Destruction and Recovery of Hermatypic Coral Communities After the Mass Bleaching Event at Ishigaki Island

Galaxea, JCRS, 4: 53 - 61 (2002) RS Jc Japanese CoralReef S ociety

Destruction and recovery of hermatypic coral communities after the mass bleaching event at Ishigaki Island

Y. Fujioka

National Research Institute of Fisheries Science, Fisheries Research Agency, 6-1-21 Sanbashidori, Kochi 780-8010, Japan e-mail address [email protected]

Abstract: Destruction and recovery of hermatypic coral communities during and after the 1998 bleaching event were compared quantitatively among three representative field sites (inner reef flat, outer reef flat, and reef slope) at Urasoko Bay, Ishigaki Island, southwestern Japan. As a result of bleaching, corals drastically decreased in number and coverage at all study sites and final coral coverage, at the heavily destroyed outer reef flat, was only 8.2 % of that prior to bleaching. Differential susceptibilities and survival were noted among coral species and among study sites. Bleaching damage most severely affected the outer reef flat, followed by the reef slope, whereas it less severely affected the inner reef flat, where sea water temperatures were highest. The degree of bleaching response among sites was not simply thermal-dependent but species-related. Several species of Acropora (A. digitifera, A. nobilis, A. formosa, A. pulchra, A. microphthalma, A. hyacinthus, A. echinata, A. subglabra, and so on) suffered the most damage from bleaching, and consequently the rich coral communities disappeared from the outer reef flat and the reef slope. At the inner reef flat, some species of Montipora (M digitata, M cactus, M stellata, M aequituberculata) recovered considerably after bleaching, despite conspicuous bleaching during the early bleaching phase. The present results indicate the possibility that some species inhabiting under thermally severe conditions have potentially acclimatized enough to endure such stress, and consequently they could live for a long period of more than six months in a bleached condition. Although the species richness (number of species) decreased at all sites, the species diversity indices often increased as a few dominant species of Acropora were reduced, and as a consequence, species evenness increased.

Key words: hermatypic coral bleaching community; diversity Ryukyu Islands

INTRODUCTION from late August to mid-September, and ended with the approach of a typhoon in late September A mass bleaching episode of hermatypic cor- which lowered the ambient coastal sea water als was extensively reported from the major temperature by mixing with the cooler offshore coral reefs of the world in the summer of 1998 water. The affected area extended to almost all (Wilkinson 1998 Wilkinson 2000 Hoegh- the coral species from the surface down to depths Guldberg 2000). In the sea around southwest- greater than 35 meters (Yamazato 1999). Details ern Japan, most notably in the Ryukyu of the bleaching damage to hermatypic corals Islands, corals reefs experienced the severest were given in the above papers, but mostly they bleaching on record (Hasegawa et al. 1999 reported on the status during the bleaching phase Taniguchi et al. 1999; Kayanne et al. 1999 between summer and autumn. Limited informa- Yamazato 1999; Sugihara et al. 1999; Fujioka tion about the subsequent status of coral com- 1999; Loya et al. 2001), which coincided with a munities in the post-bleaching phase, after the massive temperature anomaly in the summer. end of the bleaching event in the winter, is Bleaching damage to hermatypic corals started available in the literature. The main objective in mid-June (Hasegawa et al. 1999), increased of the present study is to quantify and compare 54 Y. Fujioka

