Phd Dissertation

Reproduction Patterns of Scleractinian Corals in the Central Red Sea

Dissertation/Thesis by

Jessica Bouwmeester

In Partial Fulfillment of the Requirements For the Degree of

Doctor of Philosophy

King Abdullah University of Science and Technology Thuwal, Kingdom of Saudi Arabia

December 2013

EXAMINATION COMMITTEE APPROVALS FORM

The dissertation/thesis of Jessica Bouwmeester is approved by the examination committee.

Committee Chairperson: Dr. Michael Berumen

Committee Members: Dr. Christian Voolstra

Dr. Ibrahim Hoteit

Dr. Andrew Baird

Jessica Bouwmeester

ABSTRACT

Reproduction Patterns of Scleractinian Corals in the Central Red Sea

Jessica Bouwmeester

Early work on the reproductive seasonality of corals in the Red Sea suggested that corals exhibit temporal reproductive isolation, unlike on the Great Barrier Reef where many species spawn in synchrony. More recent work has however shown high synchrony in the maturity of gametes in Acropora species, suggesting multi-specific spawning is likely to occur in the Red Sea. In this thesis I investigate the patterns of coral reproduction in the central Red Sea. The spawning season in the central Red Sea lasts four months, from

April to July and spawning occurs on nights around the full moon. During this period

Acropora species show a peak of spawning in April, with some species spawning again in May. The level of synchrony, quantified with a spawning synchrony index, is comparable to other locations where multi-specific spawning has been reported.

Observations over two consecutive years show that the synchrony of spawning was lower in spring 2012 than in spring 2011, and thus that spawning patterns are variable from one year to the other. Coral settlement patterns on artificial substrata confirmed a main spawning season in the spring but also supported reproductive data suggesting that some

Porites spawn in October-November. Settlement was studied over 2.5 years on a reef, which had suffered recently from high mortality after a local bleaching event. Settlement appeared low but post-bleaching studies from other locations indicated similar abundances and showed that recruits generally did not increase until 5 years after the 5 bleaching event. Abundance of juvenile corals however started to increase significantly three years after the bleaching. Successful recruitment, although low suggests that the coral assemblage on the affected reef will most likely recover as long as it is not affected by another disturbance.

ACKNOWLEDGEMENTS

I would like to thank my advisors Dr. Michael Berumen and Dr. Andrew Baird who have provided strong support throughout my PhD education and during my thesis writing, as well as my other committee members Dr. Christian Voolstra and Dr. Ibrahim Hoteit. I particularly thank Michael for the numerous opportunities that I have been given in the past years as a member of the Reef Ecology Lab in the King Abdullah University of

Science and Technology. I am also very grateful to the Reef Ecology Lab and of course

KAUST for strong financial support throughout my thesis.

I thank all the numerous divers from the Reef Ecology Lab, Red Sea Research Centre,

KAUST community and the Coastal Marine Resources Core Lab for the time invested in supporting me in the field mostly underwater, specifically Maha Khalil, Pedro De La

Torre, Francis Mallon and David Pallett. I would also like to thank the logistics team of

CMOR for their support with organising a big part of the fieldwork conducted for this thesis.

I am fortunate to have met many amazing people from all over the world during my PhD years in KAUST and am especially grateful to all my freediving buddies as well as the music community on campus and outside, who strongly contributed to making my time here a rich and unique multicultural experience.

Last but not least I would like to thank my family for their strong support in all my projects since I left home nearly a decade ago and I would like to specially thank my brother Daniël for his engineering knowledge and expertise (and for fixing my matlab files), and for having always been there for me. Merci Dan! 7

TABLE OF CONTENTS

EXAMINATION COMMITTEE APPROVALS FORM…………………………………… 2

ABSTRACT………………………………………………………………………………… 4

ACKNOWLEDGEMENTS………………………………………………………………... 6

TABLE OF CONTENTS…………………………………………………………………... 7

LIST OF ABBREVIATIONS……………………………………………………………… 8

LIST OF FIGURES………..………………………………………………………………. 9

LIST OF TABLES…………………………………………………………………………. 11

CHAPTER 1: INTRODUCTION………………………………...…………………………. 12

CHAPTER 2: REPRODUCTIVE SEASONALITY OF THE ACROPORA ASSEMBLAGE 32

SUPPLEMENTARY DATA……………………………………………………………….... 55

CHAPTER 3: MULTISPECIFIC SPAWNING SYNCHRONY …………………………… 59

SUPPLEMENTARY DATA……………………………………………………………….... 98

CHAPTER 4: POST-BLEACHING CORAL RECOVERY AND SETTLEMENT.……. 103

CHAPTER 5: GENERAL DISCUSSION……………………….………………………… 125

APPENDIX 1: LIST OF PUBLICATIONS……………………………………………. 136

BOUWMEESTER ET AL. 2011 CORAL REEFS…………………………………………. 137

BOUWMEESTER ET AL. 2011 GALAXEA…..…………………………………………. 138

RODER ET AL. 2013 NATURE SCIENTIFIC REPORTS………………………………. 142

RODER ET AL. 2013 SUPPLEMENTARY DATA……………………..……..…………. 152

BERUMEN ET AL. 2013 CORAL REEFS……………………………………………...…. 155

FURBY ET AL. 2013 CORAL REEFS……………………………………………………. 167

APPENDIX 2: LIST OF CONFERENCE PRESENTATIONS……………………………. 176

LIST OF ABBREVIATIONS

ANOVA Analyse of variance

CCA Crustose coralline algae

FM Full moon

GBR Great Barrier Reef

KAUST King Abdullah University of Science and Technology

NM New Moon

SM (A) Marquis’ index of synchrony at the assemblage level

SM (S) Marquis’ index of synchrony at the species level

SST Sea surface temperatures

LIST OF FIGURES

CHAPTER 1: INTRODUCTION…………………………………………………………….... 12

Figure 1 Map of the Red Sea

CHAPTER 2: REPRODUCTIVE SEASONALITY OF THE ACROPORA ASSEMBLAGE IN

THE CENTRAL RED SEA……………………….…………………………………………. 32

Figure 1 Map of the Red Sea showing Thuwal and Al Lith

Figure 2 Proportion of Acropora colonies with mature, immature or no visible eggs and proportion of Acropora species with at least one mature

Figure 3 Spawning synchrony calculated with the Marquis synchrony index SM

CHAPTER 3: MULTISPECIFIC SPAWNING SYNCHRONY WITHIN SCLERACTINIAN

CORAL ASSEMBLAGES IN THE RED SEA……………………………………………… 59

Figure 1 Map of the Red Sea showing Al Fahal and Dreams Beach Reef

Figure 2 Moon phases in Thuwal in 2011, 2012 and 2013

Figure 3 Stages of gametogenesis of some scleractinian corals throughout the reproductive season from April to July

Figure 4 Stages of gametogenesis of Porites species

Figure 5 Setting and spawning behaviour

Figure 6 Number of species observed to spawn in various locations

Figure 7 Hourly water temperatures from April to July in 2011, 2012 and 2013

Figure S1 Porites columnaris

Figure S2 Porites lobata

Figure S3 Porites lutea

Figure S4 Porites solida 10

CHAPTER 4: CORAL RECOVERY AND RECRUITMENT AFTER A BLEACHING EVENT

IN THE CENTRAL RED SEA……………………………………………………………… 103

Figure 1 Maps indicating Thuwal and Abu Shosha

Figure 2 Relative abundance of coral genera in the adult assemblage and relative distribution of each genus on the sheltered versus the exposed side

Figure 3 Average coral on the exposed and sheltered side of Abu Shosha.

Figure 4 Exposed side of Abu Shosha Reef at 1-5m

Figure 5 Relative abundance of the abundant genera of juvenile corals

Figure 6 Average number of juveniles

Figure 7 Average number of recruits per 100 cm2 tile and proportion of recruits from

Acroporidae, Pocilloporidae, Poritidae and other families

LIST OF TABLES

CHAPTER 2: REPRODUCTIVE SEASONALITY OF THE ACROPORA ASSEMBLAGE IN

THE CENTRAL RED SEA……………………………………………………………………. 32

Table S1 Proportion of Acropora colonies with white and mature eggs

Table S2 Proportion of Acropora colonies in each species with white and mature eggs

CHAPTER 3: MULTISPECIFIC SPAWNING SYNCHRONY WITHIN SCLERACTINIAN

CORAL ASSEMBLAGES IN THE RED SEA…………...…………………………………… 59

Table 1 Criteria for classification of oocytes and spermatocytes into developmental stages

Table 2 Stages of gametogenesis on various dates throughout the spawning season in 2011.

Table 3 List of corals recorded to spawn in 2011, 2012 and 2013.

CHAPTER 1: INTRODUCTION

Introduction

Coral reefs are one of the most spectacular and fragile of underwater environments, covering less than one percent of the ocean floor but supporting an estimated 25 percent of all marine life. They are one of the most diverse and productive ecosystems in the world and provide many important ecosystem services such as coastal protection, an important source of food for millions of people, a source of income from recreational activities and a source of potential medicine to cure human diseases (Connell 1978;

Moberg and Folke 1999). They are formed primarily by hermatypic scleractinian corals; commonly known as reef-building corals. Unfortunately they are considered one of the most highly impacted marine ecosystems on earth. Following the 1997-1998 ENSO event, corals all over the world suffered from intense bleaching, followed by high mortality rates in many regions (Wilkinson 1999; Baker et al. 2008). Such events are likely to happen again and the predicted consequences on coral reefs under likely climate change scenarios are extreme (Hoegh-Guldberg 1999; Hoegh-Guldberg et al. 2007;

Halpern et al. 2008). Under the current trends of climate change and the rise in disturbances to corals leading to mortality (Graham et al. 2011), successful reproduction of the adult assemblage followed by successful settlement of coral larvae on the reef and successful post-settlement survival are key factors to ensure the resilience of coral reefs on a long-term scale.

Coral Reproduction

Reproductive Traits

Scleractinian corals exhibit a range of reproductive strategies, including both sexual and asexual reproduction. Most corals grow colonies using asexual budding of individual coral polyps, and in most colonies, the budded polyps remain interconnected via the coenenchyme, allowing for some muscular and nervous communication between polyps

(Harrison 2011). Branching species such as Acropora species and some massive corals also use asexual fragmentation for colony dispersal (Highsmith 1982; Lirman 2000;

Foster et al. 2007). However sexual reproduction is the primary means of recolonization for most species. It increases genetic diversity, which positive selection can act upon, and is thus a more effective strategy (Connell 1997; Ritson-Williams et al. 2009). Two modes of sexual reproduction exist in scleractinian corals, species either brood planulae (internal fertilisation) or spawn gametes (external fertilisation) (Baird et al. 2009). Two types of sexuality can also be observed in corals, being either hermaphroditic with both sexes developed in each polyp, or gonochoric with separate sexes in each colony (Baird et al.

2009; Harrison 2011; but see Guest et al. 2012 for exceptions). Different combinations of sexuality and reproduction modes are possible but the most common system displayed is hermaphroditic spawning with the pattern found in 63 % of species examined (Baird et al. 2009).

Pelagic Phase of Coral Larvae

Brooding coral species release relatively mature planulae, which are able to initiate attachment, metamorphosis and settlement shortly after being released from their parental 14 polyps, with a competency window in the order of minutes to days (Richmond 1987;

Harrison and Wallace 1990; Carlon and Olson 1993; Sakai 1997). Spawning corals produce male and/or female gametes that are synchronously released into the water column in massive numbers, enabling them to distribute their offspring over a broad geographic area. The gametes are usually positively buoyant egg/sperm bundles, which will then float to the sea surface where the bundles will break open, releasing eggs and sperm, to allow cross-fertilization (Arai et al. 1993). Once successfully fertilized, the eggs will develop into the planulae stage. An individual planula then stays in the water column until it finds a suitable surface to which it can attach. Larvae from broadcast spawning corals generally require about 2 to 6 days of development after fertilisation during which they drift in the water column, before they become competent to settle

(Babcock and Heyward 1986; Szmant 1986). Some individuals can choose to delay the settlement phase until up to 130 days after spawning in case they don’t find any suitable substrate to settle on, the duration depending on the species (Richmond 1987; Jones et al.

