Assessing the Resilience Potential of Inshore and O ﬀ shore Coral Communities in the Western Gulf of Thailand

: Coral reefs in the Gulf of Thailand have experienced severe coral bleaching events and anthropogenic disturbances during the last two decades. This study assessed the resilience potential of coral communities at Ko Losin o ﬀ shore reef sites and Mu Ko Chumphon nearshore coral reefs, in the south of Thailand, by conducting ﬁeld surveys on the live coral cover, hard substratum composition and diversity and density of juvenile corals. Most study sites had higher percentages of live coral cover compared to dead coral cover. Some inshore and o ﬀ shore reef sites showed low resilience to coral bleaching events. The total densities of juvenile corals at the study sites were in the range of 0.89–3.73 colonies / m 2 . The density of the juvenile corals at most reef sites was not dependent on the live coral cover of adult colonies in a reef, particularly for the Acropora communities. We suggest that Ko Losin should be established as a marine protected area, and Mu Ko Chumphon National Park should implement its management plans properly to enhance coral recovery and promote marine ecotourism. Other measures, such as shading, should be also applied at some coral reefs during bleaching periods.


Introduction
Coral reefs are recognized as a high-biodiversity ecosystem containing thousands of species that provide socioeconomic benefits. The benefits include providing food and livelihoods for millions of people in tropical countries and the protection of coastal communities from extreme weather disturbances [1,2]. However, coral reefs around the world are degrading because of natural stressors (bleaching, diseases and heavy storms [3][4][5][6][7][8][9]) and anthropogenic disturbances, particularly coastal development, pollution, sedimentation and overfishing [10][11][12][13]. Human impacts have also reduced the ability of coral recovery and reef resilience after severe disturbances [14][15][16]. Knowledge about the synergistic effects of coral bleaching and human activities on the ecological processes of coral reefs, particularly coral recruitment, is very important for establishing a science-based management strategy for enhancing the resilience potential of coral reefs [17].
Coral reef management requires supporting ecosystem processes that lower sensitivity, promote recovery, and enhance the adaptive capacity of coral reefs to bleaching by reducing other human impacts [18]. The capacity of coral reefs to resist or recover from degradation and to maintain their ecosystem services is defined as coral reef resilience [19]. Resilience-based management of

Materials and Methods
The study was conducted on coral communities in the Western Gulf of Thailand in March-May 2019. Six study sites from two different groups of coral communities, i.e., three study sites from Ko Losin offshore coral assemblages on pinnacles and three study sites from Mu Ko Chumphon nearshore coral reefs in Mu Ko Chumphon National Park, were selected for this study (Figure 1). Ko Losin is a small isolated island with an old lighthouse giving signals to boat navigators, about 72 km from the mainland. It has a relatively high water clarity in the Gulf of Thailand and harbors coral reefs that are well developed in deeper water, extending from 7 to 25 m depth. Ko Losin has been affected by fishing activities as it is an unprotected remote area. Recently, it is also used as a diving site in the Gulf of Thailand during the southwest monsoon period. Mu Ko Chumphon National Park is a marine protected area that is managed by the Department of National Parks, Plant and Wildlife Conservation. There are about 40 nearshore islands in Chumphon Province in the Western Gulf of Thailand, which harbor several coral reefs in good condition with high potential for tourism, particularly snorkeling and SCUBA diving. Three reef sites in Mu Ko Chumphon, i.e., Ko Kula, Ko Ngam Yai and Ko Ngam Noi, were selected for the field surveys. The coral reefs at the study sites were in shallow water, 1-12 m in depth. Ko Kula had relatively turbid water as it was affected by high sediment load from the mainland. The location, environmental conditions and anthropogenic disturbances at each study site are summarized in Table 1 Ngam Yai and Ko Ngam Noi, were selected for the field surveys. The coral reefs at the study sites were in shallow water, 1-12 m in depth. Ko Kula had relatively turbid water as it was affected by high sediment load from the mainland. The location, environmental conditions and anthropogenic disturbances at each study site are summarized in Table 1.   At each study site, the live coral cover was observed and evaluated as colony area/unit area in three belt-transects of 50 × 1 m 2 , coral colonies (≥5 cm in diameter) were counted and identified to the species level [44], if possible, and their coverage was quantitatively estimated. Covers of dead corals, rubble, sand, rock and other benthic components were recorded. In this study, covers of dead corals, rubble, rock and other benthic components were combined as available substrate. Quadrats were also photographed with an underwater camera for reinvestigating the data. Quadrats (50 × 50 cm 2 each) were randomly placed on available substrates at each study site by SCUBA divers, and the number of juvenile coral colonies (≤5 cm in diameter) was carefully observed, identified, counted and photographed for reconfirmation in the laboratory. All juvenile coral colonies were identified to the genus level [44].