the bleaching damage between the mid- and post- from 30 to 31°C at the outer reef flat, and from bleaching phases in terms of the subsequent 29 to 30°C at the reef slope. Thus sea water community destruction andior recovery from temperatures were highest at the inner reef flat, bleaching damage. followed by the outer reef flat, and lower at the reef slope. The first field survey was carried out during MATERIALS AND METHODS September 2-20, 1998 for a detailed quantita- Ishigaki Island is located at the southwestern tive assessment of the initial status of coral com- part of the Ryukyu Islands, Japan, and is mostly munities in the early- to mid-bleaching phases. encompassed by well-developed fringing reefs. As some coral colonies were already dead due to Three representative coral communities were bleaching at this time, the status before bleach- selected for the quantitative field investigation ing (before the first survey) was estimated from at Urasoko Bay (24° 27'N, 124° 13'E), which is the recently dead portions of the colonies. situated on the northern coast of Ishigaki Recently dead portions of the colonies were dis- Island (Fig. 1). The three sites were inner reef tinguished in terms of the absence of living tis- flat, outer reef flat and reef slope. An aggre- sues and minimal algal (or cyanobacterial) gated coral reef community, dominated by sev- overgrowth on the surface of skeleton, and were eral arborescent Montipora colonies, was located ascribed as recently dead. The second survey on the inner reef flat at depths of 1-2 m. A rich using the same methods was made during coral community, dominated by arborescent and January 18-28, 1999 at the outer reef flat, nota- corymbose Acropora, was located on the exposed bly no bleached corals were observed at this habi- outer reef flat at depths of 1-2 m. A diverse tat in January. The second survey at the inner coral community, dominated by arborescent reef flat was made during March 11-21, 1999, as Acropora with elaborate branches, was located the bleaching continued partially at this site on the least exposed reef slope at depths rang- through January. ing from 8 to 18 m. Each study site was surveyed using a 10 m In the daytime at the beginning of September, x 10 m quadrat; iron pegs were driven into sea water temperatures varied from 30 to 32 the four quadrat corners to enable repeated ob- °C at the inner reef flat and reached 33°C dur- servations. Corals in the large quadrat were re- ing the daytime low water, whereas they ranged corded for each 1 m X 1 m subdivided unit

Fig 1. Map of the Ryukyu Islands (left) and Ishigaki Island (right), showing the location of the study sites (Urasoko Bay). Destruction and recovery of coral communities after bleaching 55

using the ICCE (Individual Counting and Cover ness (number of species) and diversity were cal- Estimate) method of Weinberg (1981). Thus, a culated for each study site. Coral species diver- hundred continuous replicate samplings were sity was calculated using two diversity indices, made within the total of 100 m2 area for each Simpson's Index 1/ A = 1/ (n(n-1)/N(N-1)) and study site. the Shannon-Weaver Index H'=- (nIN)log(n/N), Within each 1 m X 1 m quadrat, the per- which uses logarithmic base 2, where 'n' is the centage of coral coverage, number of colonies, number of individuals for each species and 'N' and bleaching status were visually assessed for is the total number of individuals. For these in- all hermatypic coral species. Bleached parts indices, coral coverage was substituted for the cluded the pale and completely white colored number of individuals (parameter 'n' and 'N'), to parts of colonies. Normal colored portions of the compare the relative species abundance. colonies were those with no whitened tissues and Evenness of the species component was calcu- a usual coloration. Recent background (non- lated using Pielou's Index J'=H'/log2S, in which bleaching) mortality was, if it occurred, likely to 'S' indicates the number of species . represent only in a small proportion of the present records. Thus, the surface area of coral colonies was classified as either normal, bleached, RESULTS or dead, and the percentage cover of these cate- In Fig. 2, changes in the percent coverage of gories was separately recorded for each species. live (normal + bleached) portions of colonies, Nishihira and Veron (1995) and Veron (1992) the number of live colonies, the number of were used for taxonomic identification of the cor- hermatypic coral species, and the three diversity als, with some species assigned as unidentified or indices were compared during and after the identified only to the generic level. Species rich- bleaching event.

Fig 2. Changes in the percent coverage, number of colonies, number of species, and two species diversity indices of hermatypic corals before, during (September 1998) and after (January-March 1999) the bleaching event at the three study sites. Coverage included both the normal and the bleached portions of colonies. The status 'before bleaching' was estimated from the data of the first survey ('September'). Vertical bars represent SD of the mean (n=100). • : inner reef flat, A: outer reef flat, •: reef slope 56 Y. Fujioka

Before bleaching, the highest percentage of live (39.8± 13.5) and the reef slope (20.3± 8.5) ; they coral cover was recorded at the outer reef flat decreased to only 4.4±3.2, 29.8±12.9 and 15. (80.4± 18.9), followed by the inner reef flat (72.3 5 ± 7. 7, respectively, by the next winter. The ± 22.0) and the reef slope (47.4 ± 26.0); they highest number of species (per 100 m) was re- were reduced to 57.2 ± 16.4, 66.6 ± 21.9 and 44.2 corded at the reef slope (79 spp.), followed by ± 25.2, respectively, due to the bleaching damage the outer reef flat (58 spp.) and the inner reef when recorded in early- to mid-September, and flat (36 spp.) before the bleaching they de- finally 6.6±6.0, 40.2±24.1 and 12.4±7.8 after creased to 65, 26 and 31, respectively, by the the end of the bleaching event in the next win- next winter. Thus the coral coverage, the num- ter. Final coral coverage, at the heavily de- ber of colonies, and the number of coral species stroyed outer reef flat, was only 8.2 (=6.6/80.4) decreased extensively. % of that prior to bleaching, whereas it was 26.2 Before the bleaching event, species diversity (=12.4/47.4) % at the reef slope, and 55.6 indices were highest at the reef slope, followed (=40.2/72.3) % at the inner reef flat. by the outer reef flat and the inner reef flat. At the three study sites, totals of 2027-6799 They exhibited three different trends during the colonies were distributed within the 100 m areas bleaching period they decreased slightly at the before bleaching. The highest number of colonies inner reef flat, fluctuated at the outer reef flat, per 1 m was recorded at the outer reef flat and increased at the reef slope. (68.0 ± 13.0), followed by the inner reef flat Fig. 3 shows changes in the spatial structure