2009). At this point larvae have developed cilia for motility as well as sensory and secretory cells at their aboral end to actively examine and subsequently adhere to a suitable surface (Graham et al. 2008; Connolly and Baird 2010; Figueiredo et al. 2012)

1999). However, not all larvae are active swimmers. Some larvae have the potential to disperse in plankton for an extensive period but others move along the reef with the ability to crawl rather than to swim (Gleason and Hofmann 2011). Coral larvae respond to different environmental cues such as light, depth and chemical cues while in their pelagic phase in order to disperse to a suitable reef to settle upon (Ritson-Williams et al.

2009). Pelagic coral larvae also respond to acoustic cues, mainly reef sounds produced by 15 fish and crustaceans, which may facilitate the detection of a suitable habitat from large distances (Vermeij et al. 2010).

From Pelagic to Sessile: Larval Settlement

The choice of an appropriate substrate and habitat for settlement of coral larvae is crucial for survival, corals being sessile organisms. Chemical signals associated with the reef are thought to provide important settlement cues for coral larvae which after location of an appropriate substrate must undergo a complex metamorphosis, founding a juvenile coral colony (Grasso et al. 2011). Once available substrate has been found, the motile larvae settle and subsequently undergo metamorphosis into morphologically different coral recruits, and start their sessile life (Ritson-Williams et al. 2009).

Cues for Synchronous Spawning

Sexual reproductive processes in corals appear to be strongly influenced by environmental factors, which regulate and synchronize reproductive cycles and which exert evolutionary selective pressure through enhanced reproductive success. There are proximate and ultimate controls that affect the timing of spawning events and regulate the time of year, the night of spawning and the time of spawning (Baird et al. 2009). The influence of sea surface temperatures is believed to be a strong cue in synchronising gametogenesis and maturation of gametes. For example on the Great Barrier Reef, spawning happens during warming sea surface temperatures but species on inshore reefs spawn a month earlier than species on offshore reef. In this particular case, no other environmental factors apart from temperatures appear to differ between the two sets of 16 reefs (Willis et al. 1985; Babcock et al. 1986; Baird et al. 2009). In other regions, different cues appear to regulate the time of spawning. Corals in Western Australia, although on the same latitudes as the Great Barrier Reef, spawn in a different season

(Simpson 1985). Corals in the western Atlantic time their spawning according to the average temperature and the change of insolation (van Woesik et al. 2006). Other cues are known to be responsible to determine the day of the month and the time of spawning.

Coral adjust their spawning to follow lunar patterns although environmental cues seem to interact with the lunar patterns (Baird et al. 2009).

The final cue before release of gametes is usually the sunset and most species spawn in the night, in the first hours following the sunset, although a small number of corals spawn during the daytime (Baird et al. 2009). Night-time spawning may be a consequence of predator avoidance from organisms which rely on vision for feeding, and of the avoidance from the potentially damaging UV radiation in the first hours of larval development (Babcock et al. 1986; Jeffery 1990; Guest et al. 2005). Broadcast spawning coral species may spawn on only one or a few nights each year, and though different species may spawn at different times, the spawning events for any given species happen at the same time.

Multi-specific spawning

Gametogenesis cycles in spawning corals range from 6 to 14 months, and thus each individual colony spawns only once a year with a few exceptions (Baird et al. 2009;

Rosser et al. 2013). Because fertilization success is highly dependent on sperm densities, 17 it is likely to be advantageous for individuals within a population to synchronise spawning (Babcock et al. 1986; Levitan et al. 2004; Baird et al. 2009). Subsequently, many broadcasting spawning species spawn at predictable times each year (Babcock et al. 1986).

In most coral assemblages, spawning is not only synchronous within species but also between species, leading to multi-specific synchronous spawning events. Multi-specific spawning was observed for the first time in 1981 on the Great Barrier Reef, where 30 species released their gametes in a time window of a few hours (Harrison et al. 1984).

Further observations on the Great Barrier Reef have led to the observation of 105 species spawning in the weeks following the full moon of October and November on reefs within a 500 km range (Babcock et al. 1986).

This reproductive pattern has since then been described from coral reefs in 23 locations globally (Baird et al. 2009). The cues driving the level of synchrony are however poorly understood. It has been suggested that synchrony was more likely to happen in regions with large fluctuations in environmental factors such as tides, sea surface temperatures or insolation (Oliver et al. 1988; Harrison and Wallace 1990). However, synchronous spawning is now known to occur in all specious coral assemblages, including those in places, such as Singapore, which do not have large fluctuations in environmental variables (Guest et al. 2005). The spawning season is likely to occur during months in which the environmental conditions are optimal for gametogenesis and/or fertilisation, when planktonic food is abundant and during periods of calm weather and low-amplitude 18 tides to avoid rapid gamete dilution (Babcock et al. 1986; Watson et al. 2003; Guest et al.

2005; van Woesik 2010).

The exact advantages of multi-specific spawning are not clear (Guest et al. 2005).

Species participating in multi-specific spawning events may benefit from predator satiation on the night of spawning. Alternatively, species which spawn together may independently respond to the same cues used by each species to synchronise spawning

(Guest et al. 2005).

Coral Reproduction in the Context of Climate Change

Coral reef ecosystems are particularly sensitive to changes in the physical environment such as the ones caused by climate change (Baker et al. 2008). One of the first consequences is periods of unusually high sea surface temperatures, causing coral bleaching at a frequency and magnitude which has increased over the past decades

(Hoegh-Guldberg 1999; Knowlton 2001; Baker et al. 2008). Climate change has an important impact on coral reproduction by causing changes in sea surface temperatures.

Consequences include: disruption of coral reproductive cycles, increase in larval mortality and inhibition of larval settlement as well as post-settlement survival. This can be caused by increasing carbon dioxide absorption which leads to a decrease in pH, and by increasing aragonite saturation which is likely to interfere with the calcification process in coral recruits as well as adult coral colonies (Kleypas et al. 1999; Albright et al. 2008; Jokiel et al. 2008; Munday et al. 2009; Albright and Langdon 2011).

During periods of unusually warm sea surface temperatures, corals expel most of their pigmented symbiotic algae or zooxanthellae and subsequently appear white. As a result of the loss of transfer of the photosynthetic products from the symbionts, tissue growth, regeneration and calcification are compromised during bleaching events (Baker et al.

2008; McClanahan et al. 2009). They are also more susceptible to disease following bleaching (Lesser 2004). During or following bleaching events, corals may suffer from whole colony mortality, partial mortality or they may recover fully (Baker et al. 2008).

The susceptibility to bleaching and recovery from bleaching by acquiring new symbionts appears however to be species-specific although it also varies between regions and depending on whether or not the coral has been exposed to bleaching in the past

(Marshall and Baird 2000; Brown et al. 2002; Baker et al. 2004; McClanahan et al. 2009;

Furby et al. 2013 / see annexe 1, p.167).

The pelagic phase of coral larvae offers the potential to replenish coral communities locally or travel longer distances for further dispersal and thus result in higher connectivity between populations (Ritson-Williams et al. 2009). Although a large competency window suggests further dispersal potential, most larvae seem to settle close to their parents (Sammarco and Andrews 1989; Cowen et al. 2006; Steneck 2006).

Recruitment rates must equal or exceed rates of adult mortality to sustain a local population. Consequently, although longer distance dispersal is important for gene flow, the lower density of planulae is not sufficient to sustain a population.

The reproductive output of corals that recover from bleaching events may be affected for several years after the bleaching (McClanahan et al. 2009). For example, some colonies of Montastrea annularis that suffered the previous year from prolonged bleaching lack gametes in the next reproductive cycle (Szmant and Gassman 1990), and coral fecundity can also be reduced following bleaching (Ward et al. 2002; Baird and Marshall 2002), as can egg quality (Michalek-Wagner and Willis 2001). Additionally, due to the low density of gametes during spawning events, fertilisation could be overall compromised

(Knowlton 2001).

The Red Sea

Generalities

The Red Sea is situated in a deep, narrow and elongated rift valley between Africa and the Arabian peninsula, is approximately 2000 km long and on average less than 300 km wide (Fig. 1). The Red Sea is considered an extreme environment with very little precipitation, high evaporative rates, salinity values averaging 38ppt and up to 42 ppt in its north in the Gulf of Aqaba, little exchange of water with other seas, and with temperatures regularly exceeding 32 degrees in the summer (Davis et al. 2011; Berman et al. 2003). The Red Sea is a rich and biodiverse ecosystem with high endemism (Stehli and Wells 1971; Ormond and Edwards 1987; Hughes et al. 2002). However, compared to other locations worldwide, very little research on marine ecology has been conducted in the Red Sea. Furthermore, most of the work on coral reproductive biology has been conducted in Eilat (Fig. 1), on a 6 km coast in the north of the Gulf of Aqaba (Berumen et al. 2013 / see annexe 1, p. 155). 21

Figure 1: Map of the Red Sea. Most of the research on reproductive biology of corals has been conducted in Eilat, in the northern Red Sea. Landsat image: Google, ORION-ME

Gametogenesis, Seasonality and Time of Spawning

Most early work on coral reproduction was focused on brooding corals, believed at the time to be the main mode of reproduction of scleractinian corals until the early 1980s, when the first multi-specific spawning events were described on the Great Barrier Reef.

(Harrison 2011). Rinkevich and Loya (1979a, 1979b, 1987) described the gametogenesis of one of the most abundant species in Eilat, Stylophora pistillata, a brooding coral, over 22 a period of ten years, as well as the reproductive seasonality of the species over a period of 5 years. They described a 6-9 months reproductive season, and a highly variable reproductive output from one year to the next. Fadlallah (1988) then studied the gametogenesis of Stylophora pistillata in Yanbu in the central Red Sea. He described a continuous annual cycle with two periods of oogenesis and two extended periods of embryogenesis from March to June and from September to November. He concluded that the gametogenesis and the planulation periods in Yanbu were shorter than in Eilat.

A small number of spawning species were also investigated in the Red Sea. Fadlallah

(1985) studied the gametogenesis of Pocillopora verrucosa in Yanbu, central Red Sea, and described a short annual cycle of gametogenesis in spring, pinpointed the timing of spawning on around the new moon in the end of May. Kramarsky-Winter and Loya

(1998) later studied the reproductive strategies of two Fungia species in Eilat. They describe a reproductive season from June to September for F. scutaria with night-time spawning following the full moon and a reproductive season from July to August for F. granulosa with daytime spawning uncorrelated with any lunar phase.

Synchrony of Spawning in Red Sea Coral Assemblages

Spawning in the Red Sea was thought to be asynchronous for the past decades (Harrison

1990; Richmond and Hunter 1990) with different species spawning in different seasons, on different months, and at different stages of the lunar cycle, following studies conducted in Eilat in the Gulf of Aqaba (Shlesinger and Loya 1985; Shlesinger et al.

