Establishing a baseline

To develop predictive power of the spawning time for the greatest diversity of scleractinian corals, a baseline of optimal spawning window was established for scleractinian corals in the Gulf of Thailand. Using a combination of our own unpublished data and previously published records on gamete development and broadcast spawning observations from the Gulf of Thailand [30,40,43,46], distinct spawning windows were identified. This was supported by observations by the monitoring teams of the Department of Marine and Coastal Resources (DMCR, pers. Comm.; S1 Table). The resulting dataset identified two discrete spawning periods during the year and revealed that the spawning and development information for the Agariciidae was represented by only two species of Pavona corals, both spawning during daytime. Using this baseline, a more specific investigation was carried out to 1) identify the specific spawning time for Pavona corals along the Eastern Gulf of Thailand and 2) investigate other instances of daytime spawning within the Gulf of Thailand.

Defining spawning periods

To test the reproducibility and define general spawning window of corals, in-situ roving surveys were carried out using SCUBA at 30 islands across three provinces (Chonburi, Rayong and Trat) along the Eastern Gulf of Thailand between 2021 and 2024. It is important to note that surveys were restricted to periods between January and April, and September and December, as annual monsoon conditions between May and August prevented access to sites and SCUBA surveys. These surveys were carried out at fringing coral reefs and pinnacles with the aim of observing coral spawning (gamete release), with data on time of spawning, gamete type (sperm, eggs, bundles), species (where readily identifiable), and other metadata collected (S1 Table). The majority of surveys were conducted during the daytime to identify daytime spawning corals, with night-time surveys typically carried out during established spawning periods. Night-time surveys in the months of March and April 2023 and 2024 were heavily supplemented by the use of a remotely operated vehicle (ROV) due to growing observations and incidents of Chironex sp. box jellyfish (S2 Fig) [47,48]. Spawning corals observed via ROV were initially identified to genus and tabulated during the survey, and then were later verified or disregarded based on review of the footage (S3 Fig), with the exception of a single survey where footage recording was corrupted.

Additionally, surveys with multiple parallel data collection efforts such as reef monitoring [e.g., 49,50], discarded fishing gear removal [51] etc., which were often being carried out concurrently at these same sites, are included in our dataset. These surveys, while often not focused on observing coral spawning events, did nonetheless find presence and absence of daytime spawning and are thus included in the dataset with the assumption that mass synchronised spawning of specific taxa have the potential to be detected by trained surveyors (and incidentally by recreational divers). We have thus referred to any survey in this broader dataset, or any survey explicitly looking for spawning but where spawning was not observed, as ‘null data’. Null data surveys should be considered as surveys where multiple colonies were directly observed, typically briefly and usually as transect or roving-diver surveys, where no gamete release was observed. Null data surveys also included more prolonged observations of individual colonies where an absence of spawning was noted. Therefore, null data observations may be considered as defined and specified periods of time where multiple colonies across multiple taxa were specifically assessed for varying periods of time without any indication of sexual reproductive activity. The baseline community structure of corals for all null data survey sites was recently assessed to contextualise reef composition [52]. In addition, four incidental observations of daytime coral spawning from the Gulf of Thailand by academically inclined members of the recreational diving community were submitted to the authors for review and subsequent inclusion in the dataset. No observations were included where photo/video evidence, precise timing of spawning, and number of spawning colonies were lacking.

Corals were identified to species level where possible. However, night-time spawning corals are presented only at the genus level in this study, owing to the focus on daytime observations and the difficulty of accurate taxonomic identification without sampling. Identification was conducted during daytime surveys across the reefs, based on living specimens or, where possible, the skeletal morphology of recently killed corals (S4 Fig), using original descriptions and relevant literature (e.g., [34,38,53,54]) for corals. Taxonomy was updated according to World Register of Marine Species (WoRMS) [55]. Daytime spawning corals were identified to species level through close in-situ examination of skeletons and colony morphology. Gravid colony assessments and validation were not performed due to the rapidly declining coral populations at the survey sites [51,56,57].