Inner reef flat

Outer reef flat

Reef lope

Fig 3. Changes in the percent coverage of hermatypic corals for the one hundred 1 m X 1 m quadrats, during (September 1998) and after (January-March 1999) the bleaching event at the three study sites. Vertical axis represents the coral coverage for each quadrat and two horizontal axes are the sides of a 10 m X 10 m survey area. Destruction and recovery of coral communities after bleaching 57

of percent coral coverage for the one hundred 1 m siderable, at the inner reef flat in March 1999. x 1 m subdivided units located within 10 m X 10 Thus, bleaching damage was conspicuously dif- m large quadrat. Living coral colonies decreased ferent among the sites, and the most severely largely through bleaching between September affected was the outer reef flat, followed by the 1998 and January 1999 at the outer reef flat reef slope, and least affected was the inner reef and the reef slope, and consequently the com- flat. Results of one-way repeated measures munity changed to almost flat and bare spaces, analysis of variance (ANOVA) showed that there with a small number of live colonies and partially was a significant decrease in the percent cover- living tiny coral fragments. Ridges formed by bro- age of live corals between the first and the sec- ken coral skeletons were frequently observed in ond surveys (df=2,297, F=21.59, p<0.0001). sheltered areas or on the bottom of the bay in the Fig. 4 shows changes in the percent coverage of vicinity of these sites. In contrast, bleaching normal, bleached, and recently dead portions of damage was less severe, and survival was con- coral colonies for the ten dominant species at

Inner reef flat

Outer reef flat

Reef slope

Fig 4. Species-specific bleaching damage showing changes in the percent coverage of normal (not bleached), bleached and recently dead parts of colonies for the ten predominant species at each site between the early (September 1998) and post bleaching (January-March 1999) surveys. Differences in percent coverage between the two surveys indicate the amount of coverage lost by the second survey. 58 Y. Fujioka

