events which have taken place only since the 1980s has been shown to be elevated sea surface temperatures. The response of ecosystems to extreme weather anomalies provide valuable clues to their response to climate change. It is therefore critical that the Framework Convention on Climate Change be amended in its protocols to ensure that the acceptable rate of climate change be within the tolerance limits of the most climatically sensitive ecosystems, and that these ecosystems be monitored on a world-wide scale for evidence of climatically-induced stress (Goreau and Hayes, 1994). Only through such monitoring may the guidelines for acceptable global change be continuously modified in light of the best available scientific data and be kept within the tolerance limits of critical ecosystems like coral reefs.

Coral reefs are now known to be the most sensitive ecosystem to excessive levels of sediment, nutrients, temperature, and sea level rise. In view of the critical importance of coral reefs for the biodiversity, fisheries, tourism, and shore protection of over 100 countries, we recommend that this be a major focus of the International Coral Reef Initiative. Without suitable efforts to restrain rates of global warming and sea level rise all other efforts to conserve coral reefs will ultimately prove unsustainable. This, in turn, will yield disastrous consequences for the economies of coral reef countries and the species and people who require healthy coral reefs for their very survival.

GLOSSARY OF TERMS

Acanthaster: Acanthaster planci, the "Crown of Thorns" starfish, is a coral predator which moves in swarms that kiIl large parts of the coral on reefs which they cross. It is large and beautiful, coming in many colors, and is dangerous, being protected by long venomous spines. The mode of feeding of the starfish, extruding its stomach over coral tissue and digesting it, was first described in Eritrean Red Sea reefs (Goreau, 1964b) before Acanthaster first become a major problem in the Pacific Ocean in the 1960s. In most places where mass coral bleaching has taken place local observers at first suspected the white, seemingly dead, coral was the result of fresh outbreaks of Acanthaster, but subsequently rejected this hypothesis based on field observations showing the two phenomena were not linked.

Alcyonarians: Fleshy "soft" corals of the order Alcyonaceae, with eight-fold symmetric tentacles in their polyps, lacking a solid limestone skeleton, but containing tiny limestone crystals, or spicules, in their tissue. Alcyonarians are encrusting organisms, not reef framework builders. They tend to be more frequent in disturbed habitats, and to replace corals in reefs exposed to excessive nutrients (Bell, 1992). Many have symbiotic algae, and are capable of bleaching.

Algal turf: Bottom covered by fine fuzzy filamentous algae mats from a millimeter to centimeters thick. Microscope examination is needed to identify the many microscopic algae and bacteria species present, which cannot be done in the field. A fine fuzzy green algae turf tends to grow over dead coral skeleton within days to weeks, for example after Acanthaster kills them. If nutrient levels are low the algae turf is replaced within months by encrusting hard calcareous algae, but if nutrients are high the turf is overgrown by larger fleshy algae.

Anemones: Soft bodied relatives of the corals, but lacking any form of skeleton. Anemones open up their tentacles like flowers to catch zooplankton and fish, except for the symbiotic clown anemone fish, which is impervious to the anemone's stinging cells. Some anemones, including the Heteractis which host the clownfish, also have symbiotic algae within their tissues, and can bleach. Bleached anemones lose the bright colors at their bases and turn white.

Ascidians: Fleshy, colonial marine animals which filter seawater. They are not related to the corals, having a more complex gut and nervous system. Some encrusting ascidians in the family Didemnidae have symbiotic algae in their tissues, and members of this group with white bleached patches were seen. Generally rare, they are more abundant in high nutrient environments, and were most common in Samoa.

Atoll: Ring-like arrangements of coral reefs and sandy islands common in the Indo-Pacific. Atoll reefs are free of excessive terrestrial sediment runoff which can damage reefs near high islands, and because the flat flow islands generate no local convectional rainfall, many have very limited water supplies, and hence only minor human populations and human disturbance to reefs. Coral reefs on the insides of atolls vary tremendously depending on the rate at which lagoon water exchanges with the open ocean. Fully or nearly enclosed lagoon reefs may be dead or dying, while large lagoons with many open passages have reefs that are similar to those on the outside of the atoll. Charles Darwin (1842) correctly recognized that atolls marked the trace of submerged volcanic islands when he had the chance, on the voyage of the Beagle, to see the full range of high round volcanic islands with minor reef fringes, older islands in which the volcanic material was progressively eroded down while being surrounded by a growing reef rampart, low islands with only minor vestiges of volcanic material, and pure coral rings with no surface trace of the island's original material (see Loubersac, 1992). His controversial theory was confirmed nearly a century later when deep drillings into atolls reached volcanic rock. Atolls are now even better understood in the light of plate tectonic theory. Submarine volcanos are steadily submerged as the ocean crust on which they ride moves away from submarine spreading centres, heading for the deep trench subduction zones into which they will ultimately disappear. As the crust ages, it cools and becomes more dense, seeking deeper beneath the waves. Only the unique ability of corals to grow a limestone wall upwards as fast as the volcano sinks downwards provides a ring trace marking the spot of longvanished tropical volcanic islands. The islands whose reefs were examined in this study provide a neat continuum from young volcanoes with a narrow coral fringe to fully mature atolls. Their relative age sequence, from young to mature, is: Tutuila, Tahiti, Moorea, Rarotonga, Aitutaki, Rangiroa.

Bleaching (Coral bleaching; coral reef bleaching): Marine animals with internal symbiotic algae that have lost their algae are described as "bleached" because their tissues turn transparent, revealing the white skeleton beneath, or turn to pale or unusual colors. Corals which have been exposed to bleaching agents are dead and white because the surface tissue is removed by chemical oxidation. "Bleached" corals, in contrast, have live and intact tissue despite their white appearance, but they are in a state of starvation and unable to deposit their limestone skeleton, and less able to cope with other stresses such as excessive sedimentation or overgrowth by algae, soft corals, sponges, ascidians, and other organisms. Symbiotic hard corals, soft corals, anemones, zoantharians, sponges, and ascidians are all affected by bleaching, but there is great variability between and within species, indicating a genetic component of bleaching sensitivity in the symbiotic algae or the host organism. Bleached corals may recover completely over periods ranging from days to a year, depending on the severity and duration of bleaching. If the stress is severe, many or even most of the corals may die (Glynn and Colgan, 1992). For example, corals in the Galapagos Islands were virtually entirely killed following the severe 1983 bleaching episode, when temperatures reached over 6 degrees C above normal (Glynn, 1990).