Pavona surveys

After the first year of surveys, general patterns of spawning were identified for the two species of Pavona, P. explanulata and P. varians. Therefore, alongside aforementioned generalised day and night-time surveys, dedicated surveys between 2022–2024 were carried out to study the spawning of P. explanulata and P. varians across the Eastern Gulf of Thailand, including two islands in Cambodian waters. These surveys were carried out in October, November and December, between 3 days before the lunar peak and 3 days after the lunar peak, between 12:00 and 17:00, (based on weather conditions). The purpose of these surveys was to investigate the synchrony of gamete release across the > 450km of coastline along the eastern Gulf of Thailand, encompassing both Thailand and Cambodian waters, and to develop a more robust understanding of reproductive outputs of Pavona corals in the region.

Data analysis

To review and compare findings from our surveys and explore biogeographical or taxonomic trends, spawning data were collected from all available published literature globally on Pavona spawning [31,36,43,58–62]. For all data including spawning observations from present study and previous studies compiled, and null data, the following information was collected where relevant: location; identity of spawning coral (to the most reliable taxonomic level); precise date and time of observation duration, gamete bundle setting; and release. Where possible, the number of colonies and last observed spawning time were also collected. From these spawning data were then calculated the number of minutes and hours of spawning onset since the most recent sunrise, the number of hours since the nearest full moon (FM), and the number of days since the start of the year (as a higher resolution proxy for month and season). Precise data for sunrise time and FM timing was gathered and cross-referenced from online and published sources (see [33]). It is important to note that data used for this purpose utilised lunar peak timing as defined by syzygy of the Sun-Earth-Moon system as opposed to by illumination. Therefore, we are able to greatly increase the precision of our calculations by using minutes after lunar peak, rather than nights after the nearest visible FM to greatly improve the resolution of spawning predictions. Data were then visualised with plots generated using XLstat 2024.4.1.

Synchrony in the Gulf of Thailand

Based on spawning records from 14 genera from the Gulf of Thailand, our data provide strong credence for synchrony of coral spawning across the region, with compelling evidence to suggest spawning cues based on lunar and solar timing. The majority of spawning corals clustered tightly based on lunar phase (Fig 3B) and number of hours post sunrise (Figs 3,4). Our findings expand on earlier observations, revealing that Pavona explanulata and P. varians are daytime spawners across the Gulf of Thailand and are highly synchronised across the region. Null data and night-time surveys revealed that spawning of these corals is restricted (and thus predicted by) a narrow window of hours after sunrise and the timing of the full moon. The reproductive season for these corals across the Gulf of Thailand appears to largely peak in October and November, however there are indications of timing offset between the eastern and western Gulf. The coral spawning data presented here contributes to the limited prior documentation of synchronous mass spawning events from the Gulf of Thailand. Overall, our inferences on months of development based on observed reproductive output for night-time spawning corals agree with prior findings [38,46]. Additionally, the number of days after full moon for onset of coral spawning agree with records from Koh Tao ([43], Appendix A).

Questions remain regarding the close clustering of the timing of most Pavona spawning recorded from the Gulf of Thailand as distinct from those reported from seven colonies from Koh Rangka Jeaw in Chumphon province [30]. Based on the single image shared from Chumphon [30] and our own surveys in the area, we maintain a high credence in our identification of earliest daytime spawning corals in Thailand corresponding to P. explanulata. Further exploration of Pavona spawning along this area are needed to explore possible differences in spawning time. Our dataset agrees with earlier observations [38,42,43] that night-time spawning periods differ for corals along the eastern and western Gulf of Thailand. Eastern Gulf night-time spawners typically spawn in the lunar cycles of February and March (with January and April being exceptionally rare), whereas western Gulf colonies typically spawn in March and April, with February and May records being almost absent. During days of concurrent observation, Rayong Province, Thailand and Koh Rong Sanloem Cambodia, onset of spawning was found to differ by 23 minutes for P. varians and 85 minutes for P. explanulata. Notably, on multiple occasions, where synchronicity was observed in spawning between Rayong and Trat provinces in Thailand, and Koh Rong Sanloem and Koh Koun in Cambodia, an absence of spawning was found at sites in Chonburi province. It is unclear if this is indicative of some localised disturbance of reproductive capacity, or if a breakdown in synchrony might be caused by other factors [67]. Lack of synchrony in spawning events has been shown to drastically reduce gene exchange between colonies [68–70] and therefore act as a driver for genetic divergence [70,71]. Divergences by as little as 1–3 hours in spawning time has been shown to lead to independent genetic clades in congeners [72]. The presently available data, while limited, supports synchrony in spawning within specific regions of the Gulf of Thailand, as well as to a lesser degree, between regions.