Cluster analysis and the non-multidimensional scaling method were performed to categorize study sites on the basis of the Bray-Curtis similarity of benthic components, using PRIMER version 7.0. Differences in the taxonomic composition of corals between Ko Losin and Mu Ko Chumphon were tested by analysis of similarities (ANOSIM), and the coral species contributing most to the dissimilarity between the study sites were identified by similarity percentage (SIMPER) analyses. A one-way ANOVA was used to test the differences in the percentages of live coral cover, species diversity and juvenile coral densities among study sites. Where significant differences were found, the Tukey HSD (honestly significant difference) test was employed to determine which reef site(s) differed.

Results
There were significant differences in coral cover among study sites (one-way ANOVA, p < 0.05) (Figures 2 and 3). The highest percentages of live coral cover were found at Ko Ngam Noi (77.3 ± 9.3) and Ko Kula (57.7 ± 6.9) in Mu Ko Chumphon and at Ko Losin (West) (47.0 ± 18.0), Ko Losin (East) (45.7 ± 20.5) and Ko Losin (South) (26.7 ± 10.2), while the lowest coverage was observed at Ko Ngam Yai (5.4 ± 0.6). All study sites except Ko Ngam Yai had a higher percentage of live coral cover compared to dead coral cover. At each study site, the live coral cover was observed and evaluated as colony area/unit area in three belt-transects of 50 x 1 m 2 , coral colonies (≥5 cm in diameter) were counted and identified to the species level [44], if possible, and their coverage was quantitatively estimated. Covers of dead corals, rubble, sand, rock and other benthic components were recorded. In this study, covers of dead corals, rubble, rock and other benthic components were combined as available substrate. Quadrats were also photographed with an underwater camera for reinvestigating the data. Quadrats (50 x 50 cm 2 each) were randomly placed on available substrates at each study site by SCUBA divers, and the number of juvenile coral colonies (≤5 cm in diameter) was carefully observed, identified, counted and photographed for reconfirmation in the laboratory. All juvenile coral colonies were identified to the genus level [44].
Cluster analysis and the non-multidimensional scaling method were performed to categorize study sites on the basis of the Bray-Curtis similarity of benthic components, using PRIMER version 7.0. Differences in the taxonomic composition of corals between Ko Losin and Mu Ko Chumphon were tested by analysis of similarities (ANOSIM), and the coral species contributing most to the dissimilarity between the study sites were identified by similarity percentage (SIMPER) analyses. A one-way ANOVA was used to test the differences in the percentages of live coral cover, species diversity and juvenile coral densities among study sites. Where significant differences were found, the Tukey HSD (honestly significant difference) test was employed to determine which reef site(s) differed.

Results
There were significant differences in coral cover among study sites (one-way ANOVA, p < 0.05) (Figures 2-3). The highest percentages of live coral cover were found at Ko Ngam Noi (77.3 ± 9.3) and Ko Kula (57.7 ± 6.9) in Mu Ko Chumphon and at Ko Losin (West) (47.0 ± 18.0), Ko Losin (East) (45.7 ± 20.5) and Ko Losin (South) (26.7 ± 10.2), while the lowest coverage was observed at Ko Ngam Yai (5.4 ± 0.6). All study sites except Ko Ngam Yai had a higher percentage of live coral cover compared to dead coral cover.   All reef sites except Ko Kula harbored relatively high coral diversity. The highest resilience potential site was Ko Ngam Noi, which was dominated by Acropora spp. The high potential sites included Ko Kula, Ko Losin (West) and Ko Losin (East), while the low resilience-potential sites were Ko Ngam Yai and Ko Losin (South), which were dominated by Porites lutea (Figure 4). Overall, only Ko Ngam Yai had low resilience potential in terms of survival after bleaching and anthropogenic disturbances. The Shannon-Wiener index of diversity (H') was significantly different among the six study sites (one-way ANOVA, F = 25.27, p = 0.001). Tukey HSD tests showed that Ko Losin (East) was more diverse (H' = 1.7 ± 0.2) than Ko Kula (H' = 0.5 ± 0.1) ( Figure 5). All reef sites except Ko Kula harbored relatively high coral diversity. The highest resilience potential site was Ko Ngam Noi, which was dominated by Acropora spp. The high potential sites included Ko Kula, Ko Losin (West) and Ko Losin (East), while the low resilience-potential sites were Ko Ngam Yai and Ko Losin (South), which were dominated by Porites lutea (Figure 4). Overall, only Ko Ngam Yai had low resilience potential in terms of survival after bleaching and anthropogenic disturbances.  