Calcareous algae: Algae which deposit a limestone skeleton, and hence sometimes called the coralline algae. These belong primarily to the green and the red algae. The greens, such as Halimeda, are generally made up of small platelike segments which disintegrate into sub-component crystals when the algae dies, providing the major source of coral reef sand. The reds include both branching and encrusting forms. The branching red species are also important contributors to sand, while the encrusting red species rapidly cover dead coral rubble and bind loose pieces to the reef framework, while providing the substrate on which new young coral larve settle and grow, setting the stage for reef recovery from hurricanes. Encrusting calcareous algae are characteristic of low nutrient waters. Branching calcareous algae are rare wherever parrotfish grazing is high. They are stimulated greatly in abundance by excessive shore-based nutrient sources, until they are in turn overgrown by large fleshy algae which do not contribute to building the reef or to its sand.

Ciguatera: Fish poisoning caused by a toxin produced by certain single-celled algae which live on the surface of larger algae which are eaten by herbivorous fish. These toxins are often concentrated up the food chain, but their presence is extremely variable, affecting only some species at certain locations at particular times of the year. In each area there is a variable catalogue of fish species avoided by local fishermen because of known cases of ciguatera, often including certain surgeonfish, parrotfish, and barracuda. The toxin affects people eating poisoned fish by directly affecting the flow of electrically charged ions across the nerve cell membranes, causing osmotic shock. Victims feel that their feet are on fire, and many die. The effects of ciguatera can be completely reversed if the simple sugar-like compound mannitol is administered orally or intravenously shortly after onset of symptoms. Some specialists feel that ciguatera-like poisoning is on the increase and may be related to pollution, but many episodes remain extremely unpredictable, perhaps due to poor data bases.

CLOD: Coralline Lethal Orange Disease (CLOD) is a recently described disease which attacks encrusting and branching red calcareous algae, killing it (Littler & Littler, 1995). It progresses in orange and yellow bands across the red, pink, or purple algae, leaving behind dead white limestone skeleton which is then covered by fuzzy microscopic filamentous algae. The disease-causing pathogen has not yet been isolated, but the disease is transmissible by contact. CLOD has mainly been found in the Cook Islands and Fiji, but may be in the process of spreading as it was unknown untill very recently. Previously unknown diseases are thought by many long term observers to be becoming more common in reef and other marine habitats, including red tides, coral diseases, fish kills, sea urchin diseases, and cancer-like tumors in corals, soft corals, and turtles (Williams and Bunkley-Williams, 1990).

Cnidaria: The corals and their relatives, including the stony corals, soft corals, fan corals, sea anemones, jelly-fish, etc, based on the unique presence in these animals of unique types of stinging cells known as cnidoblasts in the outer layer of the tentacles. These cells literally explode to inject a toxic or paralytic barb into minute zooplankton prey, but are rarely powerful enough to penetrate human skin, with some notable exceptions including "fire" corals and many jellyfish. Cnidaria are better known as Coelenterates.

Coelenterates: Also known as Cnidaria. Simple aquatic animals including the corals and their relatives, with a primitive anatomy featuring a ring of tentacles surrounding a mouth that opens into a bag-like stomach. The mouth also serves for excretion as the animal has no anus (coelenteron is derived from the Greek word for sack). Coeienterates may be single or live in huge colonies, and may take the form of either a flower-like animal attached to the bottom ( or polyp) such as corals, or a free floating swimming bell-shaped animal like a jellyfish (or medusa), or alternating between these forms at different stages of its life cycle. Free swimming coelenterates are jelly-like, often with powerful stinging ceils, but sometimes relying on symbiotic algae. Bottom dwelling coelenterates also have stinging cells for catching animal prey, and may or may not have symbiotic algae. They come in an astonishing variety of shapes, sizes, and colors, contributing to the overwhelming bulk of the bottom living biomass in a healthy reef. Some varieties are jelly-like, some fleshy, some horny, some have microscopic skeletal elements in their tissue, while others secrete a massive limestone skeleton beneath their tissue, into which they can draw for protection. Bottom dwelling Coelenterates are generally known as the hard and soft corals.

Coral: Generally used to refer to coelenterates with the capability of depositing hard massive limestone skeletons which build the wave-resisting reef structure, but often used to include their "soft" non-reef building relatives as well as the "hard" corals (below), and sometimes more loosely to include all other attached limestone skeleton-building animals and algae. All major reef-building corals are symbiotic, and therefore capable of bleaching, along with symbiotic soft corals and other organisms. When this happens the entire reef may look as if covered by snow.

Corymbose: Corals with an unusual form having vertical pillars rising from a horizontal plate-like base, which is sometimes on a stalk. Often found in shallow habitats, and generally members of the Acropora family, which include a tremendous range of branching as well as shelf-like morphologies, many with uncertain taxonomic identities

Cryptic: Organisms which hide in caves, crevices, and under corals. These include free-living organisms, many of which only emerge at night, as well as attached creatures. Some are adapted to dark conditions, but others are confined to cryptic habitats because of predation. An example is the calcareous algae Halimeda, which grows in prolific clumps in habitats where nutrients and grazing permit, but which are cryptic in normal reefs since parrotfish eat all which protrudes. In these habitats the heavily grazed algae are confined to crevices and cracks between and beneath corals where they cannot be reached, and chewed flat by fish teeth wherever they protrude.