Assessing non-invasive methods to study coral reproduction

Based on the data presented here, we have demonstrated the utility (and possibly the extent) of coral spawning data collection and prediction using in-situ observation-only techniques. While the sampling of colonies for the purpose of reproductive (i.e., gamete development) or taxonomic investigation will likely remain an essential tool, such activities should be carried out sparingly in areas where coral decline has been significant. Nonetheless, absence of spawning observations from other coral genera in our dataset highlights a limitation of the in-situ, non-invasive techniques employed in this study, suggesting that histological and gamete-development investigations remain the best tool for initial investigations, for taxa and regions where some baseline data is sparse or lacking. Our work thus far from Rayong and Surat Thani provinces ([43,50,68]; present study) allows for the identification of multiple abundant groups (e.g., Astreopora spp., Leptastrea spp., foliose Pavona spp., etc.) which comprise significant proportions of Thai coral communities [52] but for which regional spawning data is sparse or entirely absent. Indeed, our recent assessment of coral communities across much of Thai waters have identified Pavona (particularly foliose Pavona taxa) to be abundant throughout the region [52]. Given multiple years of surveys from across the Gulf of Thailand, the absence of spawning from these groups in the region (and sparsity from global data) highlight the value of dedicated histological assessments to investigate the role of timing among other factors on the reproduction of these corals.

Global spawning patterns of Pavona

By far the most extensive assessment of reproductive dynamics of Pavona varians was carried out by Glynn et al. [36] along the Pacific coast of Panama. Based on an eight-year dataset of spawning observations and a long-term sampling regime investigating reproductive dynamics, P. varians was found to spawn shortly before sunrise (up to 2 hours before) and to be generally hermaphroditic with only a small portion of colonies being strictly gonochoric. However, the proportion of hermaphroditic vs. gonochoric colonies appeared to vary considerably across sites. The authors categorised the species as an ‘alternating sequential hermaphrodite’ Glynn et al. [36] where the spawning of one sex is preceded by the other within a given colony, lunar phase, and breeding season. These observations alongside additional data on gametogenic development also suggested nearly year-round reproductive activity across the range from Costa Rica, Panama and the Galapagos Islands. The earliest indications of daytime spawning in P. varians were recorded in a single-page abstract by Maté [32] where colonies from Hawaii were recorded to spawn prior to sunset at 19:05–20:15. In both these reports, the spawning of the species was recorded shortly before or after a change in illumination (sunrise or sunset) but thus far no observations have demonstrated spawning exclusively during periods of peak illumination (within a 24-hour cycle), as was recorded from the Gulf of Thailand. An investigation into gamete development of P. varians from the Red Sea inferred spawning to occur in April [73], further highlighting the distinction observed from the Gulf of Thailand. Additionally, the majority of P. varians spawning observations recorded across the Indo-Pacific occur between FM and 4 days after the lunar peak. The compiled dataset does appear to suggest that the onset of spawning within a given month is mediated by the lunar phase, however synchronicity across the calendar year and within a 24-hour period coincided with longitude separations more than latitude.It is possible that future observations during the survey months not covered in our work (May-August) may yield data that improves our understanding.

In contrast to P. varians, little is known about the reproductive biology of P. explanulata. Outside of the Gulf of Thailand, the only record of spawning we found in the literature for a coral identified as Pavona explanulata comes from Taiwan [61]. The authors provide a single record of night-time (9:30 pm) egg release six days after FM in May, however no further descriptions nor details are provided. Our observations differ considerably from this record in almost every way, except for colonies being identified as gonochoric. Our observations from the eastern Gulf of Thailand demonstrate a predictable onset of spawning shortly after P. varians, typically between 15:00–16:00, rarely extending beyond 17:00, and a definitive lack of spawning immediately prior to or after sunset (during surveyed periods). The earliest record of daytime spawning from the Gulf of Thailand [30], however, differs from the remainder of our dataset of observations for the species due to spawning one day prior to the new moon, on 27th August 2003, during low tide. Curiously, the authors state this date to be the third day after the full moon, possibly suggesting an incorrect date of spawning, however for the purpose of this work we assume a misinterpretation of the lunar calendar. It is noteworthy that Glynn et al. [36] also identified a relationship between onset of spawning and the new moon as well as peak spring tides. While our study did not specifically investigate the relationship between tidal height and spawning time, the absence of spawning observations during new moon conditions across most of our dataset suggest that tidal influence in the Gulf of Thailand may be at most secondarily involved.