1998). Shlesinger and Loya (1985) studied spawning patterns of 13 ecologically important species in Eilat and concluded in temporal reproductive isolation of coral 23 species in the northern Red Sea. However, three of the 13 species were brooders and released planulae throughout the year, from January to December. Brooding corals initially release sperm that will internally fertilise the eggs from other colonies in the population. Larval development is then internal, unlike in spawning corals, which release male and female gametes synchronously in the water column for external fertilisation

(Harrison and Wallace 1990). Brooders are also known for having extended planulation periods and it is not unusual for them to spawn all year round (Harrison 2011). The ten other species in Eilat, all broadcast spawners, released gametes from May to August in

1981 and from June to September in 1982. Seven out those ten species released gametes in the week following the full moon, two species spawned around the new moon and one species released gametes the week preceding the full moon. However, the authors concluded that spawning occurred in different seasons, in different months or different lunar phases with the same month. Shlesinger et al. (1998) then studied the reproductive patterns of 24 species in Eilat, which include the 13 from the previous study. They described the location of gonads, fecundity and oocyte size and colour. They also studied spawning patterns of the 11 species additional to the ones already described in the earlier study. Nine of the studied species spawned between June and September in 1986 and the other two species spawned in April.

The authors of the above studies concluded that corals in the northern Red Sea exhibit temporal reproductive isolation and that there is no overlap in time between species.

However, they based their conclusions on results from spawning as well as brooding species. When only spawning corals are considered, reproduction in Eilat shows strong 24 seasonality and lasts 3-4 months during the summer (Guest et al. 2005). The six Acropora species studied in Eilat actually all spawned within two months. There are 47 Acropora species in the Red Sea according to Veron (2000) or 42 according to Wallace (1999), and at least 12 Acropora species have been recorded in the Gulf of Aqaba (Scheer and Pillai

1983). It has been suggested that in speciose assemblages there is often considerable overlap of spawning periods among species, resulting in multi-specific spawning events

(Guest et al. 2005). Reproductive work on more Acropora species may have indicated higher synchrony within the genus.

Years later, Hanafy et al. (2010) surveyed 23 Acropora species on reefs off the Egyptian cities of Hurghada and Marsa Alam in the northern Red Sea (Fig. 1), and in contradiction with reproductive work and spawning observations from Eilat, they found a high synchrony in the stage of maturity of gametes, suggesting that multi-specific spawning was likely to be a regular feature of Red Sea coral assemblages. Spawning was inferred to happen in April and/or May, however, no spawning was observed to confirm the assumption.

Dynamics of Coral Recruitment and Juvenile Corals

Glassom et al. (2004) studied the dynamics of coral recruitment on artificial substrata in

Eilat. Coral recruits were dominated by pocilloporids and numbers per 100 cm2 plate were low (1.9) relative to other locations in the world. Rates of recruitment were not consistent of two consecutive years. Glassom and Chadwick (2006) then studied the dynamics of juvenile corals on natural substrata. They observed high numbers of 25 juveniles 42 to 173 juveniles per m2 compared to other locations worldwide. Numbers of juvenile were different between sites of different morphology and depending on whether the available substratum was calcium-carbonate or metamorphic rock. Growth rates were variable depending on the season, and rates of mortality in different sites were variable in some genera.

Objectives of this PhD thesis

The aim of this thesis was to confirm and extend knowledge on the reproductive biology in the Red Sea and situate reproductive patterns compared to other regions in the world.

The specific subjects considered were the reproduction of broadcast spawning corals and the recruitment and juvenile survival in the context of recovery from a bleaching event

The aim of Chapter 2 was to determine the spawning period(s) of scleractinian corals in the central Red Sea and the monthly level of synchrony, based on reproductive data from the Acropora assemblage. A flowering synchrony index from plant biology was suggested to quantify spawning synchrony, and was tested in the central Red Sea and in five additional locations in the Indo-Pacific.

The aim of Chapter 3 was to pinpoint the timing of spawning of corals species from other families and collect in situ spawning observations of a wide range of species from different families.

A bleaching event occurred in Thuwal in the summer 2010 during which inshore reefs were heavily affected resulting in high mortality of the corals. The aim of Chapter 4 was to monitor the recovery processes throughout the first three years after the bleaching, focusing on coral recruitment and post-settlement survival of juvenile corals.

References Albright R, Mason B, Langdon C (2008) Effect of aragonite saturation state on settlement and post-settlement growth of Porites astreoides larvae. Coral Reefs 27:485-490 Albright R, Langdon C (2011) Ocean acidification impacts multiple early life history processes of the Caribbean coral Porites astreoides. Global Change Biology 17:2478-2487 Arai I, Kato M, Heyward AJ, Ikeda Y, Iizuka T, Maruyama T (1993) Lipid composition of positively buoyant eggs of reef building corals. Coral Reefs 12:71-75 Babcock RC, Bull GD, Harrison PL, Heyward AJ, Oliver JK, Wallace CC, Willis BL (1986) Synchronous spawnings of 105 scleractinian coral species on the Great Barrier Reef. Mar Biol 90:379-394 Baker AC, Starger CJ, McClanahan TR, Glynn PW (2004) Coral reefs: Corals' adaptive response to climate change. Nature 430:741-741 Baker AC, Glynn PW, Riegl B (2008) Climate change and coral reef bleaching: An ecological assessment of long-term impacts, recovery trends and future outlook. Estuarine, Coastal and Shelf Science 80:435-471 Baird AH, Marshall PA (2002) Mortality, growth and reproduction in scleractinian corals following bleaching on the Great Barrier Reef. Mar Ecol Prog Ser 237:133-141 Baird AH, Guest JR, Willis BL (2009) Systematic and Biogeographical Patterns in the Reproductive Biology of Scleractinian Corals. Annual Review of Ecology Evolution and Systematics 40:551-571 Berman T, Paldor N, Brenner S (2003) Annual SST cycle in the Eastern Mediterranean, Red Sea and Gulf of Elat. Geophysical Research Letters 30:1261 Berumen ML, Hoey AS, Bass WH, Bouwmeester J, Catania D, Cochran JEM, Khalil MT, Miyake S, Mughal MR, Spaet JLY, Saenz-Agudelo P (2013) The status of coral reef ecology research in the Red Sea. Coral reefs:1-12 27

Brown B, Dunne R, Goodson M, Douglas A (2002) Experience shapes the susceptibility of a reef coral to bleaching. Coral reefs 21:119-126 Carlon DB, Olson RR (1993) Larval dispersal distance as an explanation for adult spatial pattern in two Caribbean reef corals. Journal of Experimental Marine Biology and Ecology 173:247-263 Connell JH (1978) Diversity in Tropical Rain Forests and Coral Reefs. Science 199:1302-1310 Connell JH (1997) Disturbance and recovery of coral assemblages. Coral reefs 16:S101-S113 Connolly SR, Baird AH (2010) Estimating dispersal potential for marine larvae: dynamic models applied to scleractinian corals. Ecology 91:3572-3583 Cowen RK, Paris CB, Srinivasan A (2006) Scaling of connectivity in marine populations. Science 311:522 Davis KA, Lentz SJ, Pineda J, Farrar JT, Starczak VR, Churchill JH (2011) Observations of the thermal environment on Red Sea platform reefs: a heat budget analysis. Coral reefs 30:25-36 Fadlallah Y (1985) Reproduction in the coral Pocillopora verrucosa on the reefs adjacent to the industrial city of Yanbu (Red Sea, Saudi Arabia). Fifth International Coral Reef Congress 4:313-318 Fadlallah YH, Lindo RT (1988) Contrasting cycles of reproduction in Stylophora pistillata from the Red Sea and the Arabian Gulf, with emphasis on temperature. 6th International Coral Reef Symposium 3:225-230 Figueiredo J, Baird AH, Connolly SR (2012) Synthesizing larval competence dynamics and reef- scale retention reveals a high potential for self-recruitment in corals. Ecology 94:650-659 Foster NL, Baums IB, Mumby PJ (2007) Sexual vs. asexual reproduction in an ecosystem engineer: the massive coral Montastraea annularis. Journal of Animal Ecology 76:384- 391 Furby KA, Bouwmeester J, Berumen ML (2013) Susceptibility of central Red Sea corals during a major bleaching event. Coral reefs 32:505-513 Glassom D, Chadwick N (2006) Recruitment, growth and mortality of juvenile corals at Eilat, northern Red Sea. Marine Ecology Progress Series 318:111-122 Glassom D, Zakai D, Chadwick-Furman NE (2004) Coral recruitment: a spatio-temporal analysis along the coastline of Eilat, northern Red Sea. Mar Biol 144:641-651 Gleason DF, Hofmann DK (2011) Coral larvae: From gametes to recruits. Journal of Experimental Marine Biology and Ecology 408:42-57 28

Graham EM, Baird AH, Connolly SR (2008) Survival dynamics of scleractinian coral larvae and implications for dispersal. Coral Reefs 27:529-539 Graham NAJ, Nash KL, Kool JT (2011) Coral reef recovery dynamics in a changing world. Coral reefs 30:283-294 Grasso LC, Negri AP, Fôret S, Saint R, Hayward DC, Miller DJ, Ball EE (2011) The biology of coral metamorphosis: Molecular responses of larvae to inducers of settlement and metamorphosis. Developmental Biology 353:411-419 Guest JR, Baird AH, Goh BPL, Chou LM (2005) Seasonal reproduction in equatorial reef corals. Invertebrate Reproduction & Development 48:207-218 Guest JR, Baird AH, Goh BPL, Chou LM (2012) Sexual systems in scleractinian corals: an unusual pattern in the reef-building species Diploastrea heliopora. Coral reefs 31:705-713 Halpern BS, Walbridge S, Selkoe KA, Kappel CV, Micheli F, D'Agrosa C, Bruno JF, Casey KS, Ebert C, Fox HE, Fujita R, Heinemann D, Lenihan HS, Madin EMP, Perry MT, Selig ER, Spalding M, Steneck R, Watson R (2008) A Global Map of Human Impact on Marine Ecosystems. Science 319:948-952 Hanafy MH, Aamer MA, Habib M, Rouphael AB, Baird AH (2010) Synchronous reproduction of corals in the Red Sea. Coral Reefs 29:119-124 Harrison PL (2011) Sexual Reproduction of Scleractinian Corals. In: Dubinsky Z, Stambler N (eds) Coral Reefs: An Ecosystem in Transition. Springer Netherlands, pp59-85 Harrison PL, Wallace CC (1990) Reproduction, dispersal and recruitment of scleractinian corals. Dubinsky Z (ed) Ecosystems of the world: coral reefs. Elsevier, Amsterdam, pp133–207 Harrison PL, Babcock RC, Bull GD, Oliver JK, Wallace CC, Willis BL (1984) Mass Spawning in Tropical Reef Corals. Science 223:1186-1189 Highsmith RC (1982) Reproduction by fragmentation in corals. Marine Ecology Progress Series 7:207-226 Hoegh-Guldberg O (1999) Climate change, coral bleaching and the future of the world's coral reefs. Marine and Freshwater Research 50:839-866 Hoegh-Guldberg O, Mumby PJ, Hooten AJ, Steneck RS, Greenfield P, Gomez E, Harvell CD, Sale PF, Edwards AJ, Caldeira K, Knowlton N, Eakin CM, Iglesias-Prieto R, Muthiga N, Bradbury RH, Dubi A, Hatziolos ME (2007) Coral Reefs Under Rapid Climate Change and Ocean Acidification. Science 318:1737-1742 Hughes TP, Bellwood DR, Connolly SR (2002) Biodiversity hotspots, centres of endemicity, and the conservation of coral reefs. Ecology Letters 5:775-784 29