Doldrums: Long extended calm periods of low winds, low clouds, high sunshine, and high temperatures. They allow the sea surface temperature to rise to excessive levels since more sunlight penetrates the flat ocean surface and heats the water below. Formerly the curse of sailing ships, they are now recognized as "bleaching weather", pre-cursor events to mass coral bleaching when they occur at the warmest time of year. Local observers have almost invariably noted that bleaching followed unusually long doldrum periods. Doldrum weather episodes have always existed and may have caused sporadic small-scale natural bleaching events in the past, but these appear to have been too rare before the 1980's to have been noted by divers, fishermen, and boaters. It is possible that doldrum conditions are becoming more frequent in the tropics as the result of global warming. Fishermen in many parts of the Caribbean and the Pacific have noted a decline in the catch of fish species which are brought in by frontal storms, which they attribute to a decrease in wind strength and increase in calm weather conditions since the late 1970's. Interestingly, the old wind patterns returned for a few years after the 1991 eruption of Mount Pinatubo cooled the earth, but are now dissipating.

El Nino: The popular name for an irregular multi-year cycle called the Southern Oscillation or ENSO, which affects atmospheric pressure, winds, sea levels, ocean currents, deep water upwelling, rainfall, and temperatures in certain parts of the earth. It's Spanish name, meaning the Child, refers to observations of large Christmas-time changes in fish and sea bird distributions off Peru caused by changes in deep ocean upwelling. The major oscillation is reflected in the atmospheric pressure differences between Darwin, North Australia, and Tahiti, French Polynesia, with one being high while the other is low. This affects the ocean currents and depth of surface water mixing across the Pacific, heating some ocean areas and cooling others. Bleaching occurred in areas which were strongly warmed by the 1983 El Nino, the strongest ever measured, but it also took place in areas which are normally cooled by the El Nino, in areas which are unaffected by El Nino, and in non El Nino years. In recent years the Pacific circulation was apparently stuck in an unusual permanent mild El Nino, but this has since moderated (Climate Diagnostics Bulletin, 1995).

Eutrophication: The massive and excessive growth of fast-growing nuisance algae in response to elevated levels of nutrients. All marine ecosystems can become eutrophic when excessive amounts of nutrients enter the water, usually from sewage, fertilizer, and animal wastes, but sometimes from natural sources (Lapointe et al., 1992). No ecosystem is more sensitive to eutrophication than coral reefs, which are smothered and killed by fleshy and filamentous algae at nutrient concentrations far below those which would cause problems in any other marine, river, lake, or bog ecosystems. Eutrophication is now ubiquitous in coral reefs adjacent to the populated shorelines everywhere in the tropics, but its effects have often been incorrectly blamed on overfishing or lack of herbivorous sea urchins by researchers who have not examined the spatial and temporal relationships of weedy algae abundance to nutrient concentrations. Eutrophication results in massive growth of fleshy algae, often following a successional sequence of species, which overgrow and kill corals and other attached marine animals. Eutrophication converts living reefs with immense biodiversity to dead reefs with only a handful of weedy algae and fish types incapable of building the waveresisting reef structure which provides the home and food for all other reef organisms and which protects the shoreline from erosion. In the early stages of eutrophication calcareous algae proliferate, but they are then overgrown and smothered by fleshy algae, ending the supply of new sand to replace that removed by wave erosion.

Fleshy algae: Algae which do not make limestone skeletons, and which are not composed of microscopic filaments, generally called seaweeds. These dominate most eutrophic habitats, but are normally very rare in healthy coral reefs except at sites protected from grazing such as in crevices or on the shallow top of the reef flat. Fleshy algae include hundreds of brown, red, and green algae species, each with distinctive ecological preferences which may be hard to recognize, especially when they are rare.

Fluorescence: The emission of light by tissue pigments in marine organisms. The origin and significance of fluorescence is uncertain, but there are many hypotheses. Many south-eastern Pacific corals, especially of the Acropora family, show intense fluorescence after bleaching, but this is due to fluorescent pigments which are normally present in healthy corals, often in a ring around the mouth, which become more obvious after the masking colors of the symbiotic algae are lost by bleaching (Jardin et al., in preparation).

Genera: The plural of genus, a term indicating the family to which a species belongs. Every species has two names: the first is its genus or immediate family, the second is its individual species name in that family. Genera are groups of species which are distinctively more closely related to each other than to any outside species. In general the genus name is unique, but in a few rare and confusing cases the same name has been used in unrelated groups: the coral Turbinaria and the brown algae Turbinaria may occur in the same habitat, but are not relatives. Most coral and algae genera have only one species in a given area, or are monotypic, and only a few genera have large numbers of very similar and closely related species which can be hard to tell apart. While observers can usually be easily trained to tell genera apart, identification of the actual species may be impossible in the field if they require microscopic anatomical measurements. In these cases it is necessary to take samples for study by taxonomists specializing in the genus.

Hard corals (stony corals): Coelenterates with massive limestone skeletons. Almost all hard corals are species of the "true" corals, the Scleractinia (with six-fold symmetry in their polyps), along with a few species of Octocorallia which have hard limestone skeletons (relatives of the true corals with eight-fold polyp symmetry, most of whose members are soft corals) and Millepora, or fire coral, a unique massive skeleton-producing member of the Hydrozoa, fiercely stinging branching coelenterates. While some hard corals lack symbiotic algae, these are small and slow growing and of negligible importance in coral reefs, which are largely built by the symbiotic coelenterates. Non-symbiotic hard corals can occur in deep and cold water, but symbiotic hard corals only occur where the water is sufficiently warm and shallow since symbiotic algae will not remain in association with the coral host outside certain species-dependent tolerance ranges. Photosynthesis by symbiotic algae greatly speeds up the rate at which corals deposit limestone skeleton, and reef building corals which bleach essentially halt growth of their skeleton.

"Hot Spot" anomaly: A region of the ocean surface which is heated up by more than 1.0 degrees C above the long term average value. Hot spots which occurred in the warmest season of the month preceded all major mass bleaching episodes (Goreau and Hayes, 19941.