Given the wide range of spawning seasonality identified for P. varians across its range, it is unsurprising to find variability in spawning of P. explanulata. Indeed, when compared with other Pavona species, considerable variability is noted in the spawning period during the calendar year and within a given day. For example, alongside P. varians, Glynn et al. [36] also provide considerable data on P. chiquiriensis (initially identified as Pavona sp. prior to its description), which was found to spawn around sunset approximately 12 hours offset from P. varians from the same sites. Meanwhile P. clavus colonies from nearby Panamanian islands was also found to spawning around sunset [59], however consistently between August to October, as opposed to in January, April and July as recorded for P. chiquiriensis. Similarly, P. cactus, P. decussata, and P. explanulata have all been recorded to spawn after sunset in Taiwan, however the latter species spawning in May while the foliose taxa were found to spawn in August [58,61]. It is noteworthy that relatively little is known about the spawning of the abundant foliose members of the genus, with only four colonies of P. cactus and two colonies of P. decussata found in the literature, and nothing for the others. Our observations in this study lend credence to P. frondifera and P. decussata (abundant across the Gulf of Thailand) are possibly night-time spawning corals based on our null data. Across the genus-level, some possible trends of Pavona spawning are hinted at across the Indo-Pacific. Firstly, the vast majority of spawning observations appear to be in proximity to the lunar peak, with some possible indication that colonies spawning earlier in the day (hours after sunrise) are somewhat correlated with spawning earlier in the lunar cycle (Fig 5B). This of course is possibly skewed due to biases in observation time being higher between 3–7 days after the full moon. Across the biogeographic range (Fig 5A) there is a possible indication that, overall, colonies spawning later in the year are more likely to spawn earlier in the day (hours after sunrise). Of course, data remains lacking from the majority of locations from which Pavona corals are recorded and from many described species. Thus, the current dataset is far from a comprehensive review of the entirety of Pavona corals.

Broader insights into daytime spawning

It is at present unclear if the consistent daytime spawning of Pavona corals in the Gulf of Thailand represents a recent population-wide adaptation or a localised shift in spawning time when compared to the remainder of spawning corals and locations. A key assumption made in the present study and review of global Pavona spawning, is the taxonomic stability of species identified by authors across the globe. Variation in reproductive outputs becomes considerably more likely if multiple cryptic species are represented within a single pool of data incorrectly attributed to a single species [36]. At present however, a comprehensive taxonomic review of the genus Pavona is somewhat limited by the need for considerably more data on ecological, morphological, and molecular data across much of its range [39]. The most recent species to be described in the genus, P. giannii, is yet another example of a taxon that bears close morphological resemblance to P. varians, likely leading to it being overlooked as an independent taxon to date [39]. With multiple junior synonyms of P. varians (P. repens (Brüggemann, 1877), P. intermedia (Gardiner, 1898), and P. percarinata (Ridley, 1883)) being described from different locations, future application of integrated morphological and molecular techniques to this group may challenge the currently accepted diversity of taxa.