Jeffery WR (1990) Ultraviolet irradiation during ooplasmic segregation prevents gastrulation, sensory cell induction, and axis formation in the ascidian embryo. Developmental Biology 140:388-400 Jokiel PL, Rodgers KS, Kuffner IB, Andersson AJ, Cox EF, Mackenzie FT (2008) Ocean acidification and calcifying reef organisms: a mesocosm investigation. Coral Reefs 27:473-483 Jones GP, Almany GR, Russ GR, Sale PF, Steneck RS, van Oppen MJH, Willis BL (2009) Larval retention and connectivity among populations of corals and reef fishes: history, advances and challenges. Coral Reefs 28:307-325 Kleypas JA, Buddemeier RW, Archer D, Gattuso J-P, Langdon C, Opdyke BN (1999) Geochemical Consequences of Increased Atmospheric Carbon Dioxide on Coral Reefs. Science 284:118-120 Knowlton N (2001) The future of coral reefs. Proceedings of the National Academy of Sciences of the United States of America 98:5419-5425 Kramarsky-Winter E, Loya Y (1998) Reproductive strategies of two fungiid corals from the northern Red Sea: environmental constraints? Marine Ecology Progress Series 174:175- 182 Lesser MP (2004) Experimental biology of coral reef ecosystems. Journal of Experimental Marine Biology and Ecology 300:217-252 Levitan DR, Fukami H, Jara J, Kline D, McGovern TM, McGhee KE, Swanson CA, Knowlton N (2004) Mechanisms of reproductive isolation among sympatric broadcast spawning corals of the Montastraea annularis species complex. Evolution 58:308-323 Lirman D (2000) Fragmentation in the branching coral Acropora palmata (Lamarck): growth, survivorship, and reproduction of colonies and fragments. Journal of Experimental Marine Biology and Ecology 251:41-57 Marshall PA, Baird AH (2000) Bleaching of corals on the Great Barrier Reef: differential susceptibilities among taxa. Coral reefs 19:155-163 McClanahan TR, Weil E, Cortés J, Baird AH, Ateweberhan M (2009) Consequences of Coral Bleaching for Sessile Reef Organisms. In: Oppen MH, Lough J (eds) Coral Bleaching. Springer Berlin Heidelberg, pp121-138 Michalek-Wagner K, Willis BL (2001) Impacts of bleaching on the soft coral Lobophytum compactum. II. Biochemical changes in adults and their eggs. Coral reefs 19:240-246 Moberg F, Folke C (1999) Ecological goods and services of coral reef ecosystems. Ecological economics 29:215-233 30

Munday PL, Leis JM, Lough JM, Paris CB, Kingsford MJ, Berumen ML, Lambrechts J (2009) Climate change and coral reef connectivity. Coral Reefs 28:379-395 Oliver J, Babcock R, Harrison P, Willis B (1988) Geographic extent of mass coral spawning: clues to ultimate causal factors. Proc 6th int coral Reef Symp 2:803-810 Omori M, Fukami H, Kobinata H, Hatta M (2001) Significant drop of fertilization of Acropora corals in 1999: An after-effect of heavy coral bleaching? Limnol Oceanogr 46:704-706 Ormond R, Edwards A (1987) Red Sea fishes. In: Edwards AJ, Head SM (eds) Red Sea, pp251- 287 Richmond RH (1987) Energetics, competency, and long-distance dispersal of planula larvae of the coral Pocillopora damicornis. Mar Biol 93:527-533 Richmond RH, Hunter CL (1990) Reproduction and recruitment of corals: Comparisons among the Caribbean, the Tropical Pacific, and the Red Sea. Marine ecology progress series 60:185-203 Rinkevich B, Loya Y (1979) The reproduction of the Red Sea coral Stylophora pistillata. I. Gonads and planulae. Mar Ecol Prog Ser 1:133-144 Rinkevich B, Loya Y (1979) The reproduction of the Red Sea coral Stylophora pistillata. II. Synchronization in breeding and seasonality of planulae shedding. Mar Ecol Prog Ser 1:145-152 Rinkevich B, Loya Y (1987) Variability in the pattern of sexual reproduction of the coral Stylophora pistillata at Eilat, Red Sea: a long-term study. The Biological Bulletin 173:335-344 Ritson-Williams R, Arnold SN, Fogarty ND, Steneck RS, Vermeij MJ, Paul VJ (2009) New perspectives on ecological mechanisms affecting coral recruitment on reefs. Smithsonian Contributions to the Marine Sciences Rosser NL (2013) Biannual coral spawning decreases at higher latitudes on Western Australian reefs. Coral reefs 32:455-460 Sakai K (1997) Gametogenesis, spawning, and planula brooding by the reef coral Goniastrea aspera (Scleractinia) in Okinawa, Japan. Marine Ecology Progress Series 151:67-72 Sammarco PW, Andrews JC (1989) The Helix experiment: Differential localized dispersal and recruitment patterns in Great Barrier Reef corals. Limnology and Oceanography 34:896- 912 Scheer G, Pillai CSG (1983) Report on the stony corals from the Red Sea. Zoologica 45:1-198 Shlesinger Y, Loya Y (1985) Coral community reproductive patterns: Red Sea versus the Great Barrier Reef. Science 228:1333-1335 31

Shlesinger Y, Goulet T, Loya Y (1998) Reproductive patterns of scleractinian corals in the northern Red Sea. Mar Biol 132:691-701 Simpson C (1985) Mass spawning of scleractinian corals in the Dampier Archipelago and the implications for management of coral reefs in Western Australia. Bulletin 244:1-35 Stehli FG, Wells JW (1971) Diversity and Age Patterns in Hermatypic Corals. Systematic Biology 20:115-126 Steneck RS (2006) Staying Connected in a Turbulent World. Science 311:480-481 Szmant A (1986) Reproductive ecology of Caribbean reef corals. Coral reefs 5:43-53 Szmant AM, Gassman NJ (1990) The effects of prolonged “bleaching” on the tissue biomass and reproduction of the reef coral Montastrea annularis. Coral reefs 8:217-224 Van Woesik R (2010) Calm before the spawn: global coral spawning patterns are explained by regional wind fields. Proceedings of the Royal Society B: Biological Sciences 277:715- 722 Van Woesik R, Lacharmoise F, Köksal S (2006) Annual cycles of solar insolation predict spawning times of Caribbean corals. Ecology Letters 9:390-398 Veron JEN (2000) Corals of the world. Australian Institute of Marine Science, Townsville Vermeij MJA, Marhaver KL, Huijbers CM, Nagelkerken I, Simpson SD (2010) Coral larvae move toward reef sounds. Plos One 5:e10660 Wallace CC (1999) Staghorn corals of the world: a revision of the coral genus Acropora (Scleractinia; Astrocoeniina; Acroporidae) worldwide, with emphasis on morphology, phylogeny and biogeography. CSIRO Publishing Ward S, Harrison P, Hoegh-Guldberg O (2002) Coral bleaching reduces reproduction of scleractinian corals and increases susceptibility to future stress. Proceedings of the Ninth International Coral Reef Symposium, Bali, 23-27 October 2000 2:1123-1128 Watson GJ, Bentley MG, Gaudron SM, Hardege JD (2003) The role of chemical signals in the spawning induction of polychaete worms and other marine invertebrates. Journal of Experimental Marine Biology and Ecology 294:169-187 Wilkinson CR (1999) The 1997-1998 mass bleaching event around the world. Compilation of Internet Reports. Global Coral Reef Monitoring Network, Australian Institute of Marine Science special publication Willis BL, Babcock RC, Harrison PL, Oliver JK, Wallace CC (1985) Patterns in the mass spawning of corals on the Great Barrier Reef from 1981 to 1984. 5th International Coral Reef Symposium 4:343-348

CHAPTER 2: REPRODUCTIVE SEASONALITY OF THE ACROPORA

ASSEMBLAGE IN THE CENTRAL RED SEA

Abstract

Reproduction in the Red Sea was until recently thought to be asynchronous with ecologically significant species spawning on different days and months of the lunar calendar. Spawning observations in the central Red Sea have since confirmed that multi- specific synchronous spawning does occur in the Red Sea. Here we report and quantify spawning synchrony in the Acropora coral assemblage in two regions in the Saudi

Arabian Red Sea over two consecutive years and compare the level of synchrony with five other locations in the Indo-Pacific. Populations of Acropora species take between 1-

3 months for all individuals to release their gametes in the central Red Sea. High proportions of colonies with mature eggs each month suggest that some colonies from some species spread the release of gametes over two consecutive months. In the central

Red Sea, synchrony was not consistent in different years with lower values in 2012.

Comparisons with other location in the Indo-Pacific suggest that latitude does not play a role in the level of synchrony, and furthermore that different locations with similar synchrony can be subject to very different environmental variables. It is not known whether the level of synchrony and the length of the reproductive season is the result of adaptation to local environmental conditions or whether it is the consequence of the presence or absence of certain environmental cues. More research on the influence of 33 each environmental variable is still necessary in different regions worldwide before we can attempt to understand the cues responsible for spawning synchrony.

Keywords Coral reefs – reproduction – Saudi Arabia – Scleractinia – spawning – synchrony

Introduction

Reproductive synchrony is known to occur in animal as well as in plant populations.

Some examples with strikingly high synchrony include some species of bamboos, which will grow during a species-specific period of 3-120 years, then reproduce synchronously before dying and letting the cycle restart (Janzen 1976) and the Pacific Ridley turtle which, in numbers close to several tens of thousands, lays eggs during synchronous mass nestings called “arribadas” (Hughes and Richard 1974). In seasonal environments, the clustering of reproductive activities at a certain period is caused by different species or populations selecting the same favourable environmental conditions for reproduction.

However, reproductive synchrony also occurs in aseasonal environments such as tropical forests. Furthermore, even in seasonal environments, synchrony is often higher than expected based solely on environmental variability (Ims 1990). These examples suggest that reproductive synchrony is also be driven by selective pressures other than environmental ones, such as inner biological clocks or predation avoidance (Langerhans

2007; Naylor 2001).

High reproductive synchrony between several species of scleractinian corals was observed for the first time on the Great Barrier Reef in 1981 (Harrison et al. 1984; Willis et al. 1985) and since then on reefs in many other locations around the world (Baird et al.

2009a). Cues involved in synchronising the release of gametes in corals at different levels are thought to be a combination of water temperatures, insolation, moon cycles, photoperiod, and regional wind fields (Baird et al. 2009a; Rosser 2013). Different cues seem to be stronger in some regions as well as in some species and not all cues seem to have the same influence on reproductive synchrony. For example, in Western Australia the proportion of Acropora corals spawning during the main reproductive season is similar at all latitudes but the proportion of species involved in the second spawning season decreases with higher latitudes (Rosser 2013). Also, selective pressures on the timing of the final gamete release have resulted in most scleractinian corals spawning at night, a strategy believed to have been selected to avoid predation of fish on the gametes

(Baird et al. 2001). However, Pocillopora species on the GBR and P. verrucosa in the

Red Sea all spawn in the daytime, not releasing egg-sperm bundles but free gametes

(Bouwmeester et al. 2011 / see annexe 1, p. 138; Schmidt-Roach et al. 2012).

Reproductive synchrony of coral assemblages has been recorded in a wide range of regions. In contrast, some other regions report temporal reproductive isolation where species of the assemblage appear not to overlap in their spawning periods (Baird et al.

2009a). Both patterns occur in the Red Sea: coral spawn asynchronously in Eilat in the

Gulf of Aqaba (Shlesinger and Loya 1985) but synchronously in the central Red Sea

(Bouwmeester et al. 2011 / see annexe 1, p. 137). However, spawning synchrony is rarely 35 defined quantitatively (Baird and Guest 2009). Baird et al. (2009c) proposed to quantify spawning synchrony with an index modified from Simpson’s diversity index, which measures the probability that two colonies taken randomly in the population spawn during the same month. However, this index is not suited to population were individual colonies spawn on more than one occasion (e.g. Bauman et al. 2011; and see Chapter 3).

Here, we use a synchrony index developed to measure flowering synchrony, which allows for individual spawning in consecutive months (Marquis 1983).

The aim of this study was to (a) study the reproductive condition of Acropora corals in the central Red Sea over an entire year to determine when species were likely to be spawning, (b) determine the level of synchrony within and between species, (c) compare spawning synchrony at a number of different regions in the Indo-Pacific and (d) assess stability of spawning synchrony between years.