each site. During the second survey (January- survived at the inner reef flat, and 22.3 % sur- March 1999), bleaching was rarely observed vived at the reef slope. Similarly, survival was and most dead coral skeletons had already comparatively low in some species belonging to been lost from the reef flat sites, while they re- Pocillopora (25.0-33.3 %), Seriatopora (29.0 %) mained considerably within the reef slope site. and Millepora (10.0-15.4 %). In contrast, the There were striking differences in the effect most tolerant to bleaching was the Agariciidae of the bleaching event among coral species. At (Pavona, Leptoseris and Pachyseris), followed by all the study sites, most species of Acropora the Poritidae (Porites and Goniopora), the Faviidae were severely affected by the bleaching, and re- (Favia, Favites, Goniastrea, Platygyra, Cyphastrea sulted in a decrease in percent coverage. In par- and Echinopora), the Dendrophylliidae (Turbinaria), ticular, high mortality rates occurred to six and the Siderastreidae (Psammocora). Species predominant species of Acropora, i.e. Acropora within the genera Montipora, Fungia and digitifera (Dana), Acropora nobilis (Dana), Acropora Lobophyllia, had survival differing considerably formosa (Dana), Acropora pulchra (Brook), Acropora among study sites; in all cases they were higher microphthalma (Verrill) and Acropora hyacinthus at the reef slope than at the reef flat. (Dana), and only a few tiny fragments, up to 1 % in coverage, remained alive in January 1999. Bleaching damage of Acropora formosa differed DISCUSSION among the study sites; it was greatest at the Prior to 1998, coral bleaching events in the inner reef flat, followed by the outer reef flat Ryukyu Islands of Japan had received little at- with least damage at the reef slope. At the reef tention because they were generally considered slope, bleaching for Seriatopora hystrix Dana and to be temporary, local, and sublethal events, ex- three Acropora species, Acropora echinata (Dana), cept for some thermally sensitive species that Acropora subglabra (Brook) and Acropora carduus sustained partial damage (Yamazato 1981; (Dana), was not severe during the early bleach- Kamezaki & Ui 1984; Kato 1987; Environment ing phase, but final survival ranged from only Agency 1997). However, the coral bleaching 1.3-16.2 %, indicating that the damages pre- event of the 1998 summer was catastrophic com- dominantly occurred during the late bleaching pared with previous events (Hasegawa et al. phase. At the inner reef flat, Acropora tennis 1999; Taniguchi et al. 1999; Kayanne et al. (Dana) and Acropora aspera (Dana) had recov- 1999; Yamazato 1999; Sugihara et al. 1999; ered partly from the bleaching by the second Fujioka 1999; Loya et al. 2001). As has shown survey. in this study, some coral communities were Four species of Montipora, i.e. Montipora digitata mostly destroyed while others had more or less (Dana), Montipora cactus (Bernard), Montipora recovered from the bleaching. stellata (Bernard) and Montipora aequituberculata Differential susceptibilities and survival was Bernard, were severely affected by bleaching noted among coral species and among study sites. during the early bleaching phase, but they had In the case of the same species, survival seemed recovered considerably to the normal state by to be related to the environmental gradient the second survey. Massive colonies of Porites among habitats. However, mean survival of en- lutea Edwards & Haime and Psammocora contigua tire coral communities did not correspond with (Esper) were observed to be bleached during the thermal gradient among the three study the first survey and they had recovered com- sites, as shown in Fig. 2 and Fig. 3. Differences pletely by the second survey. Other massive in bleaching damage among sites seems not to be and encrusting corals, Favites abdita (Ellis & simply due to the thermal gradient among sites Solander), Goniastrea retiformis (Lamarck), but also markedly affected by the species compo- Platygyra ryukyuensis Yabe & Sugiyama, sition of the coral communities, because the least Echinophyllia aspera (Ellis & Solander) and damaged experienced the highest temperatures. Pachyseris speciosa (Dana), which were bleached to Under such severe conditions, arborescent varying degrees during the first survey, had Montipora species, such as M digitata, M stellata, partially recovered by the second survey. and M cactus, are frequently exposed to high In Table 1, final survival of coral species are water temperatures of more than 32°C in the summarized for the main genera and compared daytime during summer. The present results among study sites. Corals most susceptible to indicate the possibility that some species inhab- bleaching were Acropora, they disappeared from iting under thermally severe conditions have the outer reef flat, 19.1 % of the former coverage potentially acclimatized enough to endure such Destruction and recovery of coral communities after bleaching 59

Table 1. Comparisons of the tolerance to bleaching among the major coral genera showing the mean (± standard deviations) survival of the distributed species at each study site. Survival of species was calculated from the percent coverage during (September 1998) and after (January-March 1999) the bleaching event.

stress, and consequently they could live for a Sugihara et al. 1999 Fujioka 1999). However, in long period of more than six months in a comparison of the two main genera, Acropora bleached condition. and Montipora belonging to the Acroporidae, they Similar species specific susceptibilities were re- exhibited striking differences during the post- ported by Hoeksema (1991). He demonstrated bleaching phase a near total destruction of that mushroom corals in relatively stagnant Acropora communities from reef flat shallower water of lagoons and on the inner reef flat than 2 m deep, while Montipora had recovered showed little bleaching. Marshall and Baird considerably from the bleaching. The tolerance of (2000) also suggested that much of the spatial Montipora species to bleaching stress has hitherto variation in bleaching response was due to the been unreported. Some authors reported that assemblage composition and thermal acclima- the Pocilloporidae was the most thermally sensi- tion. The present results also show the degree of tive (Jokiel & Coles 1974; Jokiel & Coles 1990 bleaching response is not simply thermally- Coles 1975; Yamazato 1981; Yamazato 1999). dependent but species-related. This difference might be caused by either differ- The species of the Acroporidae are most ther- ences in the taxonomic assemblages of the study mally susceptible of all scleractinian corals sites and the degree of thermal stress, or both. (Fisk & Done 1985; Brown & Suharsono 1990; During the bleaching event, the species di- Gleason 1993; Hoegh-Guldberg & Salvat 1995; versity indices exhibited three different trends; Taniguchi et al. 1999; Kayanne et al. 1999; they decreased slightly at the inner reef flat, 60 Y. Fujioka