All reef sites except Ko Kula harbored relatively high coral diversity. The highest resilience potential site was Ko Ngam Noi, which was dominated by Acropora spp. The high potential sites included Ko Kula, Ko Losin (West) and Ko Losin (East), while the low resilience-potential sites were Ko Ngam Yai and Ko Losin (South), which were dominated by Porites lutea (Figure 4). Overall, only Ko Ngam Yai had low resilience potential in terms of survival after bleaching and anthropogenic disturbances. The Shannon-Wiener index of diversity (H') was significantly different among the six study sites (one-way ANOVA, F = 25.27, p = 0.001). Tukey HSD tests showed that Ko Losin (East) was more diverse (H' = 1.7 ± 0.2) than Ko Kula (H' = 0.5 ± 0.1) ( Figure 5).   ANOSIM indicated significant differences in the taxonomic composition of corals between Ko Losin and Mu Ko Chumphon (R = 0.52, p < 0.001, Figure 6). The average similarity in the composition of coral species between Ko Losin and Mu Ko Chumphon ranged from about 41.64% to 69.62%, whereas dissimilarity between Ko Losin and Mu Ko Chumphon was 59.74% (SIMPER analysis), Table 2.   ANOSIM indicated significant differences in the taxonomic composition of corals between Ko Losin and Mu Ko Chumphon (R = 0.52, p < 0.001, Figure 6). The average similarity in the composition of coral species between Ko Losin and Mu Ko Chumphon ranged from about 41.64% to 69.62%, whereas dissimilarity between Ko Losin and Mu Ko Chumphon was 59.74% (SIMPER analysis), Table 2.   ANOSIM indicated significant differences in the taxonomic composition of corals between Ko Losin and Mu Ko Chumphon (R = 0.52, p < 0.001, Figure 6). The average similarity in the composition of coral species between Ko Losin and Mu Ko Chumphon ranged from about 41.64% to 69.62%, whereas dissimilarity between Ko Losin and Mu Ko Chumphon was 59.74% (SIMPER analysis), Table 2.   The two-dimensional non-metric multidimensional scaling (NMDS) plot of the study sites based on the live corals, dead corals and other benthic components revealed that there were three groups of study sites, i.e., all three study sites of Ko Losin, Ko Kula and Ko Ngam Noi study sites, and Ko Ngam Yai study site (Figure 7). The two-dimensional non-metric multidimensional scaling (NMDS) plot of the study sites based on the live corals, dead corals and other benthic components revealed that there were three groups of study sites, i.e., all three study sites of Ko Losin, Ko Kula and Ko Ngam Noi study sites, and Ko Ngam Yai study site (Figure 7). Underwater photographs of the six study sites are shown in Figure 8. All study sites at Ko Losin and Ko Ngam Noi still displayed high live coral cover of Acropora spp., indicating that these reef sites were highly resilient to the coral bleaching events in 1998, 2010 and 2016. Underwater photographs of the six study sites are shown in Figure 8. All study sites at Ko Losin and Ko Ngam Noi still displayed high live coral cover of Acropora spp., indicating that these reef sites were highly resilient to the coral bleaching events in 1998, 2010 and 2016. The total densities of juvenile corals, i.e., those less than 5 cm in diameter, at the study sites were in the range of 0.89-3.73 colonies/m 2 . The highest average density of juvenile corals was found at Ko Ngam Yai (3.73 colonies/m 2 ), while the lowest average density was found at Ko Losin (West) (0.89 colonies/m 2 ). The total density of juvenile corals at Ko Ngam Yai was significantly higher than that at Ko Ngam Noi, Ko Kula and all study sites of Ko Losin (one-way ANOVA; Tukey's HSD test; p < 0.05) (Figure 9  The total densities of juvenile corals, i.e., those less than 5 cm in diameter, at the study sites were in the range of 0.89-3.73 colonies/m 2 . The highest average density of juvenile corals was found at Ko Ngam Yai (3.73 colonies/m 2 ), while the lowest average density was found at Ko Losin (West) (0.89 colonies/m 2 ). The total density of juvenile corals at Ko Ngam Yai was significantly higher than that at Ko Ngam Noi, Ko Kula and all study sites of Ko Losin (one-way ANOVA; Tukey's HSD test; p < 0.05) (Figure 9 Figure 9. Densities of juvenile corals (mean ± SD) on available substrate at the study sites (one-way ANOVA, p < 0.05). Different letters above bars indicate statistical differences (p < 0.05), as determined by Tukey's HSD. Figure 9. Densities of juvenile corals (mean ± SD) on available substrate at the study sites (one-way ANOVA, p < 0.05). Different letters above bars indicate statistical differences (p < 0.05), as determined by Tukey's HSD. ANOSIM indicated significant differences in the composition of juvenile corals between Ko Losin and Mu Ko Chumphon (R = 0.63, p < 0.001, Figure 11). The average similarity in the composition of juvenile corals between Ko Losin and Mu Ko Chumphon ranged from about 43.17% to 73.68%, whereas dissimilarity between Ko Losin and Mu Ko Chumphon was 63.81% (SIMPER analysis), Table 3.