Methods

Study Sites

The reproductive condition of scleractinian coral colonies was examined in 2011 and

2012. Two to four reefs covering most of the regional Acropora assemblage were surveyed monthly 1-2 weeks prior to each full moon, starting from before the April full moon for a total of 2-5 consecutive months, in two regions of the Saudi Arabian Red Sea separated by 266 km (Fig. 1): Thuwal (22°17'43.24"N, 39°05'59.11"E), and Al Lith (20

°09'1.78"N, 40°16'2.36"E). Additionally, one to two reefs were surveyed prior to the full moons of December 2012 in Thuwal and of October 2012 in Al Lith. 36

Figure 1 Map of the Red Sea showing the 2 study areas in this chapter: Thuwal and Al Lith

Assessing the reproductive condition of coral colonies

At each reef, the reproductive condition of every Acropora colony encountered in 60 min haphazard dives at 0-15m was recorded. The reproductive condition was established by breaking coral branches below the sterile apical zone to expose the developing oocytes

(Wallace 1985). The level of gametogenesis synchrony within a colony is generally high

(Wallace 1985). However, not every fracture will penetrate a polyp (Baird et al. 2009b).

As a consequence three branches were broken in each colony and a colony was only recorded as empty if none of the three broken branches exposed oocytes. Mature oocytes are easily visible in the field in Acropora species as they are pigmented in pink or red 37

(Harrison 1984) and are expected to be released with the subsequent full moon (Harrison and Wallace 1990; Hayashibara et al. 1993). Each surveyed colony was classified into one of three reproductive conditions: mature (oocytes pigmented), immature (oocytes white) or empty (oocytes too small to be detected or absent) (Baird et al. 2002). To look at reproductive patterns at the population scale, we selected three common species in the

Red Sea, Acropora humilis, A. hyacinthus and A. eurystoma, and examined them individually.

Synchrony index

A definition of synchrony on the scale of months is appropriate to test for proximate cues involving in determining the length of the spawning season (Baird et al. 2009a). To compare the level of spawning synchrony between Thuwal and Al Lith during two consecutive years, we used the following index originally developed for flowering synchrony by Marquis (1988).

m1 m2 m3 mn SM  t n  p1  t n  p2  t n  p3  ...  t n  pn mt mt mt mt t 1 t 1 t 1 t 1

Where t is the month of survey, n in the number of months surveyed, mt is the number of mature coral colonies in month t, pt is the proportion of colonies sampled in month t that

tn are mature, and  mt is the total number of coral colonies recorded as mature during the t1 period studied. This index is practical for determining synchrony in coral spawning as it allows for different sample sizes each month. If SM is close to 0, the spawning synchrony as well as the proportion of colonies reproducing is low. A SM value equal or close to 1 38 means the proportion of colonies spawning in one month is close to 100%, with most colonies spawning in the same month(s).

Marquis’ synchrony index was calculated for each Acropora species for Thuwal and Al

Lith. The individual SM values were then averaged to calculate a species level index SM

(S). A second index SM (A) was also calculated for the assemblage level by grouping reproductive observations of all colonies of all species. When no mature colonies were observed in a species, the species was removed from the species level index but kept for the assemblage level index. Data was analysed from 2011 as well as 2012 to verify the similarity or difference in synchrony in two consecutive years.

Spawning synchrony throughout the Indo-Pacific

To compare the level of synchrony of corals between the central Red Sea and other locations in the Indo-Pacific, we chose to quantify spawning synchrony in the Acropora assemblage from data collected in Singapore (Guest et al. 2005a), in the Philippines

(Guest, unpublished data), in Lord Howe Island, Australia (Baird, unpublished data), in

Kimbe Bay, Papua New Guinea (Baird, unpublished data), and in Okinawa, Japan (Baird et al. 2009a). We calculated SM at the assemblage level and at the species level.

Synchrony in the Red Sea was averaged over the two years for comparison with other locations. We also chose to compare in more detail the synchrony of three common species distributed in most locations of the Indo-Pacific: Acropora humilis, A. hyacinthus and A. tenuis. In the Red Sea, A. tenuis is replaced by A. eurystoma. Additionally, A. hyacinthus is not clearly differentiated from A. lamarcki and thus we combine data from 39 both morphs together. At Lord Howe Island, only data for A. hyacinthus is shown as no mature colonies of A. humilis were found and A. tenuis was not sampled.

Coral Taxonomy

Coral taxa were identified in the field, from images taken in the field, or from skeletons, according to Veron (2000). A field identification catalogue for Acropora species in the central Red Sea was established throughout this study and contains all the surveyed species.

Environmental Data

Hourly measurements of wind velocity were extracted from a meteorological buoy 50 km offshore from Thuwal (22° 9'38.16"N, 38°30'4.32"E) from October 2008 to December

2010. Van Woesik (2010) suggested that corals may couple gamete release when winds are light, that long spawning seasons are expected in regions with long periods of calm weather (winds under 6m/s during 20% of the year or more), and that spawning was unlikely to happen in months with winds higher than 6 m/s during 70% of the month or more. We thus calculated the proportion of the year during which winds are under 6 m/s and the number of months in the year during which the wind speed is less than 6 m/s.

Results

Reproductive patterns in Red Sea assemblages

Mature Acropora colonies were first observed prior to the April full moon in 2011 and

2012 in Thuwal and in Al Lith, ranging from 39% of colonies sampled in 2012 to 65% in 40

2011 (Fig. 2). In Thuwal, many mature colonies were observed again prior to the May full moon (33% in 2011 and 44% in 2012) whereas in Al Lith no, or few colonies, still contained visible oocytes (0% in 2011 and 12% in 2012). In 2011, all colonies (n=110) in

Al Lith were expected to have spawned in April. However, an intermediate reproductive survey conducted a day after the full moon indicated that the proportion of mature colonies had dropped from 65% to 33%. Prior to the following full moon no mature colonies were observed, indicating that although colonies had all spawn in April, they had spawned over at least two nights. The proportion of species with at least one mature colony was the highest in Al Lith with 80-85% species involved in the April spawning. In

Thuwal, 60-67% of species were expected to spawn in April and 54-63% of species in

May. Immature colonies were recorded in both regions prior to the April full moon, ranging from 2-5% in 2011 to 45-61% in 2012. In Al Lith in 2012, 45% colonies had immature eggs prior to the April full moon but by the next month, only 16% colonies still had mature or immature eggs, suggesting that at least 29% colonies released white eggs in April, if not more.

Reproductive surveys were conducted again prior to the full moon of October 2012 in Al

Lith and of December 2012 in Thuwal, and a rapid survey was conducted prior to the

October full moon 2012 in Thuwal. The absence of mature or immature oocytes during these surveys indicates that Acropora species in the central Red Sea only reproduce in the spring, between April and June.

a b 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% April May June April April FM+1 May

c d 2 0 1 1 1 0 2 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% April May June April April FM+1 May

e f 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% April May June July Aug Dec April May June July Oct

g h 2 0 1 2 1 0 2 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% April May June July Aug Dec April May June July Oct

Figure 2 a-b and e-f Proportion of Acropora colonies with mature, immature or no visible eggs in a-b spring 2011 and in e-f 2012 c-d and g-h Proportion of Acropora species with at least one mature colony a-c-e-g in Thuwal b-d-f-h in Al Lith. Black sections are the proportion of colonies with mature oocytes, grey sections are the proportion of colonies with immature eggs and white sections are the proportion of colonies with no oocytes observed. In order to be closer to the lunar calendar, the month indicated is of the subsequent full moon, not of the reproductive survey itself.

Reproductive patterns in individual species in the Red Sea

In Thuwal, A. humilis does not spawn with most of the Acropora assemblage in April, but rather in June, although in some years a few colonies spawned in May. In Al Lith, however A. humilis spawns mostly in April, along with the rest of the Acropora assemblage. Similarly, A. eurystoma spawns mostly in May in Thuwal, but in April in Al

Lith. Acropora lamarcki spawns mostly in April in both locations, although a few colonies might spawn in May as well.

Interestingly, in Thuwal, Acropora humilis had immature oocytes from April through

June 2011 (respectively 38%, 75% and 83%). In July all colonies were empty (n=30) suggesting the species had either released white eggs or had matured just before spawning. In 2012, all A. humilis colonies were immature in April and May (n=10 and n=8), 69% colonies (n=13) were mature in June and all had spawned by July (n=25).

Although most colonies spawned in June about 31% had spawned in May. In Al Lith,

41% colonies in 2011 were mature in April (n=22) and colonies were all empty in May

(n=6). In 2012 92% (n=13) had immature oocytes in April, 28 % (n=18) had immature 43 oocytes in May (no mature colonies were observed) and in June 6% (n=16) were mature.

This means that this species spawned in three consecutive months in Al Lith, with a peak in April, and released mostly white eggs.

In Thuwal, 89% (n=37) A. lamarcki colonies were mature in April 2011 and 20% (n=5) colonies were mature in May, indicating some colonies either spawned twice or retained their eggs until a month later. In 2012, 73% colonies were mature in April (n=11) and

25% were mature in May (n=4). In Al Lith 87% colonies were mature in April. The species was unfortunately not resampled in May to verify spawning. However in 2012,

62% colonies were mature and 38% were immature (n=8) prior to the April full moon and all were empty by the May full moon (n=4).

Acropora eurystoma generally does not appear to release white eggs. However, in 2012 in Al Lith, although 31% colonies (n= 13) had mature eggs in April, another 62% had immature colonies. The following month, only 8% had mature eggs (n=12), indicating some white eggs could have been released around the April full moon or that the eggs matured just before spawning.

Spawning synchrony in the central Red Sea

Spawning synchrony, as calculated with Marquis’ synchrony index SM, was similar between Thuwal (SM (A) = 0.37-0.56; SM (S) = 0.67-0.73) and Al Lith (SM (A) = 0.37-0.53;

SM (S) = 0.60-0.63), at the assemblage level as well as at the average species level (Fig.

3a-b). Some variation in spawning synchrony was evident between 2011 and 2012 in the 44 central Red Sea. In 2012, the spawning synchrony at the assemblage level was 34% lower in Thuwal and 30% lower in Al Lith than in 2011 (Fig. 3a). The difference was less marked at the average species level although it was high for A. hyacinthus (lamarcki)

(32% decrease in Thuwal and 26% decrease in Al Lith). Synchrony was highly variable in A. humilis, both in Thuwal and Al Lith, but this seems to be due to very low values in one of the years (2011 in Thuwal and 2012 in Al Lith), most probably due to white eggs being released on those years. The difference in synchrony in A. tenuis (eurystoma) is clearly visible between Thuwal and Al Lith (Fig. 3e) with a much lower synchrony in Al

Lith. This is however an average between 2011 and 2012 and interestingly the difference is less prominent in 2011 where SM = 0.67 in Thuwal and SM = 0.43 in Al Lith.

a b 1 1

0.8 0.8 AL 0.6 0.6

0.4 0.4 TH

(S) (S) of species

(A) (A) assemblge of

M S M M 0.2 0.2 S AL

0 0 2011 2012 2011 2012

c d

1 1

0.8 0.8 TH LH SG AL LH 0.6 SG ALTH 0.6 OK PH OK

0.4 KB PH 0.4 KB

(S) species of (S)

(A) assemblge of (A)

M 0.2 0.2 S 0 0 0 10 20 30 0 10 20 30 Latitude Latitude

e 1 AL TH LH

0.8

SG OK 0.6 KB PH A. humilis

perspecies 0.4

A. hyacinthus

M S 0.2 A. tenuis

0 0 10 20 30 Latitude

Figure 3 Spawning synchrony calculated with the Marquis synchrony index SM: a at the assemblage level in the central Red Sea (Thuwal and Al Lith) over two consecutive years, b at the species level (average of SM of each species) in the central Red Sea (Thuwal and Al Lith) over two consecutive years, c at the assemblage level as function of latitude, d at the species level as function of latitude, e for Acropora humilis, A. hyacinthus and A. tenuis as function of latitude. 46

LH, Lord Howe Island, Australia; TH, Thuwal, Saudi Arabia; AL, Al Lith, Saudi Arabia; SG,

Singapore; OK, Okinawa, Japan; KB, Kimbe Bay, Papua New Guinea; PH, Philippines.