Necrosis: Tissue death. As an adjective, necrotic. Dead or dying coral tissue, as the result of digestion by predators, attack by pathogenic organisms, or starvation. Necrotic tissue can look white and be confused with bleaching unless close visual inspection is made to determine whether the tissue is intact. Necrotic tissue is white, with stringy wisps of tissue and mucus. Distinct communities of bacteria have been found to occur on necrotic, bleached, and normal coral tissue in both the Pacific and Atlantic by Jindal et al. (1994, 1995).

Nutrients: Essential elements required for the growth of all organisms, in particular the plants and algae which form organic matter out of water and carbon dioxide, using the energy of sunlight. Limiting nutrients are those whose insufficiency in the environment reduces growth rates of organisms below their potential, but in excessive levels they cause eutrophication. They are most generally dissolved inorganic nitrogen (DIN, as ammonium or nitrate) and soluble reactive phosphorus (SRP, as orthophosphate and dissolved organic phosphorous). When nutrients are low, an increase in them stimulates rapid growth of algae, which will overgrow and kill corals when the nutrients are above a critical eutrophication level of 1.0 micromole per liter of DIN and 0.1 micromole per liter of SRP (Bell, 1992; Lapointe and Clark, 1992). Excessive nutrient levels almost always result from discharge of sewage directly into the coastal zone or indirectly via rivers, groundwater seepage, and surface runoff, or from agricultural organic wastes and fertilizers. Reduction of coastal zone nutrients below the critical levels is neccessary if reefs are to recover from eutrophication. This can be accomplished through a combination of upgrading sewage collection and treatment to the tertiary level, which removes nutrients from sewage effluents, or use of dry and composting toilets which allow the nutrients to be recycled into organic fertilizer taken up by plants rather than contaminating the groundwater supply and ultimately the coastal zone (Goreau and Thacker, 1994).

pH: The acidity of water, expressed as the negative logarithm of the hydrogen ion concentration. Typically measured with an electrode meter, pH values below 7 are acidic, values above are basic. Seawater values are very well buffered' held to a nearly constant basic pH by the solution and precipitation of limestone from seawater, largely by biological means.

Phytoplankton: Microscopic single celled floating algae living in the water. Corals do not feed on them, focusing instead on the zooplankton, the small floating animals which eat the phytoplankton, such as small shrimps and worms. Water low in nutrients is clear and blue unless it contains suspended sediments, while water high in nutrients is green and turbid due to massive blooms of microscopic single celled algae, or phytoplankton, reducing light levels and negatively affecting coral growth. Increased numbers of jellyfish, zooplankton, bacteria, and their feeders are also typically noted in eutrophic coral reefs.

Polyps: The basic unit of organization of the attached coelenterates is the polyp, a single mouth and stomach ringed by tentacles. These may be free living individual polyps or colonial, formed by the sideways growth of polyps which divide but remain attached, forming large sheets, branches, or mounds. Coral colonies, growing from minute sizes up to 15 meters or more across, may have vast numbers of polyps, each ranging in size from a millimeter to more than 10 centimeters. These are capable of sharing food and behaving as a unit through interconnections between their stomachs and nervous systems. Each species has a uniquely shaped polyp which develops its tentacles and skeleton in characteristic patterns. Although there can be tremendous variation in color, polyps of a single species are very similar even though the shape of the colony as a whole is generally highly variable, depending on local physical and biological stresses. This is analogous to variations in the form of trees of a single species in different locations, even though the leaf is constant in shape. Coral colonies create the entire physical structure of the coral reef ecosystem, just as trees create the forest habitat, allowing insects, bird, mammals, etc. to survive. Loss of the corals or trees imperils all the rest of biodiversity directly or indirectly dependent on the habitat they create.

Scleractinians: The "true" hard corals, members of the order Scleractinia. Their polyps have a six-fold symmetry, and they are grouped with various anatomically similar "soft" coral relatives (such as sea anemones and zoanthids) in the subclass Zoantharia, which is combined with the "soft" corals having polyps with eight-fold symmetry to make up the class Anthozoa, the coelenterates which develop a free living larva, or planula, that settles and converts into a polyp, without going through a medusa stage. All Scleractinians are hard corals, and most of them are symbiotic, including all the reef-building sceleractinians.

Soft corals: Bottom dwelling coelenterates which do not form a massive limestone skeleton, but which may contain microscopic limestone spicules in the tissue. Soft corals include a large range of organisms with variable appearance and color, including encrusting, branching, whip-like, and fan-shaped forms. Food sources can vary from symbiotic algae, fish, zooplankton, and detritus. They include the Alcyoncacea and Gorgonacea in the Octocorallia, sea anemones, zoanthids, and several other groups. While often very abundant, they do not contribute to building the reef framework. Soft corals tend to become more abundant in stressed or polluted reefs.

 Sponges: Marine animals even simpler than corals, which pump water through their bodies in order to filter out suspended detritus for the bacteria on which they feed, thus acting to keep the water clear (Reiswig, 1971). Sponges are extremely variable in form and color, lack a nervous system, and have unique powers of regeneration even after the whole animal is disaggregated into individual cells. Some are soft, but most contain microscopic spicules made of silica. A few sponge species also contain symbiotic algae, and may bleach. Sponges are generally more abundant on eutrophic reefs with large amounts of suspended organic matter and bacteria. Sponges were uncommon at most Pacific sites in this study, in contrast to their abundance in the Caribbean. This could be related to the large number of sponge-eating fishes found in the Pacific

Statistical significance: A statistical correlation between variables is regarded as significant if the probability of it emerging by chance is less than one part in 20, or the calculated P (for probability) value is less than 0.05. This is equivalent to a 95% probability that it does not result from chance. A statistical correlation is regarded as highly significant when P is less than 0.01 or a 99% likelihood of not being random, and is regarded as very highly significant when P is less than 0.001, or 99.9% likely not to emerge by chance. These three statistical categories are represented here by the symbols *, **, and *** respectively. A correlation with a P value which is greater than 0.05 could result by chance more than once in 20, and is regarded as not statistically significant.