fluctuated at the outer reef flat, and increased warming in the Thousand Islands, Indonesia. Coral at the reef slope (see Fig. 2). These differences Reefs 8: 163-170 are due to the relative tolerances of several of Coles SL (1975) A comparison of effects of elevated temperature versus temperature fluctuations on the predominant species, including Acropora reef corals at Kahe Point, Oahu. Pac Sci 29: 15-18 pulchura, A, hyacinthus, and A. microphthalma at the Connell JH (1978) Diversity in tropical rain forests and outer reef flat and A. subglabra, A. carduus, and coral reefs. Science 199 1302-1310 A, echinata on the reef slope, which suffered the Environment Agency (1997) Studies on preservation highest damage and decreased in abundance, of coral reef ecosystem. Final Report of Global whereas previously subordinate species, belong- Environmental Research Fund, Environment Agency of Japan, 149pp (In Japanese) ing to the families Poritidae, Fungiidae, Fisk DA, Done TJ (1985) Taxonomic and bathymetric Pectiniidae, Mussidae, and Faviidae suffered less patterns of bleaching in corals, Myrmidon Reef severe damage, and as a consequence, species (Queensland). Proc 5th Int Coral Reef Congr 6: 149- evenness increased. 154 Fujioka Y (1999) Mass destruction of the hermatypic According to the intermediate disturbance hy- corals during a bleaching event in Ishigaki Island. pothesis, the very high diversity of coral reefs is Galaxea, JCRS 1: 41-50 maintained only in a non-equilibrium state Gleason MG (1993) Effects of disturbance on coral (Connell 1978). They are subjected to distur- communities bleaching in Moorea, French Polynesia. bances often enough that equilibrium may never Coral Reefs 12: 193-201 Hasegawa H, Ichikawa K, Kobayashi M, Kobayashi be attained, and form archetypical non- T, Hoshino M, Mezaki S (1999) The mass-bleaching equilibrium communities whose structure is de- of coral reefs in the Ishigaki lagoon, 1998. Galaxea, termined by interacting rates of competitive JCRS 1: 31-39 displacement and by a range of disturbance Hoegh-Guldberg O, Salvat B (1995) Periodic mass- types (Huston 1985). The different trends in the bleaching and elevated sea temperatures bleaching species diversity that were observed in the pre- of outer reef slope communities in Moorea, French Polynesia. Mar Ecol Prog Ser 121 181-190 sent study are considered to be caused by the Hoegh-Guldberg O (2000) Global climate change and difference in the degree of disturbance by the thermal tolerance of corals. Galaxea, JCRS 2: bleaching. If disturbances are moderately se- 1-11 vere as at the reef slope, a higher degree of diver- Hoeksema BW (1991) Control of bleaching in mushroom coral populations (Scleractinia: Fungiidae) in sity may be sustained due to the increasing the Java Sea: stress tolerance and interference by evenness of the species assemblage. The present life history strategy. Mar Ecol Prog Ser 74: 225-237 results provide a good example of the interme- Huston MA (1985) Patterns of species diversity on coral diate disturbance hypothesis under a non- reefs. Ann Rev Ecol Syst 16: 149-177 equilibrium state. Jokiel PL, Coles SL (1974) Effects of heated effluent on hermatypic corals at Kahe Point, Oahu. Pac Sci 28: 1-18 ACKNOWLEDGEMENTS Jokiel PL, Coles SL (1990) Response of Hawaiian and other Indo-Pacific reef corals to elevated temperature. Coral Reefs 8: 155-162 The author would like to express his sincere Kamezaki N, Ui S (1984) Bleaching of hermatypic corals in Yaeyama Islands. Marine Park J 81: 10-13 thanks to Dr. Haruki Ochi, Kyoto University, (In Japanese) for his assistance during the field surveys. The Kato M (1987) Mucus-sheet formation and discolora- original draft of the manuscript was improved tion in the reef-building coral, Porites cylindrica: ef- by Dr. Barbara E. Brown, University of fects of altered salinity and temperature. Galaxea Newcastle, whose efforts the author sincerely 6: 1-16 Kayanne H, Harii S, Yamano H, Tamura M, Ide Y, appreciates. Thanks also extend to Dr. Michio Akimoto F (1999) Changes in living coral coverage Hidaka, Dr. Robert van Woesik, Dr. Hideo before and after the 1998 bleaching event on coral Ohba and anonymous reviewers for their helpful reef flats of Ishigaki Island, Ryukyu Islands. Galaxea, comments on the manuscript. This work was sup- JCRS 1: 73-82 Loya Y, Sakai K, Yamazato K, Nakano Y, Sambali ported by the Global Environment Research H, Woesik R (2001) Coral bleaching the winners Fund, Environment Agency of Japan. and losers. Ecol Let 4: 122-131 Marshall PA, Baird AH (2000) Bleaching of corals on the Great Barrier Reef: differential susceptibilities REFERENCES among taxa. Coral Reefs 19: 155-163 Nishihira M, Veron JEN (1995) Hermatypic corals of Japan. Kaiyusya, Tokyo, 440pp Brown BE, Suharsono (1990) Damage and recovery Sugihara K, Iryu Y, Nakamori T (1999) Coral of coral reefs affected by El Nino related seawater Destruction and recovery of coral communities after bleaching 61