Spawning synchrony throughout the Indo-Pacific

At the assemblage level, synchrony was highest at Lord Howe Island (SM (A) = 0.50), followed by Thuwal (SM (A) = 0.46), Al Lith (SM (A) = 0.45), Singapore (SM (A) = 0.45),

Okinawa (SM (A) = 0.38), Kimbe Bay (SM (A) = 0.28) and the Philippines (SM (A) = 0.24)

(Fig. 3c). At the species level, the pattern is similar, although the amplitude of synchrony values is broader, going from SM (S) = 0.25 in Kimbe Bay to SM (S) = 0.70 in Thuwal

(Fig. 3d). The index of synchrony at the species level in Kimbe Bay is low compared to other locations but at the same time also very similar to its index of synchrony at the community level, indicating that Acropora species in Kimbe Bay might be more evenly distributed throughout the assemblage than in the other locations which have higher synchrony values at the species level than at the assemblage level. The level of synchrony is more variable when examining species separately (Fig. 3e). The variability was the largest in the central Red Sea (SM = 0.33-0.79 in Thuwal and SM = 0.24-0.78 in Al Lith).

The synchrony of Acropora humilis, A. hyacinthus and A. tenuis in the Philippines was less variable with SM = 0.06-0.38, followed by Kimbe Bay (SM = 0.12-0.41) and Okinawa

(SM = 0.30-0.55). The less variability between species was found in Singapore (SM =

0.43-0.58). The synchrony of A. hyacinthus is very high (SM = 1) in Lord Howe, possibly due to the low number of samples collected throughout the season (n=3). There was no consistent pattern in synchrony according to latitude, neither at the assemblage level, at the average species level, or at the individual species level (Fig. 3c-e). 47

Environmental Data

The wind velocity, between October 2008 and December 2010, ranged from 0.04-13.18 m/s, but was on average 5.03 m/s. During this same period, the winder was under 6 m/s for 66% of the time. Each month, the wind velocity was under 6 m/s for 41-86% of the month, and was on average under 6 m/s for 66% of the month. The monthly average was under 6 m/s for every month of the year. Half of the months had an average of 5-6 m/s winds (January, March, May, June, September, December) and the other half had an average of 4-5 m/s winds (February, April, July, August, October, November).

Discussion

This study has shown that reproduction of the Acropora assemblage in the central Red

Sea happens once a year, in the spring, between April and June. A second reproductive season, as observed in Western Australia, Indonesia (Gilmour et al. 2009; Permata et al.

2012) cannot be excluded. However, if a second reproductive season does exist, it does not include any Acropora species. The clustering of reproductive activities in a three- month period has led to a relatively high level of synchrony compared to other locations in the world, thus invalidating the earlier assumption that coral spawning is asynchronous in the Red Sea.

The reproductive season of Acropora in the central Red Sea lasts from April to June, with a strong peak in April. Spawning synchrony was similar at Thuwal and Al Lith. At Al

Lith, most Acropora species and colonies spawn in April but fewer colonies seemed to be 48 involved in spawning throughout the season. Reproductive surveys prior to the March full moon in Al Lith would be useful to verify if some colonies had already spawned in

March or whether a small proportion of the Acropora assemblage does not take part each year in spawning activities.

Reproduction of the Acropora community is spread out over April and May in Thuwal with at least half of the species involved in each month. In contrast, the reproductive season was shorter in Al Lith where a high proportion of colonies were mature in April.

A similarly short reproductive season occurs on the Egyptian coast in the northern Red

Sea, where all Acropora species but one are expected to spawn around the full moon in late April or early May (Hanafy et al. 2010). Spawning over two months could be a strategy selected naturally over time in the region around Thuwal, due to different environmental conditions. However, reefs in Thuwal suffered from heavy bleaching in the summer 2010 (Furby et al. 2013 / see annexe 1, p.167), and although corals surveyed in the present study were all on midshore or offshore reefs which significantly suffered less from bleaching, this raises the question whether the reproductive success of the

Acropora assemblage might have been affected and the length of the reproductive season altered. High levels of thermal stress can directly compromise the reproductive capacities of corals by affecting the reproductive output of populations (Baird and Marshall 2002), fecundity (Ward et al. 2002) and egg quality (Michalek-Wagner and Willis 2001). The influence on thermal stress on the length of the reproduction period of each species is however unknown. Alternatively, phases of the moon cycle are shifted by 11-12 days every year, which could result in a year with split spawning every 2-3 years, to realign 49 the lunar cycle with the time of the year (Willis et al. 1985; Baird et al. 2009c). The

Acropora assemblage was surveyed on different years between the northern Red Sea

(Hanafy et al. 2010) and the central Red Sea and differences in numbers of spawning months and synchrony could be due to differences in the alignment of the lunar month. A split spawning realignment year could also explain why in 2012, spawning was more equally split between April and May at Thuwal and spread out over an additional month at Al Lith, compared to 2011. This could also explain a general lower synchrony in 2012, as calculated with Marquis’ synchrony index. Further reproductive studies in the region should assist in determining the annual spawning rhythms in the Red Sea, and in identifying realignment years.

The high proportion of white eggs in some months followed by the absence of eggs a month later strongly suggests white eggs were released in a few occasions in Thuwal as well as in Al Lith, notably in Acropora humilis. Alternatively, in some species, egg might mature very quickly. Repeated reproductive surveys closer to the expected spawning date would reveal whether those species releasing eggs which were white within two weeks of the full moon, actually released white eggs or we able to undergo a rapid pigmentation of their eggs prior to spawning.

Although environmental factors have some influence on the spawning synchrony of coral communities, it is not known which factors drive the synchrony and whether the same factors are involved in different locations worldwide. It has been suggested that the degree of synchrony of spawning in corals can be predicted by regional wind fields and 50 that in regions with long calm periods, such as in Kenya, the reproductive season is longer, whereas in regions with calm weather during less than 20% of the year such as on the GBR, spawning occurs with much higher synchrony (van Woesik 2010). Winds off the coast from Thuwal are under 6 m/s for significantly more than 20 % of the year

(66%). Van Woesik (2010) also suggested that spawning would most probably happen during calm periods in months with winds below 6 m/s during 30% of the month or more.

Each month of the year in the central Red Sea fulfills the requirement and offers wind conditions suitable to spawning. However, spawning in the central Red Sea occurs during a three-month season and do not spawn in the calmest months of the year.

Furthermore on reefs of the Whitsundays on the GBR, a region affected by stronger winds (van Woesik 2010) mature colonies have been found in five consecutive months, indicating that spawning on the GBR might not be as synchronous as originally thought

(Baird et al. 2009b) and thus that the degree of synchrony on those reefs is not necessarily correlated with a short calm season.

High spawning synchrony is generally expected in broadcast spawning species to maximize fertilization of gametes at the sea surface (Harrison et al. 1984; Oliver and

Babcock 1992; Guest et al. 2005b). It is however not known whether coral communities which don’t spawn with high synchrony have selected that strategy over time or whether they simply lack the environmental cues necessary to cluster their reproductive activities further. In any case, synchrony is likely to vary from one year to the next, as shown in the central Red Sea in two consecutive years. This variability is not unique, on the GBR, recruitment rates and fecundity varied greatly between years as well as between regions 51

(Hughes et al. 2000). In the central Red Sea, the beginning of the spawning season, the length of the spawning season, and the spawning synchrony, are similar between Thuwal and Al Lith, although sea surface temperatures in Al Lith are on average 1.4°C warmer than in Thuwal (E. Tyler and F. Cagua pers. comm.). Winds are similar in both regions for most of the year and blowing from the north, but every year for a few weeks starting in November, Al Lith is subjected to the Arabian monsoon being at its northern limit, and gets stronger southern winds (Sofianos 2003). However, the difference in sea surface temperatures and winds surprisingly does not seem to create differences in spawning patterns. On the Great Barrier Reef, corals on inshore reefs spawn a month earlier than corals on offshore reefs at the same latitude, which has been explained by sea surface temperatures which start warming a month earlier on inshore reefs (Willis et al. 1985;

Babcock et al. 1986).

Red Sea Acropora assemblages show a spawning synchrony as determined by the

Marquis index very similar to those in Singapore and in Lord Howe, although the three locations are in very different latitudes (20-22°N, 1°N and 31°S respectively).

Assemblages in Okinawa, Kimbe Bay and the Philippines are also located at various latitudes (26°N, 5°S and 16°N respectively) but have all generally a lower spawning synchrony, with the lowest synchrony in Kimbe Bay and in the Philippines. This confirms further the lack of consistent reproductive synchrony according to latitude

(Baird et al. 2009c). The factors that determine the level of synchrony are however still to be identified and might be not be consistently the same in different locations. For example when comparing the central Red Sea with Singapore, two locations with similar 52 levels of synchrony, the environmental variables are strikingly different. They have very different latitudes (20-22°N and 1°N respectively), different annual water temperature variations (7°C (Raitsos et al. 2011) and 3-4°C (Guest et al. 2005a)) and different tidal amplitudes (20.3 cm (Sultan et al. 1996) and 240cm (Guest et al. 2005a). More data analysis from a wide range of locations along with a large number of environmental variables are still necessary to attempt to understand the driving forces behind the synchrony of spawning in scleractinian corals.

Multi-specific synchronous spawning is probably a characteristic of all speciose coral assemblages (Guest et al. 2005a; Baird et al. 2009c). The reproductive patterns of corals of the Red Sea seem to support the hypothesis, with high synchrony of mature gametes in the Acropora assemblage. Temporal variation in synchrony is most probably explained by lunar cycles, which don’t fall at the same time of the month each year, causing realignment years with generally lower synchrony. Spatial variation in reproductive synchrony is however less well understood and possibly, environmental variables interact differently on coral assemblages from different regions.

Acknowledgements

We thank K.A. Furby and J. Masterman for helping in collecting reproductive data in Al

Lith in 2011 and M.T. Khalil for assisting in the field during the remaining surveys. We also thank E. Tyler and F. Cagua for providing temperature data from Al Lith, and J.

Churchill and S. Lentz for facilitating access to environmental data in Thuwal. This work 53 was conducted as part of the PhD thesis of J. Bouwmeester in the King Abdullah

University of Science and Technology (KAUST).

