**1. 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 [3–9]) and anthropogenic disturbances, particularly coastal development, pollution, sedimentation and overfishing [10–13]. Human impacts have also reduced the ability of coral recovery and reef resilience after severe disturbances [14–16]. Knowledge about the synergistic e ffects of coral bleaching and human activities on the ecological processes of coral reefs, particularly coral recruitment, is very important for establishing a science-based managemen<sup>t</sup> strategy for enhancing the resilience potential of coral reefs [17].

Coral reef managemen<sup>t</sup> 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 managemen<sup>t</sup> of coral reefs includes assessing spatial variation in resilience potential and implementing appropriate

managemen<sup>t</sup> plans [18,20,21]. The assessment of the resilience potential of coral reefs was first developed after the coral bleaching event in the year 1998 and it focused on the physical and ecological characteristics of coral reefs that provide some reefs with greater resistance to and/or recovery from coral bleaching [22,23]. Several resilience indicators have been widely developed and proposed for assessing the ecological resilience of coral reefs [24–27].

Successful coral recruitment and juvenile survivorship play an important role in the maintenance of coral populations under normal natural conditions and following mass mortality from bleaching events [28–30]. The planktonic larval stage, settlement and juvenile coral are critical periods in the coral life cycle and have high mortality rates, particularly under stressful environments. Following coral bleaching events, most surviving adult corals show reduced fecundity and growth as well as decreased reproductive outputs and recruitment rates [18,31]. Therefore, coral recruitment is often used as a bioindicator of coral reef health, recovery rate and resilience potential after severe disturbances such as bleaching events. A high coral recruitment rate or high density of juvenile corals on natural substrates can lead to quick coral recovery of degraded reefs after coral bleaching events and anthropogenic disturbances [32]. Coral recovery is also controlled through grazing by herbivores, which limits algal growth [33]. Several environmental factors influence coral recruitment rates, particularly water pollution, overfishing and coastal development, which can a ffect coral competition ability, fecundity, fertilization success, settlement and survival of juvenile corals [34–36]. Coral recovery and the resilience potential of coral reefs are usually controlled by coral larval supply, recruitment rate, the survival rate of juvenile corals and high resistance/tolerance to environmental stresses [17,37,38].

Mass coral bleaching events in the Gulf of Thailand were reported in 1998, 2010 and 2016 [39–41]. There were significant di fferences in the susceptibility of coral species to bleaching events in the Gulf of Thailand between the years 1998 and 2010. The 2010 coral bleaching phenomenon at some reef sites, such as Ko Samui in the Western Gulf of Thailand, was more severe than the 1998 bleaching event [39]. The intensive study of coral bleaching in the Gulf of Thailand in the year 2016 revealed that the levels of coral bleaching varied significantly among the reef sites. A high severity level of coral bleaching, of about 70%, was recorded at Ko Ngam Noi, Chumphon Province, in the south of Thailand. The coral mortality following the 2016 bleaching event was approximately 18%, which was much lower than that of the 2010 coral bleaching event because the southwest monsoon started earlier, and therefore the seawater temperature dropped rapidly [41]. Previous studies defined resilience as the capacity of a system to absorb or withstand stressors, maintain its structure and functions in the face of disturbance and change and adapt to future challenges [42,43]. This study aims to assess the resilience potential, based on coverages of live coral, dead coral, rubble and other benthic organisms, of coral communities at Ko Losin o ffshore reef sites in Pattani Province and Mu Ko Chumphon nearshore coral reefs in Chumphon Province, in the south of Thailand. Field surveys on the live coral cover, hard substratum composition and diversity and density of juvenile corals were conducted to determine the resilience of the coral communities in the south of Thailand.

#### **2. 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 di fferent groups of coral communities, i.e., three study sites from Ko Losin o ffshore 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 a ffected 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.

**Figure 1.** Map of the study sites at Mu Ko Chumphon National Park, Chumphon Province and Ko Losin, Pattani Province.



At each study site, the live coral cover was observed and evaluated as colony area/unit area in three belt-transects of 50 × 1 m2, 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<sup>2</sup> 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.