bleaching, geological ranges, and adaptation to 白化によって出現種数はすべての地点で減少したも high sea surface temperatures. Galaxea, JCRS 1 : 89-95 のの、多様度指数は調査地点ごとに異なった変化パター Taniguchi H, Iwao K, Omori M (1999) Coral bleach- ンを示した。これは白化による撹乱の程度に起因する ing around Akajima, Okinawa, I. a report of the ものと考えられ、優占していた全てのミドリイシ類が September 1998 survey. Galaxea, JCRS 1: 58-64 壊滅的な被害を被った外側礁原部では多様度指数が急 Veron JEN (1992) Hermatypic corals of Japan. 激に減少したが、一部のミドリイシ類の群体が生き残っ Australian Inst Mar Sci, 234pp Weinberg S (1981) A comparison of coral reef survey た礁斜面では均等度が増して多様性が高くなるという methods. Bijd Dierk 51: 199-218 現象が認められた。この結果は、サンゴ礁における高 Wilkinson C (1998) The 1997-1998 mass bleaching い多様性が、非平衡状態下で維持されているという従 event around world. In: Wilkinson C (ed) Status 来からの考え方を支持するものである。 of coral reefs of the world: 1998. Australian Inst Mar Sci, pp 15-38 Wilkinson C (2000) The 1997-98 mass coral bleaching and mortality 2 years on. In: Wilkinson C (ed) Status of coral reefs of the world: 2000. Australian Inst Mar Sci, pp 21-34 Yamazato K (1981) A note on the expulsion of zooxanthellae during summer, 1980 by the Okinawan reef building corals. Sesoko Mar Sci Lab Tech Rep 8: 9-18 Yamazato K (1999) Coral bleaching in Okinawa, 1980 vs 1998. Galaxea, JCRS 1: 83-87

(Received 2 June 2002/Accepted: 4 September 2002)

石垣島における大規模白化後にみられた造礁サンゴ群集の崩壊と回復

藤岡義三(中央水産研究所)

1998年夏に起こった造礁サンゴの白化による群集の崩壊とその後の回復について、石垣島浦底湾の代表的な3定点(内側礁原部、外側礁原部、礁斜面)において比較調査を行った。すべての調査地点で、造礁サンゴの被度、群体数、種数は大きく減少し、白化終了時の最終的な被度は白化以前の状態に比べて、外側礁原部でわずか8.2%、礁斜面で同26.2%、内側礁原部で同55.6%となった。白化による影響は種や地点によって異なり、外側礁原部で最も被害が大きく、礁斜面がこれに続いたが、最も水温が高かった内側礁原部では被害が小さかった。このことから白化による影響は、単に水温に依存するのではなく、群集の種組成によって異なるものと推察された。ミドリイシ属数種 (Acropora digitifera, A. nobilis, A. formosa, A. pulchra, A. microphthalma, A. hyacinthus, A. echinata, A. subglabra 等)は最も被害が著しく、外側礁原部および礁斜面のミドリイシ群集はほぼ完全に消失した。一方、内側礁原部に生息するコモンサンゴ属数種 (Montipora digitata, M cactus, M stellata, M aequituberculata) は、夏期の高水温時に急速に白化したにもかかわらず、白化終了時には比較的高い割合でもとの正常な状態まで回復した。このことは高水温下に曝される機会が多いサンゴほど、潜在的に温度耐性が高いという可能性を示すものであり、そのような種では6ヶ月以上の長期間にわたって白化した状態のまま生存することが可能であった。