Symbiosis: Two or more unrelated organisms, tissues or cells, living together in a mutually beneficial arrangement, which may even be obligatory. Reef building corals are a symbiosis between a marine organism, usually a sceractinian coelenterate, and single celled algae, often called zooxanthellae, living within its tissues. The algae makes organic material by photosynthesis, using carbon dioxide and nutrients excreted by the host organism, and in turn providing it with oxygen. The symbiotic algae releases most of the organic matter it makes to the host organism, providing it with much of its food and the color of its tissue. In skeleton-forming organisms such as the corals, photosynthesis of symbiotic algae greatly accelerates the rate at which the coral host forms its skeleton. Any environmental stress which is sufficiently severe causes the organism to lose its symbiotic algae, causing it to become bleached. Symbiosis is the essential feature which allows corals to grow so fast, building coral reefs which provide the homes for all other reef species. Without it reefs would be vastly smaller and less diverse, as they were before the modern symbiotic corals evolved. A large number of other reef animals besides the scleractinians may also have symbiotic algae and are capable of bleaching, including the giant clams.

Ultraviolet radiation: Short wavelength solar radiation, more energetic than blue light. High levels of ultraviolet light are capable of damaging the DNA genetic material of all organisms. Ultraviolet light is able to penetrate significant depths of clear ocean water in the tropics, and high ultraviolet radiation is capable of causing bleaching in experiments (Gleason and Wellington, 1993). Corals contain high levels of ultraviolet-absorbing pigments, which are contained in the host animal and remain after the coral is bleached. Since corals are so efficient at absorbing ultraviolet light it is unlikely that small increases in ultraviolet light penetration is an important cause of bleaching in the field. Coral bleaching has been seen as deep as 300 feet down, well below the depth to which ultraviolet light can penetrate, but still in the warm upper surface layer overlying the thermocline boundary to colder deep water below. Since the high total sunlight conditions that raise the temperature of the ocean to excessive levels for corals always include high ultraviolet light exposure, it is difficult to separate the effects of temperature and ultraviolet radiation on corals from field data. Ultraviolet light is found experimentally to increase the rate of bleaching, but only when temperatures are above a threshold level, implying that it is a contributing rather than a primarily causal stress (Glynn and D'Croz, 1992). There is little long term data available which allows reliable estimates of changes in ultraviolet radiation anywhere in the tropics, but the tropics are thought to have suffered the smallest decline in the overlying ozone layer. Global models of ozone layer depletion show larger estimates of ozone loss near the edges of the tropics than over the equator. Decreasing global ozone could exacerbate the effects of bleaching at higher latitudes in the tropics in coming years if UV exposure is a significant contributor to bleaching, and cause corals in equatorial zones to have relatively higher apparent bleaching threshold temperatures. If ozone depletion-linked UV increases were a significant cause of mass bleaching most episodes should have occurred at high latitude reefs such as Hawaii, Bermuda, Okinawa, Mauritius, etc, and little or no bleaching should have taken place at more equatorial sites. These cooler areas do not appear to have bleached excessively compared to warmer sites, but unfortunately there is little data to test these hypotheses from the reef sites at which bleaching information is lacking near the equator.

Verrucae: Spiny limestone projections between polyps of corals in the Pocillopora family. When the coral is bleached and dies, the entire surface becomes covered with filamentous green algae after a few days to weeks. within a few months the tips of verrucae are colonized by encrusting calcareous red algae. If conditions are not eutrophic, the calcareous algae completely cover the entire dead coral skeleton within a few years, otherwise fleshy algae or algal turf will dominate.

 Zoanthid: "Soft" corals in the same subclass as the "true" corals, but which form encrusting mats, often on recently disturbed hard bottom. Many zoanthids have symbiotic algae, and those in the group Palythoa are one of the earliest reef organisms to become visibly bleached and one of the last to recover.

Zooplankton: Minute animals which feed on microscopic algae, or phytoplankton, including a wide range of shrimp relatives, the larvae of free swimming and bottom dwelling reef animals, and worms. Some zooplankton are passively transported by currents, while most are demersal, living in the reef by day and swimming in huge swarms at night. Corals rely on zooplankton as their major source of food to supplement that received from their symbiotic algae, and are efficiently equipped with tentacles and stinging cells to capture them. Most corals expand their tentacles at night to catch the large numbers of zooplankton which emerge from the reef into the water column only in the dark.

Zooxanthellae: A general term applied to symbiotic algae within coral tissues, related to the free-living yellow-brown algae, the Dinoflagellates. This term is now regarded as inaccurate since a very wide range of symbiotic species from both dinoflagellates as well as other algae groups have been discovered (Trench, 1993; Banaszak et al., 1993; Trench and Long, 1995), and shown that each has different environmental tolerances (Iglesias-Prieto et al., 1992) and host specificities. A large part of the variability in bleaching responses seen within many coral species may be related to genetic variation in the different strains and species of symbiotic algae (McNally et al., 1994), which is still inadequately known.

LITERATURE CITED

Banaszak, A., R. Iglesias-Prieto, and R. Trench (1993) Scrippsiella velellae sp. nov. (Peridiniales) and Gloeodinium viscum sp. nov. (Phytodiniales), dinoflagellate symbionts of two hydrozoans (Cnidaria). J. Phycol. 29: 517-528.

Bell, P. (1992) Eutrophication and coral reefs - some examples in the Great Barrier Reef Lagoon. Water Resources 26: 553-568

Carleton, J. H., and T. Done (1995) Quantitative video sampling of coral reef benthos: large scale application. Coral Reefs 14: 35-46.

Climate Diagnostics Bulletin (1994) National Weather Service, NOAA, Washington, DC.

Craig, P. (1994) State of our local environment. Samoa News, April 22, 1994.