References

Baird AH, Pratchett MS, Gibson DJ, Koziumi N, Marquis CP (2001) Short Communication: Variable palatability of coral eggs to a planktivorous fish. Marine and Freshwater Research 52:865-868 Baird AH, Marshall PA (2002) Mortality, growth and reproduction in scleractinian corals following bleaching on the Great Barrier Reef. Mar Ecol Prog Ser 237:133-141 Baird A, Marshall P, Wolstenholme J (2002) Latitudinal variation in the reproduction of Acropora in the Coral Sea. 9th International Coral Reef Symposium 1:385-389 Baird AH, Birrel CL, Hughes TP, McDonald A, Nojima S, Page CA, Prachett MS, Yamasaki H (2009a) Latitudinal variation in reproductive synchrony in Acropora assemblages: Japan vs. Australia. Galaxea, Journal of Coral Reef Studies 11:101-108 Baird AH, Guest JR (2009b) Spawning synchrony in scleractinian corals: Comment on Mangubhai & Harrison (2008). Marine Ecology Progress Series 374:301-304 Baird AH, Guest JR, Willis BL (2009c) Systematic and Biogeographical Patterns in the Reproductive Biology of Scleractinian Corals. Annual Review of Ecology, Evolution, and Systematics 40:551-571 Bouwmeester J, Khalil MT, De La Torre P, Berumen ML (2011) Synchronous spawning of Acropora in the Red Sea. Coral Reefs 30:1011-1011 Furby KA, Bouwmeester J, Berumen ML (2013) Susceptibility of central Red Sea corals during a major bleaching event. Coral reefs 32:505-513 Guest JR, Baird AH, Goh BPL, Chou LM (2005a) Reproductive seasonality in an equatorial assemblage of scleractinian corals. Coral Reefs 24:112-116 Guest JR, Baird AH, Goh BPL, Chou LM (2005b) Seasonal reproduction in equatorial reef corals. Invertebrate Reproduction & Development 48:207-218 Hanafy MH, Aamer MA, Habib M, Rouphael AB, Baird AH (2010) Synchronous reproduction of corals in the Red Sea. Coral Reefs 29:119-124 Harrison P, Wallace C (1990) Reproduction, dispersal and recruitment of scleractinian corals. Dubinsky Z (ed) Ecosystems of the world: coral reefs. Elsevier, Amsterdam, pp133–207 Harrison PL, Babcock RC, Bull GD, Oliver JK, Wallace CC, Willis BL (1984) Mass Spawning in Tropical Reef Corals. Science 223:1186-1189 Hayashibara T, Shimoike K, Kimura T, Hosaka S (1993) Patterns of coral spawning at Akajima Island, Okinawa, Japan. Marine Ecology Progress Series 101:253-253 Hughes DA, Richard JD (1974) The nesting of the Pacific ridley turtle Lepidochelys olivacea on Playa Nancite, Costa Rica. Mar Biol 24:97-107 Hughes TP, Baird AH, Dinsdale EA, Moltschaniwskyj NA, Pratchett MS, Tanner JE, Willis BL (2000) Supply-side ecology works both ways: the link between benthic adults, fecundity, and larval recruits. Ecology 81:2241-2249 Ims RA (1990) The ecology and evolution of reproductive synchrony. Trends in Ecology & Evolution 5:135-140 Janzen DH (1976) Why bamboos wait so long to flower. Annual Review of Ecology and Systematics 7:347-391 54

Langerhans RB (2007) Evolutionary consequences of predation: avoidance, escape, reproduction, and diversification. In: Elewa AT (ed) Predation in Organisms. Springer Berlin Heidelberg, pp177-220 Marquis RJ (1988) Phenological variation in the neotropical understory shrub Piper arielanum: causes and consequences. Ecology 69:1552-1565 Michalek-Wagner K, Willis BL (2001) Impacts of bleaching on the soft coral Lobophytum compactum. II. Biochemical changes in adults and their eggs. Coral reefs 19:240-246 Naylor E (1999) Marine animal behaviour in relation to lunar phase. Earth, Moon, and Planets 85:291-302 Oliver JK, King BA, Willis BL, Babcock RC, Wolanski E (1992) Dispersal of coral larvae from a lagoonal reef--II. Comparisons between model predictions and observed concentrations. Continental Shelf Research 12:873-889 Raitsos DE, Hoteit I, Prihartato PK, Chronis T, Triantafyllou G, Abualnaja Y (2011) Abrupt warming of the Red Sea. Geophys Res Lett 38:L14601 Rosser NL (2013) Biannual coral spawning decreases at higher latitudes on Western Australian reefs. Coral reefs 32:455-460 Schmidt-Roach S, Miller KJ, Woolsey E, Gerlach G, Baird AH (2012) Broadcast Spawning by Pocillopora Species on the Great Barrier Reef. PLoS ONE 7:e50847 Shlesinger Y, Loya Y (1985) Coral community reproductive patterns: Red Sea versus the Great Barrier Reef. Science 228:1333-1335 Sofianos SS, Johns WE (2003) An Oceanic General Circulation Model (OGCM) investigation of the Red Sea circulation: 2. Three-dimensional circulation in the Red Sea. Journal of Geophysical Research: Oceans 108:3066 Sultan SAR, Ahmad F, Elghribi NM (1996) Sea level variability in the central Red Sea. Oceanologica acta 19 Van Woesik R (2010) Calm before the spawn: global coral spawning patterns are explained by regional wind fields. Proceedings of the Royal Society B: Biological Sciences 277:715- 722 Veron JEN (2000) Corals of the world. Australian Institute of Marine Science, Townsville Wallace CC (1985) Reproduction, recruitment and fragmentation in nine sympatric species of the coral genus Acropora. Mar Biol 88:217-233 Ward S, Harrison P, Hoegh-Guldberg O (2002) Coral bleaching reduces reproduction of scleractinian corals and increases susceptibility to future stress. Proceedings of the Ninth International Coral Reef Symposium, Bali, 23-27 October 2000 2:1123-1128 Willis BL, Babcock RC, Harrison PL, Oliver JK, Wallace CC (1985) Patterns in the mass spawning of corals on the Great Barrier Reef from 1981 to 1984. 5th International Coral Reef Symposium 4:343-348

SUPPLEMENTARY DATA

Table S1 Proportion of Acropora colonies (%) in each species with white eggs (im.) and mature eggs (mat.) and the number (n) of colonies sampled in Thuwal

April May June Species 2011 2011 2011 immature mature n immature mature n immature mature n A. cytherea - - - 0 100 1 0 0 1 A. eurystoma 0 0 1 0 67 3 0 0 8 A. formosa 0 0 1 - - - 0 0 2 A. gemmifera 0 100 24 0 0 7 0 0 2 A. hemprichii 0 0 2 0 75 8 0 0 7 A. humilis 38 0 13 75 6 16 83 4 24 A. hyacinthus 0 100 2 0 100 3 0 0 4 A. lamarcki 0 89 37 0 20 5 0 0 11 A. listeri ------0 0 8 A. loripes 0 100 1 0 0 3 0 0 3 A. lutkeni - - - 0 78 9 0 0 8 A. maryae ------0 0 2 A. microphthalma - - - 0 0 1 - - - A. nasuta 0 100 4 ------A. nobilis 0 50 6 ------A. pharaonis 0 17 12 0 13 8 0 0 4 A. plantaginea 0 0 3 0 0 4 0 0 3 A. polystoma - - - 0 0 2 0 0 2 A. samoensis 0 100 1 - - - 0 0 3 A. secale - - - 0 0 1 25 0 4 A. valida 0 0 1 0 50 8 7 0 14 A. variolosa 0 100 2 0 29 7 0 0 8

April May June July August December Species 2012 2012 2012 2012 2012 2012 im mat n im mat n im mat n im mat n im mat n im mat n

A. anthocercis - - - 0 0 2 0 0 3 0 0 2 0 0 5 0 0 3 A. austera - - - 0 100 2 - - - 0 0 1 ------A. clathrata 100 0 1 ------0 0 2 0 0 1 A. cytherea 50 50 4 0 0 3 0 0 6 0 0 3 0 0 1 - - - A. donei ------0 0 1 0 0 2 0 0 3 0 0 2 A. downingi 100 0 2 0 0 3 0 0 2 ------0 0 5 A. elseyi ------0 0 1 - - - 56

A. eurystoma 89 11 9 0 100 9 0 0 9 0 0 6 0 0 3 0 0 2 A. gemmifera 57 43 7 0 25 4 0 0 4 0 0 4 0 0 3 0 0 2 A. haimei - - - 0 0 1 ------A. hemprichii 22 56 9 0 33 3 0 0 8 0 0 4 0 0 6 0 0 3 A. humilis 100 0 10 100 0 8 8 62 13 0 0 25 0 0 2 0 0 3 A. hyacinthus 60 40 5 0 0 1 ------A. lamarcki 27 73 11 0 25 4 0 0 5 0 0 4 - - - 0 0 1 A. listeri 100 0 1 ------A. loripes 33 67 3 0 75 4 - - - 0 0 2 0 0 2 0 0 2 A. lutkeni 80 0 5 0 100 2 0 14 7 0 0 4 0 0 3 0 0 3 A. microclados 67 33 3 0 60 10 0 0 8 0 0 9 0 0 1 0 0 3 A. microphthalma 100 0 2 0 100 2 0 0 5 0 0 2 0 0 2 - - - A. monticulosa - - - 0 50 2 0 0 5 0 0 2 ------A. nasuta ------0 0 1 0 0 1 ------A. nobilis ------0 0 1 ------A. ocellata - - - 0 0 1 ------A. parapharaonis 0 100 1 100 0 2 0 0 5 0 0 1 0 0 4 0 0 3 A. pharaonis 75 0 4 17 0 6 0 0 2 0 0 3 ------A. plantaginea 50 50 4 0 0 0 0 0 2 0 0 3 0 0 1 - - - A. polystoma 75 25 4 0 100 3 0 0 1 0 0 1 0 0 2 - - - A. rufus - - - 0 0 1 ------A. samoensis 50 50 2 0 0 2 0 0 2 0 0 1 0 0 2 0 0 2 A. secale 100 0 2 0 80 5 0 0 5 0 0 5 ------A. selago 50 50 2 - - - 0 0 1 ------A. valida 70 30 10 0 58 12 0 8 12 0 0 11 0 0 2 0 0 3 A. variolosa 43 57 7 0 0 4 0 0 3 0 0 6 0 0 1 0 0 3 A. verweyi 0 100 2 0 100 1 0 0 1 0 0 1 ------

Table S2 Proportion of Acropora colonies (%) in each species with white eggs (im.) and mature eggs (mat.) and the number (n) of colonies sampled in Al Lith

April April FM+1 May Species 2011 2011 2011 immature mature n immature mature n immature mature n A. austera 0 100 7 ------A. anthocercis 0 60 5 ------A. cytherea 0 100 3 ------A. downingi 0 0 8 ------A. eurystoma 0 43 14 0 12 8 0 0 1 A. formosa 0 75 4 - - - 0 0 1 A. gemmifera 0 86 37 0 100 1 0 0 1 57

A. haimei - - - 0 0 2 - - - A. hemprichii - - - 0 0 5 - - - A. humilis 14 41 22 10 10 10 0 0 6 A. hyacinthus 0 100 4 0 100 1 0 0 1 A. lamarcki 0 87 16 0 100 1 - - - A. lutkeni - - - 0 0 1 - - - A. maryae 0 0 1 0 0 1 - - - A. microclados 0 67 6 ------A. 0 14 7 0 0 1 0 0 1 microphthalma A. monticulosa 0 33 3 ------A. nasuta - - - 0 0 1 - - - A. nobilis 0 20 5 ------A. pharaonis 0 0 10 0 0 4 - - - A. plantaginea 0 0 6 0 0 1 - - - A. samoensis 0 25 4 ------A. selago 0 100 2 ------A. valida - - - 0 0 5 0 0 4 A. variolosa 0 10 10 0 0 2 - - -

April May June July October Species 2012 2012 2012 2012 2012 im mat n im mat n im mat n im mat n im mat n A. anthocercis 25 50 4 0 0 1 0 0 2 0 0 2 0 0 6 A. austera 0 100 2 0 33 3 0 0 3 0 0 1 0 0 4 A. clathrata 80 20 5 0 0 2 0 0 4 0 0 1 0 0 2 A. cytherea 50 0 2 0 0 2 - - - 0 0 2 0 0 2 A. donei 0 100 1 0 0 1 0 0 1 ------A. downingi 25 50 4 0 0 1 0 0 2 ------A. elseyi ------A. eurystoma 62 31 13 0 8 12 0 0 5 0 0 4 0 0 3 A. formosa - - - 0 0 1 ------A. gemmifera 27 73 11 0 0 2 0 0 1 0 0 4 0 0 2 A. haimei - - - 0 0 1 ------A. hemprichii 0 75 4 0 0 7 0 0 3 0 0 8 0 0 5 A. humilis 92 8 13 28 0 18 6 6 16 0 0 8 0 0 6 A. hyacinthus 0 100 1 0 0 2 ------A. lamarcki 37 63 8 0 0 2 - - - 0 0 1 - - - A. listeri - - - 0 100 1 ------A. loripes 33 33 3 0 0 1 0 0 4 0 0 2 0 0 5 A. lutkeni 67 17 6 0 75 8 0 0 1 0 0 4 - - - 58