Ko Losin
Mu Ko Chumphon Figure 10. Composition of the juvenile corals on available substrate at the study sites. Error bars indicate standard deviation.
ANOSIM indicated significant differences in the composition of juvenile corals between Ko Losin and Mu Ko Chumphon (R = 0.63, p < 0.001, Figure 11). The average similarity in the composition of juvenile corals between Ko Losin and Mu Ko Chumphon ranged from about 43.17% to 73.68%, whereas dissimilarity between Ko Losin and Mu Ko Chumphon was 63.81% (SIMPER analysis), Table 3.  1.33 Figure 11. Two-dimensional NMDS plot of the composition of juvenile corals at the study sites. The juvenile coral densities of the brooder Pocillopora were relatively high at Ko Ngam Noi (0.37 ± 0.15 colonies/m 2 ), Ko Losin Pinnacle (South) (0.66 ± 0.08 colonies/m 2 ) and Ko Losin (West) (0.44 ± 0.05 colonies/m 2 ). The juvenile coral densities of broadcast spawners at the study sites of Mu Ko Chumphon were much higher compared to those at the study sites of Ko Losin (Figure 12). Underwater photographs of the dominant juvenile corals, Pocillopora, Porites and Tubastraea, at the six study sites are shown in Figure 13. The juvenile corals were in healthy conditions without any signs of partial mortality or stress from competitors, diseases and bleaching.

Discussion
The coral reefs in the Gulf of Thailand are developed in high turbidity and have experienced severe coral bleaching events during the last two decades. The impacts of coastal development, destructive fishing and the expansion of tourism on coral reefs are documented [12,40]. The coral communities at Ko Losin (West), Ko Losin (East) and Ko Ngam Noi are interesting due to their high percentages of live coral cover and the fact that the dominant corals of these reef sites are several species of Acropora, which are susceptible to abnormal high-temperature-driven coral bleaching [7,45]. The coral communities at the study sites of Ko Losin are in relatively deep water, which may have protected them from high temperatures during the severe coral bleaching events in 1998 and 2010. Some Acropora corals also showed a high degree of bleaching but they did not die after bleaching. Intensive studies on ocean currents and other related issues of physical oceanography are required for understanding high resistance to bleaching events. Protection of the coral communities at Ko Losin from negative impacts of human activities, particularly fishing, boat anchoring and diving, is urgently needed to enhance coral reef resilience in the Gulf of Thailand.
The density of juvenile corals in the Gulf of Thailand is usually lower compared to that of other reef sites in the Indo-Pacific region [46]. Therefore, the coral communities in the Gulf of Thailand can maintain their community structures through the survival of resistant and/or tolerant coral species.

Discussion
The coral reefs in the Gulf of Thailand are developed in high turbidity and have experienced severe coral bleaching events during the last two decades. The impacts of coastal development, destructive fishing and the expansion of tourism on coral reefs are documented [12,40]. The coral communities at Ko Losin (West), Ko Losin (East) and Ko Ngam Noi are interesting due to their high percentages of live coral cover and the fact that the dominant corals of these reef sites are several species of Acropora, which are susceptible to abnormal high-temperature-driven coral bleaching [7,45]. The coral communities at the study sites of Ko Losin are in relatively deep water, which may have protected them from high temperatures during the severe coral bleaching events in 1998 and 2010. Some Acropora corals also showed a high degree of bleaching but they did not die after bleaching. Intensive studies on ocean currents and other related issues of physical oceanography are required for understanding high resistance to bleaching events. Protection of the coral communities at Ko Losin from negative impacts of human activities, particularly fishing, boat anchoring and diving, is urgently needed to enhance coral reef resilience in the Gulf of Thailand.