Craig, P., P. Trail, G. Grant, J. Craig and D. Itano (1993) American Samoa: natural history and conservation topics. Biol. Rep. Ser. 42, Department of Marine and Wildlife Resources, Pago Pago, American Samoa

Crosby, M. P., S. Drake, C. Eakin, N. Fanning, A. Paterson, P. Taylor, and J. Wilson (1995) The United States Coral Reef Initiative: an overview of the first steps. Coral Reefs 14: 1 -3.

Dahl, A. (1984) Coral reef monitoring handbook. Regional Seas Programme Reference Methods for Marine Pollution Studies 25: 1-25. United Nations Environment Programme, Nairobi, Kenya.

Darwin, C. (1842) On the structure and distribution of coral reefs, London.

DeVantier, L. M., E. Turak, J. Davidson, & T. J. Done (1995) Recurrent seasonal bleaching on nearshore reefs of the Central Great Barrier Reef. p. 31 in Biology and Geology of Coral Reefs, Proc. Int. Soc. Reef Studies European Mtg. Newcastle upon Tyne, U. K. (Abstract).

Drollet, J. H., M. Faucon, S. Maritorena, and P. Martin (1994) A survey of environmental physico-chemical parameters during a minor coral mass bleaching event in Tahiti in 1993. Aust. J. Mar. Freshwater Res. 45: 1149-1156.

Enright, J. (1993) Fa'atautaiga, Traditonal Samoan Fishing Methods, American Samoa Council on Culture, Arts and Humanities.

Flora, J. and R. Flora (1994) Observations on Manihiki Atoll. (unpublished report), pp 1-13.

Gleason, D., and G. Wellington (1993) Ultraviolet radiation and coral bleaching. Nature 365: 836-838

Gleason, M. G. (1993) Effects of disturbance on coral communities: bleaching in Moorea, French Polynesia. Coral Reefs 12 193-201.

Gleeson, M.W. and A.E. Strong (1994) Applying MCSST to coral reef bleaching. COSPAR-94, A1-085. Hamburg, Germany, July, 1994.

Gleeson, M.W. (1994) A Trident scholar project report #215. Correlation of coral reef bleaching events and remotely sensed SSTs. U.S. Naval Academy, Annapolis, MD.

Glynn, P. (1990) Coral mortality and disturbances to coral reefs in the tropical eastern Pacific. In "Global ecological consequences of the 1982-83 El Nino Southern Oscillation'., ed. P. Glynn, Elsevier Press.

Glynn, P. (1991) Coral reef bleaching in the 1980s and possible connections with global warming. Trends in Ecology and Evolution 6: 175-179.

Glynn. P. t1993) Coral reef bleaching: ecological perspectives. Coral Reefs 12: 1-17.

Glynn, P.W. and M.W. Colgan (1992) Sporadic disturbances in fluctuating coral reef environments: El Nino and coral reef development in the eastern Pacific. Amer. Zool. 32: 707-718.

Glynn, P.W. and L. D'Croz (1990) Experimental evidence for high temperature stress as the cause of El Nino-coincident coral mortality. Coral Reefs 8: 181-191.

Goreau, T. F. (1964a) Mass expulsion of zooxanthellae from Jamaican reef communities after Hurricane Flora. Science 145: 383-386.

Goreau, T. F. (1964b) On the predation of coral by the spiny starfish Acanthaster planci in the Southern Red Sea. Bull. Sea Fisheries Research Station Haifa, Israel 35: 23-26.

Goreau, T. J. (1992) Bleaching and reef community change in Jamaica: 19511991. Amer. Zool.32: 683-695

Goreau, T. J., R.L. Hayes, J.W. Clark, D.J. Basta and C.N. Robertson (1993) Elevated sea surface temperatures correlate with Caribbean coral reef bleaching, In "A global warming forum: scientific, economic and legal overview", pp. 225-255, ed. R. Geyer, CRC Press, Boca Raton, FL

Goreau, T. J. and R. Hayes (1994) Coral bleaching and ocean "hotspots". AMBIO 23: 176-180.

Goreau, T. J. and K. Thacker (1994) Coral reefs, sewage, and water quality standards. (In Press) Proc. Carib. Water and Wastewater Assoc. Conf., Kingston, Jamaica.

Goreau T. J. and R. Hayes (1995) Monitoring and calibrating sea surface temperature anomalies using satellite and in-situ data to study the effects of weather extremes and climate change on coral reefs. (In Press) Proc. Conf. Remote Sensing and Environmental Monitoring for the Sustainable Development of the Americas, San Juan, Puerto Rico.

Goreau T. J., R. Hayes, and A. C. Strong Tracking South Pacific coral reef bleaching by satellite and field observations. (Submitted) Proc. 8th. International Coral Reef Symposium, Panama. (Abstract).

Hoegh-Guldberg, O. (1994) Mass-bleaching of coral reefs in French Polynesia. Greenpeace Internat. pp.1-36.

Hoegh-Guldberg, O., and B. Salvat (1995) Periodic mass bleaching and elevated sea temperatures: Bleaching of outer reef slope communities in Moorea, French Polynesia. Mar. Ecol. Proo. Ser.121: 181-190.

Iglesias-Prieto, R., J. Matta, W. Robins, and.R. Trench (1992) Photosynthetic response to elevated temperature in the symbiotic dinoflagellate Symbiodinium microadrieficum in culture. Proc. Nat. Acad. Sci. USA 89: 10302-10305.

Jaap, W. (1979) Observations on zooxanthellae expulsion at Middle Sambo Reef, Florida Keys. Bull. Mar. Sci. 29: 414-422.

Jardin, C., B. Salvat, R. Hayes, T. Goreau and C. Mazel (1994) Fluorescence, color and mortality of bleached corals in French Polynesia. (manuscript in preparation) .

Jindal, A., R.L. Hayes, T.J. Goreau and G.W. Smith (1994) Bacterial communities from bleached, unbleached and necrotic south Pacific scleractinian corals. Proc. SE Br. Amer. Soc. Microbiol. (Abstract).