The identity of Pavona spp. (S4 Fig) in our dataset was determined based on skeletal morphology, agreeing with the brief and un-illustrated original descriptions [74] and subsequent literature [75], based on in-situ examination of skeletons in recently killed polyps), ecology (species-specific associations as in [76,77]) and, in the case of P. varians, a nearly identical description of gametes released during spawning between our data and that of Glynn et al. [36]. Spawning during the daytime raises significant questions regarding the adaptation of specific corals to changes in light availability in inducing spawning [21]. It is apparent from our dataset for example that the number of hours after sunrise is a more precise predictor for P. explanulata and P. varians, as opposed to hours before or after the full moon. This same aspect is notable in night-time spawning corals too which initiate spawning in general within a predictable window a given spawning night, regardless of month or year. There is no shortage of literature on importance of solar and lunar cycles in mediating night-time spawning time, and the varying roles of changes in light intensity and colour (wavelength) correlating with the onset of spawning timing [2,9,17,20,24–26,78,79]. If these mechanisms are to be considered as equally applicable to daytime spawning Pavona spp. in the Gulf, then this would imply a highly consistent ‘countdown’ several hours after dramatic change in illumination (sunrise, or perhaps sunset the previous day). Given the considerable variation of spawning time for Pavona spp. across the Indo-Pacific but high synchrony within populations, it is unlikely that this variation is a result of the major illumination-associated disruptor of synchrony, artificial light. Chronic exposure to artificial light, particularly white and blue light, have been shown to disturb gametogenic development in corals and disrupt synchrony [78,80–83]. The use of red-light is commonly applied to coral spawning monitoring in an effort to minimise disturbance of corals and other marine life, benefits of which are attributed to the shorter distance of attenuation of longer wavelengths [16,17,84]. In the present study, night-time SCUBA surveys were conducted using standard underwater torch light (white light) as opposed to using red-light filters, due to the already challenging prospect of spotting the near-transparent and potentially lethal individuals of Chironex sp. with white light, with red-light filters further reducing visibility and contrast considerably. Surveys using the ROV also used bright white light as surveys were restricted to in-built lighting with no modifications. While no studies have yet been conducted to assess the impact of short-term, close-proximity exposure to white light during spawning periods, the potential for a disruptive influence cannot be discarded. Nonetheless, patterns of spawning over the survey period appeared to be reinforced instead of being disrupted, and largely agree with earlier observations ([43], Appendix A).

Additionally, exposure of gametes to ultraviolet radiation of daylight spawners suggests possible photoprotective mechanisms that may differ considerably from conventionally night-time spawning corals. Little is known about the photoprotective mechanisms of gametes and larvae for agariciid corals, let alone differences between taxa based on spawning time or type, however insights may be gleaned from investigations into UV stress in non-agariciid taxa. Among the most informative may be the work of Aranda et al. [83], who reported differential UV tolerance in Orbicella faveolata (Ellis & Solander, 1786) larvae, with low sensitivity to UV radiation during the larval stage but high sensitivity in planulae. In contrast, Zhou et al. [85], found natural (overall) UV radiation had no noticeable effect on the survival, metamorphosis, or settlement on the larvae of Seriatopora caliendrum Hemprich & Ehrenberg, 1834. However, they did identify that that UV-A radiation alone did cause a decline in metamorphosis and settlement rates, which disappeared when UV-B was also included, which indicated a possible effect of UV-A on the duration of the planktonic phase of these larvae. Both these studies provide two examples of the seemingly widely differing and variable responses of coral larvae to different intensities and wavelengths of UV radiation among the few taxa that have been investigated [86–89], highlighting the need for more investigations on this topic.

In a recent investigation into the morphology of P. varians [90], the authors identified the species as one of a minority of taxa that may benefit from deeper water thermal refugia due to its light associated plasticity. It is possible therefore that this same plasticity and adaptation may extend to gametes and juveniles, allowing for a wide range of spawning timings. A possible explanation emerges, therefore, for the local adaptation to earlier daytime spawning in the Gulf of Thailand. Our surveys consistently found both P. varians and P. explanulata predominantly at greater depths of the reef slope, often as a ‘band’ of colonies along the edge of fringing reefs. Colonies were only found in shallower water when shaded by overhanging corals or rocks, or at high-turbidity sites. The Gulf of Thailand is well known to be a high-turbidity region due to the multiple major rivers exiting into the Gulf [91–94]. This may therefore drive localised ‘availability’ of daylight hours allowing for a broader window of spawning for these species, however, this is not seen in other high-turbidity regions, nor would it explain the noon-time spawning observed in our data (a period of peak illumination). This theory is also unsupported by the spawning patterns of the majority of other coral taxa in our dataset, which show minimal diverge from the spawning timings found elsewhere in the West-Pacific [33]. An alternative explanation may lie in the possible trade-off between illumination and thermal load, with Pavona corals spawning far later in the year than all other groups in the Gulf, and congeners across the globe. Water temperatures in the Gulf are typically coolest between October and January, thus spawning during the cooler months may reduce direct thermal-stress to gametes and larvae. Recent documentation from Ningaloo Reef, Western Australia [95] identified eutrophication and subsequent deoxygenation leading to widespread fish mortality associated with elevated temperature post-spawning and gametes/larvae being unable to disperse. Cooler temperatures may thus provide an ecological benefit to the survival of larvae in addition to a thermal benefit. However, this again does not explain the prevalence of daytime spawning during the cooler months, with night-time conditions obviously receiving no direct solar irradiation.