The density of juvenile corals in the Gulf of Thailand is usually lower compared to that of other reef sites in the Indo-Pacific region [46]. Therefore, the coral communities in the Gulf of Thailand can maintain their community structures through the survival of resistant and/or tolerant coral species. The results of this study suggest that highly resistant and tolerant coral species at Ko Losin, Ko Ngam Noi and Ko Kula play a major role in the high resilience potential of coral communities after coral bleaching events. The Acropora communities at Ko Ngam Noi, Mu Ko Chumphon National Park, are particularly important to the high resilience potential of nearshore reef sites. These coral communities may provide larval supply to nearshore reefs along the Western Gulf of Thailand through the connecting sea surface current in the Gulf of Thailand [47].
The poor coral condition at Ko Ngam Yai and the high percentage of dead corals at Ko Kula in Mu Ko Chumphon National Park imply the need for urgent investigation on how to restore these reef sites. The densities of juvenile corals at Ko Ngam Yai and Ko Kula from this study were relatively high compared to those of other reef sites in the Gulf of Thailand. The dominant juvenile corals at Ko Ngam Yai were Pocillopora, Porites, Favites and Pavona, whereas the dominant juvenile corals at Ko Kula were Fungia, Porites and Pachyseris. Enhancing the survival rates of juvenile corals is crucial for coral recovery following bleaching events [38]. Sediment loaded from coastal development and tourism impacts should be carefully mitigated for passive coral reef restoration. A high diversity of healthy corals in a coral reef ecosystem is an important factor for enhancing reef resilience potential because it occupies the reef substrates and inhibits the settlement of other benthic organisms that are coral competitors [17]. The coral communities at Ko Kula and Ko Ngam Yai also require an adequate supply of coral larvae from other coral reefs in the Gulf of the Thailand to enhance their coral diversity.
The density of juvenile corals recorded in our study was 0.89-3.73 colonies/m 2 , which is comparable to that of the Palk Bay reef in the northern Indian Ocean [17] but is much lower than that of several reef sites in the Indo-Pacific region, in which the juvenile coral density at some reef sites was over 50 colonies/m 2 [48,49]. Variation in the juvenile coral density between the study sites of Mu Ko Chumphon and Ko Losin was obviously shown in this study. Several factors may influence this spatial variation in juvenile coral density, such as larval supply from the parent reef, larval mortality, reef connectivity, settlement and post-settlement mortality, grazing and sedimentation [50,51]. The density of the juvenile corals at Ko Losin (West), Ko Losin (East), Ko Ngam Noi and Ko Kula was not dependent on the live coral cover of adult coral colonies in a reef. Moreover, the Acropora communities at Ko Losin and Ko Ngam Noi had no juvenile corals in their communities.
This study shows that several coral reefs at Ko Losin and Mu Ko Chumphon in the south of Thailand had high resilience potential to coral bleaching events and anthropogenic disturbances because of their survival rates, although they had relatively low densities of juvenile corals. We suggest that Ko Losin should be established as a marine protected area under Thai laws to protect the healthy corals as well as to provide coral larvae to other coral reefs in the Gulf of Thailand. The results from this study also imply that Mu Ko Chumphon National Park should implement its management plans properly to enhance coral recovery at Ko Ngam Yai and Ko Kula. Resilience-based management may be applied to support natural processes that promote the resistance and recovery of corals [43]. The promotion of marine ecotourism can protect coral communities at tourist destinations as well as keep the tourist numbers below the carrying capacity of the reef sites. Other measures to enhance the resistance of corals during bleaching events and appropriate coral restoration projects should be also considered. The field shading experiments that were carried out on coral communities of Ko Ngam Noi should be applied at other reef sites to protect corals during bleaching periods [41].
Author Contributions: All of the authors collected data; M.S. and T.Y. conceived the idea; M.S., T.Y., C.C., S.P. and W.K. analysed the data and wrote the manuscript.
Funding: This research was funded by Thailand Science Research and Innovation (TSRI), National Science and Technology Development Agency (NSTDA) and a budget for research promotion from the Thai Government to Ramkhamhaeng University.