Jindal, A., K. Ritchie, R. Hayes, T. J. Goreau, and G. Smith (1995) Bacterial ecology of selected corals following the 1994 south central Pacific bleaching event. (In press) Proc. Ass. Mar. Labs. Carib., St. Thomas. USVI

Jokiel, P. & S. Coles (1990) Response of Hawaiian and other Indo-Pacific corals to elevated temperatures. Coral Reefs 8:155-162

Lapointe, B., and M. Clark (1992) Nutrient inputs from the watershed and coastal eutrophication in the Florida Keys. Estuaries 15: 465-476

Lapointe, B., M. Littler, and D. Littler (1992) Modification of benthic community structure by natural eutrophication: the Belize Barrier Reef. Proc. 7th International Coral Reef Symposium, 317-328, University of Guam.

Littler, M. and D. Littler (19953 Impact of CLOD pathogen on Pacific coral reefs. Science 267: 1356-1360.

Loubersac, L. (1992) Orama Nui: la Polynesie vue de ltespace. STP Multipress, Papeete, Tahiti.

Magruder, W.H. and J.W. Hunt (1979) Seaweeds of Hawaii: A Photographic Identification Guide. Oriental Publ. Co., Honolulu. HA.

 Mayor, A. (1918) Toxic effects due to high temperature. Carnegie Institute Wash. Pap. Mar. Biol. 12: 175-178.

McNally, K., N. Govind, P. Thome, and R. Trench (1994) Small subunit ribosomal DNA sequence analyses and a reconstruction of the inferred phylogeny among symbiotic dinoflagellates (Pyrrophyta). J. Phycol. 30: 316-329.

Migotto, A. (1995) Anthozoan bleaching on the southern coast of Brazil, summer 1994. Proc. Intern. Coelenterate Symp. Groningen, the Netherlands (in press).

National Research Council (1995) Understanding Marine Biodiversity: a research agenda for the nation. pp. 1 -114. National Academy Press, Washington DC.

Oceanographic Monthly Summary (1994) National Ocean Service, NOM, National Environmental Satellite Data and Information Service (NESDIS), Camp Springs, MD.

Reiswig, H. (1971) Particle feeding in natural populations of three marine Demospongiae. Biol. Bull. Mar. Biol. Lab. Woods Hole 141: 568-591.

Reynolds, R.W. and D.C. Marisco (1993) An improved real-time global sea surface temperature analysis. J. Climate 6: 1 14-1 19.

Reynolds, R.W. and T.M. Smith (1994) Improved global sea surface temperature analyses using optimum interpolation . J. Climate 7: 929-948.

Rogers, C. (1990) Responses of coral reefs and reef organisms to sedimentation. Mar. Ecol. Pron. Ser. 62: 185-202.

Rogers, C. (1993) Hurricanes and coral reefs: the intermediate disturbance hypothesis revisited. Coral Reefs 12: 127-137.

Rogers, C (1994) Coral Reef Monitoring Manual for the Caribbean and the Western Atlantic, U.S. National Park Service, Virgin Islands National Park.

Salvat, B. (1991) Blanchissement et mortalite des scleractiniaries sur les recifs de moorea (archipel de la Societe) en 1991. C. R. Acad. Sci. Paris 314: 105-111.

Salvat, B. (1992) Coral reefs, a challenging ecosystem for human societies. Global Environ. Changes 2: 12-18

Stanley, D. (1993) South Pacific Handbook, Fifth Edition, Moon Publ., Chico, CA (source for South Pacific maps in text).

Trench, R. (1993) Microalgal-invertebrate symbioses: a review. Endocytobiosis & Cell Res. 9: 135-175

Trench, R., and L. Van Thinh (1995) Gymnodinfum linuchae sp. nov.: the dinoflagellate symbiont of the jellyfish Linuche unguiculata. Eur. J. Phycol. 30: 149-154.

U.S. Department of State (1994) The U.S. Coral Reef Initiative: Forging partnerships for effective management.

Veron, J.E.N. (1986) Corals of Australia and the Indo-Pacific. Univ. Hawaii Press, Honolulu.

Whistler, W.A. (1992) Flowers of the Pacific Island Seashore. Isle Botanica, Honolulu. HA

Williams, E.H. and L. Bunkley-Williams (1990) The world-wide coral bleaching cycle and related sources of coral mortality. Atoll Res. Bull. 335:1-71.

Williams, E., C. Goenaga, and V. Vicente (1987) Mass bleaching on Atlantic coral reefs. Science 238: 877-878.

Yonge, C. M., and A. G. Nicholls (1931) Studies on the physiology of corals. Great Barrier Reef Exp. Sci. Pap. 1: 135-211.

ACKNOWLEDGMENTS

We thank the following individuals who have assisted us before, during and after this survey. They contributed immeasurably to our mission by providing us with information, supplies, or advice. They also guided us in the selection of interviews, sites to visit, and documents to review. Without the support, help, background observations, and knowledge of these many individuals, this survey would have been impossible to complete. We are grateful to them all, even if their contributions were too many to be acknowledged individually in the text. While most of our assessments will come as no surprise to them, being based on their own observations in part, we hope that the rapid and widespread assessment in this study will assist them in evaluating bleaching and comparative stresses in their own and other reefs. Despite our deep debt to them, the authors are fully responsible for the overall conclusions drawn in this study. Financial sponsorship from the Coral Reef Initiative of the U.S. Department of State is gratefully acknowledged for making this study a reality. Special thanks go to Pete Ely for providing permission to use some of his photographs from Aitutaki.