While daytime spawning of scleractinian corals is far from the norm, numerous (and growing) examples of sporadic or periodic, isolated or synchronised, daytime spawning cases are documented in the literature, often however as brief descriptions [27–30,44,96–98]. Among these are cases of Fungiidae corals releasing gametes ranging from a few hours after sunrise [29] to shortly before and during sunset [43,98], and widespread synchronised daytime spawning in Porites rus [21,27]. Our documentation represents the first time Cycloseris corals have been found to spawn in Thailand, and thus it is unknown how widespread daytime spawning may be for the genus. For Porites however, a closer examination is needed to ascertain the species identity of colonies at Sail Rock spawning during the daytime to investigate if asynchrony is due to taxonomic or some other localised variation. Given the remoteness of the single site from where it was recorded, Sail Rock, 15 km from the nearest coastal illumination, it is unlikely to be affected from coastal development-induced artificial lighting. However, abundant night-time squid fishing activities are carried out offshore the islands of the Samui archipelago (and the wider Gulf of Thailand) which utilise extremely bright lights visible from kilometres away. The possible impact of night-time illumination used in such fishing activities on coral spawning general reef ecology needs further study, and may disproportionately affect isolated pinnacle sites where offshore fishing is permitted.

Further investigation is needed for the geographical extent of daytime Pavona spawning. Questions also arise whether fertilisation and recruitment remain viable under UV exposure or if the daytime spawning nature of the population interrupts the population turnover of these corals. Presence and even growth in reef cover is not indicative of population turnover or recruitment [99], and thus abundances of mature colonies, recruits and larvae should be assessed independently for species-level population assessments. Glynn et al. [36] identified recruitment of Pavona corals to be limited by unusually high temperature anomalies, at thresholds of between 1.6 and 1.9 C°, but is unexpectedly recruitment was favoured by anomalies of 0.5 to 1.5 C°. In the case of the Gulf of Thailand, the spawning period of P. varians and P. explanulata being offset by at least five months has major implications for the dispersal of larvae, due to significant hydrodynamic changes at different times of the year [100–102]. Our study did not document fertilisation and dispersal kinetics, however P. varians larvae have been known to survive at least up to 10 days [36] allowing for a significant window of dispersal between source and sink sites. Indeed, little is known about the status or biogeography of Agariciid taxa across Thailand, with population changes being typically documented at highly localised scales and rarely with species-specific differentiation [103,104]. In a recent assessment on the efficacy of micro-fragmentation techniques for coral restoration [105], P. varians was found to have the slowest growth rate of the four taxa and significantly higher mortality than Diploastrea heliopora, Lithophyllon undulatum (but near-identical to P. decussata). The findings of regular synchronised spawning of P. varians documented in the present work provide a baseline for possible larval propagation or similar sexual-reproduction conservation efforts to be carried out in the Gulf of Thailand. Such techniques are increasingly being utilised [106,107] as an alternative to the widespread and often damaging (when unregulated) asexual fragmentation and propagation techniques, often freely used in Thailand and the wider tropics which dilute local genetic diversity and likely reduce fertilisation rates due to saturation of reef areas with fewer genotypes [108,109].

At present, coral spawning observations in the Gulf of Thailand rely on incidental observations or predictions from previous years, and do not provide the precision allowed by direct assessment of internal gametogenic development by sampling. Therefore, spawning timing and predictive potential for many genera remain unknown and may require more active examination. Alternatively, temporary relocation of medium sized colonies to ex-situ conditions may allow for assessment without the restrictions of microscopy, SCUBA, or safety concerns from envenomation. In a recent survey [66] utilising environmental DNA (eDNA), daytime spawning corals were reliably recorded in daytime water samples but not in night-time samples. Such an eDNA approach may therefore be a valuable tool in identifying the diversity and seasonality of day and night-time spawning corals across a wider geographic range.