Tahiti, Society Islands:

Dr. Francis Rougerie, ORSTOM, Tahiti, French Polynesia

Catherine Jardin, Laboratoire d'Ecologie Marine, Universite Francaise du Pacifique, Tahiti, Society Islands, French Polynesia

Henri Poliquen, Owner/lnstructor, Tahiti Plongee, Papeete, Tahiti, Society Islands, French Polynesia

Christian Durocher, Journaliste, La Depeche de Tahiti, Papeete, Tahiti, French Polynesia

Moorea, Society Islands:

Dr. Bernard Salvat, Director, E.P.H.E., and Professor, Universite de

Perpignan, Perpignan, France

Yannick Chancerelle, E.P. H. E. , Moorea, Society Islands, French

Polynesia

Rangiroa, Tuomotu Islands:

Phillippe Cabral, Biologist, Departement Aquaculture, Etablissement

pour la Valorisation des Activities Aquacoles et Maritimes

(EVMM), Rangiroa, Tuamotu, French Polynesia

Terii Seaman, Biologist, Pearl Oyster Culture, Department, EVAAM,

Rangiroa, Tuamotu, French Polynesia

Francois Tavita ("Totot'), EVAAM, Rangiroa, Tuamotu, French Polynesia

Laurence Mani, EVMM, & Faculty of Agriculture, University of Bordeaux,

France

Yves Lefevre, Owner/ Instructor, Centre de Plongee Subaquatique,

Raie Manta Club, Rangiroa, French Polynesia

Lance Leonhardt, Bethlehem PA

Rarotonga, Cook Islands:

Wayne Tamangaro King, Chief Conservation Officer, Environmental

Planning and Resource Management, Conservation Service,

Tu'anga Taporoporo, Rarotonga, Cook Islands

Jenny Shaw, Senior Conservation Officer, Cook Island Conservation

Service, Rarotonga, Cook Isiands

Teina Rongo, Volunteer, Cook Island Conservation Service, Rarotonga,

Cook Islands

Aitua Kuro Isaia, Cook Island Conservation Service, Penrhyn,

Cook Islands

Greg Wilson, Owner/lnstructor, Cook Island Divers, Ltd., Titikaveka,

Rarotonga, Cook Islands

Francis Baronie, volunteer, Cook Island Conservation Service,

Rarotonga, Cook Islands

David J. Simms, Instructor, Cook Island Divers and Underwater

Photographer, Aquapic, Avarua, Rarotonga, Cook Islands

Aitutaki, Cook Islands:

Dr. Gustav Paulay, Professor of Marine Science, University of Guam, Mangilao,GU, USA

Dr. Charles J. Flora, President Emeritus, and Professor Emeritus of Biology,Western Washington University, Bellingham, WA ,USA

Peter Ely, Student, Western Washington University, Blaine,WA, USA

Bobby Bishop, Conservation Officer, Aitutaki

Daniel loane Amuri, Aitutaki SCUBA, Ltd., Aitutaki

Neil Mitchell, Owner/lnstructor, Aitutaki SCUBA, Ltd., Aitutaki

Michael Henry, Director, Aitutaki Tourist Board, Aitutaki

lan Bertram, Department of Fisheries. Aitutaki

Tutuila, American Samoa:

Dr. Peter Craig, Chief Biologist, Department of Marine and Wildlife

Resources, Pago Pago

Lelei M. Pelau, Manager, American Samoa Coastal Management

Program, Office of Economic Development Planning, American

Samoa Government, Pago Pago

Dr. Alison Green, Biologist, Department of Marine and Wildlife

Resources, Pago Pago

Susan Saucerman, Fisheries Biologist, DMWR, Pago Pago

Fa'asega Kuresa, Giant Clam Mariculturist, DMWR, Pago Pago

Fale Tuilagi, Captain, DMWR, Pago Pago

John Enright, President, Le Vacmatua, Pago Pago

Karla Kluge, Wetland Specialist, Coastal Management Program,

Pago Pago

Larry Ward, Environmental Planner, Coastal Management Program,

Pago Pago

Chuck Brugman, Owner, Dive Samoa, Inc., Pago Pago

Lewis Wolman, Editor, Samoa News, Pago Pago

Niue:

Terry Coe, Minister of Agriculture, Forestry and Fisheries, Public Works,

Post and Telecommunications, Alofi

Kirabati:

Tererei Abete, Ministry of Environment and Natural Resources, Bairiki, Tarawa

Brazil:

Dr. Alvaro Migotto, Universidade de Sao Paulo, Sao Sebastiao, S.P.

United States of America:

Elinor Constable, Assistant Secretary of State, Bureau of Oceans and

International Environment and Scientific Affairs, U.S. Department of

State, Washington, DC

Rafe Pomerance, Deputy Assistant Secretary of State, Bureau of Oceans

and International Environment and Scientific Affairs, U.S.

Department of State, Washington, DC

Peter Thomas, Coordinator, International Coral Reef Initiative, US

Department of State, Washington, DC

Susan Drake, Coordinator (former), International Coral Reef Initiative, US

Department of State, Washington, DC

Grenville Sewell, Coordinator (former), Coral Reef Initiative, US

Department of State, Washington, DC

Angela Barber-Wilson, Secretary (former), Coral Reef Initiative, US

Department of State, Washington, DC

Sebia Hawkins, Pacific Campaign Coordinator, Greenpeace

International, Washington, DC

Stephanie Mills, Pacific Campaign, Greenpeace International,

New Zealand

Jeanne Kirby, Greenpeace International, Portland, OR

J.R. Scanlon (All Mau), Senior Congressional Staff Aide to Rep.

Faleomaevenga, Territory of American Samoa, Washington, DC

Scientific Reviewers:

Dr. Ernest H. Williams, Professor of Marine Sciences, University of Puerto Rico, Mayaguez, Puerto Rico

Dr. Peter Glynn, Professor of Marine Science, University of Miami, Miami, Florida

Dr. Alan Strong, Professor of Oceanography, US Naval Academy, Annapolis, Maryland

Dr. Terry Done, Australian Institute of Marine Sciences, Cairns, Queensland, Australia

Dr. Robert Trench, Professor of Zoology, University of California, Santa Barbara, California

We wish to thank the scientific reviewers for their detailed and helpful comments on an early draft of the manuscript, and in particular Dr. Ernest Williams for the exceptional time he kindly devoted to it. Finally, we express our deep appreciation to our respective families who have patiently endured our absences as we conducted the survey and completed this report.