A. maryae 100 0 1 0 0 2 0 0 3 - - - 0 0 1 A. microclados 25 75 8 0 0 8 0 0 1 0 0 3 0 0 4 A. microphthalma 50 50 4 0 0 3 - - - 0 0 1 0 0 6 A. monticulosa 0 100 2 ------0 0 1 0 0 1 A. nasuta - - - 0 100 2 0 0 2 - - - 0 0 2 A. ocellata 100 0 2 ------A. parapharaonis - - - 0 0 3 0 0 1 0 0 2 - - - A. pharaonis 50 50 2 0 0 2 0 0 3 ------A. plantaginea 50 0 2 0 0 3 - - - 0 0 1 0 0 5 A. polystoma 50 50 2 0 0 1 0 0 1 0 0 2 - - - A. rufus - - - 0 0 1 0 0 1 ------A. samoensis 0 100 2 0 0 1 0 0 1 - - - 0 0 1 A. secale 75 25 4 0 100 2 0 0 4 0 0 1 - - - A. selago 0 86 7 0 0 1 - - - 0 0 1 - - - A. squarrosa - - - 0 0 1 ------A. valida 64 18 11 0 8 12 0 0 13 0 0 11 0 0 5 A. variolosa 25 37 8 0 0 6 0 0 1 0 0 3 0 0 3 A. verweyi - - - 0 0 5 - - - 0 0 3 0 0 3

CHAPTER 3: MULTISPECIFIC SPAWNING SYNCHRONY WITHIN

SCLERACTINIAN CORAL ASSEMBLAGES IN THE RED SEA

Abstract

Early work on spawning times of corals in the Red Sea suggested that spawning times of the ecologically abundant species did not overlap, unlike on the Great Barrier Reef where many species spawn with high synchrony. More recent work in the northern and central

Red Sea has suggested a high degree of synchrony in the reproductive condition of

Acropora species: over 90% of species sampled in April/May contain mature gametes.

However, it has yet to be determined when most Acropora release there gametes. In addition, there is a lack of data for other ecologically important scleractinian species such as merulinids and poritids. Here, we document the date and time of spawning for 51 species from 8 clades in the central Red Sea over three years. Spawning occurs on nights around the full moon, the spawning season lasts at least four months from April until

July, and observations are consistent with the few other records from the Red Sea. The number of Acropora species spawning was highest in April with 13 species spawning two nights before the full moon in 2011 and 13 species spawning on the night of the full moon in 2012. The total number of species spawning was high in April, May and June and involved 15 to 19 species per month in 2012. Only four species spawned in July

2012. Spawning patterns in the central Red Sea have similarities with patterns in other locations worldwide and we compare here spawning observations with two other locations for which there is relatively complete understanding of annual patterns of spawning. Corals in the Philippines and Okinawa, Japan, spawn on nights around the full 60 moon over a period of three to 4 months. In all three locations Acropora are the first species to spawn. Timing of coral spawning is valuable knowledge and should be taken into account when implementing protection measures on local reefs or when developing new projects in coastal reef areas.

Keywords Coral reefs, reproduction, Saudi Arabia, multi-specific spawning

Introduction

Knowing the spawning day of different coral species and the level of spawning synchrony between species allows for much important ecological research into topics such as fertilisation success (Negri et al. 2007), larval development (Palmer et al. 2012,

Graham et al. 2013), pelagic larval durations (Gilmour 1999, Graham et al. 2008), modeling larval dispersal (Connolly & Baird 2010, Tay et al. 2011, Figueiredo et al.

2013) and restoring coral reefs (Guest et al. 2011, Baria et al. 2012). All this information provides valuable information for long-term coral conservation and management (Shanks et al. 2003, Steneck et al. 2009).

Most scleractinian coral colonies have an annual gametogenesis cycle and reproduce only once a year with high synchrony within species. In numerous locations around the world, numerous coral species spawn synchronously during mass- or multi-specific spawning events involving different species from different families (Baird et al. 2009a, Harrison

2011). The first mass spawning observations were on the GBR and revealed a high number of species spawning over a few nights between the full and last quarter moon in late spring (Harrison et al. 1984; Willis et al. 1985; Babcock et al. 1986). Since then, 61 many multi-specific spawning observations were made in other locations of the Indo-

Pacific and in the Caribbean (Guest et al. 2005; Baird et al. 2009).

The Red Sea has been described as a marine biodiversity hotspot with high coral richness and endemism (Hughes et al. 2002; Roberts et al. 2002). However it remains a highly under-studied region (Berumen et al. 2013 / see annexe 1, p. 155). In particular, the reproductive ecology of scleractinian species has received relatively little attention.

Reproductive surveys and night-time observations in the early eighties in Eilat, Gulf of

Aqaba, northern Red Sea, suggested the absence of synchronous spawning with different species reproducing in different seasons and months and throughout the lunar cycle

(Shlesigner and Loya 1985). Later, Shlesinger et al. (1998) added data for 11 additional broadcast spawning coral species, bringing the total to 21 species studied at Eilat. Except for two species which were found to spawn in April, the reproductive season lasted three to four months, from June to September (Shlesinger & Loya 1985; Shlesinger et al.

1998). Due to the lack of data for additional species and the lack of similar work in the rest of the Red Sea it was assumed for many years that coral reproduction was asynchronous in the entire Red Sea basin and that Red Sea corals show temporal reproductive isolation (Richmond & Hunter 1990, Wilson & Harrison 2003, Mangubhai

& Harrison 2008), reinforcing the earlier assumption that synchronous multi-specific spawning is not a characteristic of all coral communities (Harrison 1990, Richmond &

Hunter 1990). However, a recent study conducted on the Egyptian coast in the northern

Red Sea revealed 85% of Acropora colonies from 12 species had mature oocytes in

Marsa Alam in April 2008 and 99% of Acropora colonies from 17 species had mature 62 oocytes in Hurghada in April 2009. Subsequent sampling revealed the absence of oocytes in all but one of these species, indicating that spawning had occurred sometime in the previous couple of weeks, most likely around the full moon (Hanafy et al. 2010). This assumption was confirmed by the observation of ten Acropora species spawning on the same night in April 2011 on the Saudi Arabian coast in the central Red Sea

(Bouwmeester et al. 2011 / see annexe 1, p. 137). Nonetheless, there are few other observations on the timing of gamete release for scleractinian corals in the Red Sea

In this study, we (a) present reproductive data from a wide range of scleractinian species in order to assist in pinpointing the timing of spawning and (b) describe direct coral spawning observations in the central Red Sea. We then (c) examine the timing of spawning between different coral groups and the consistency of the timing in consecutive years where possible and (d) compare spawning observations from the Red Sea to other locations with a similar reproductive season length.

Materials and Methods

Study Sites

The reproductive condition of a variety of species of the coral assemblage was examined from April to July 2011 in Al Fahal (22°13'26.51"N, 38°58'10.12"E), a mid-shore reef in the central Red Sea located 12 km off the coast of Thuwal (Fig. 1). Night time observations were conducted from April to June 2011 and April to July 2012 in Al Fahal or, when weather conditions did not allow boat access to the reef (May 2011, June and

July 2012), in Dreams Beach Reef (21°45'24.43"N, 39° 3'7.05"E), a fringing reef north of 63

Jeddah with a similar coral assemblage. In 2013 night time observations were conducted in April only, each night on a different reef within a range of 30km from Al Fahal.

Figure 1 Map of the Red Sea showing the two study sites: Al Fahal, 10 km off the coast form

Thuwal and Dreams Beach Reef on a fringing reef in Jeddah.

Moon phases and environmental variables

Scleractinian corals are known to reproduce around the full moon (Babcock et al. 1986), in the spring during warming sea surface temperatures that act as one of the cues to synchronising gamete maturation (van Woesik et al. 2006). Moon cycles from 2011 to

2013 are presented in Figure 2. To record the exact temperature variation throughout the year and during the spawning season, seawater temperatures were recorded hourly with a

HOBO Water Temp Pro v2 logger installed at the depth of approximately 0.7 m on a mooring between Thuwal and Jeddah (21°58'45.48"N, 38°50'41.28"E) from April to May in 2011 and on the reef edge in Al Fahal from December 2011 to August 2013.

Figure 2 Moon phases in Thuwal in 2011, 2012 and 2013. The grey bars above the moon phases represent the nights where the coral assemblage was surveyed for spawning behaviour and the inverted black triangles represent the days/nights when spawning was observed.

Histological examination

To narrow down the spawning window of different species from different families, samples from several abundant as well as less abundant coral species were collected and taken back to the lab for histological examination. Visible oocytes were recorded when observed in the field.

Coral colonies were examined and sampled between one and three times at different periods from April to July 2011 and two Porites species again in October and November

2011. Except for Porites colonies for which the same colonies were sampled each time, random colonies were examined in different months to allow for more freedom of movement to look for some of the less abundant species. The development of gametes is usually highly synchronous between polyps of a same species (Harrison and Wallace

1990) so one fragment is generally sufficient to assess the reproductive condition of a 65 colony (Baird et al. 2011). Fragments containing several polyps were collected on

SCUBA with a hammer and a chisel, were then brought back to the lab, and fixed in 10% seawater formalin for at least 24 hours. Coral tissue was then decalcified in 5% hydrochloric acid, embedded in wax, sectioned at 5-7μm thickness and mounted on slides. The slides were then stained for 8-20 minutes with haematoxylin stain triple strength, which stains nuclei blue, and for 5 minutes with Eosin Y solution alcoholic, which stains cytoplasmic elements in various shades of red (Baird et al. 2011), were dehydrated and mounted in DPX (Di-N-Butyle Phthalate in Xylene). The histological sections were then examined with a Leica DM3000 microscope to determine the stage of maturity of gametes, photographed with a Leica DFC495 camera and each colony was classified into one out of four stages of oogenesis or spermatogenesis (Table 1). Stage IV gametes were expected to be released during the following full moon.

Table 1 Criteria for classification of oocytes and spermatocytes into developmental stages adapted from Baird et al. (2011)

Stage Oogenesis Spermatogenesis Enlarged interstitial cells in the mesoglea of mesenteries. Nucleus Small clusters of interstitial cells near or I makes up the bulk of the oocyte. entering the mesoglea. Stain deep blue Oocytes well-spaced Accumulation of blue staining yolk Clusters of spermatocytes with distinct II around the nuclei which remains in the spermary boundary. Large nuclei centre of the oocyte arranged peripherally Spermatocytes are smaller with smaller Oocytes of variable size. Nucleus nuclei. Number of cells within the III migrates to periphery of oocyte spermary is larger. Prominent central lacunae Oocytes very crowded. Nucleus at IV Spermatozoa with tails periphery

Coral spawning observations

Following the appearance of mature oocytes in surveys of the reproductive condition in

Acropora (Baird et al. 2002) night-time surveys commenced 2 nights before the full moon (Fig. 2) in April 2011 for 5 nights and continued 3 nights before the May full moon for 9 nights and from 3 nights before the June full moon for 6 nights. In 2012, surveys commenced each month 3 nights before the full moon and lasted in total 10 nights in

April, 8 nights in May, 11 nights in June and 6 nights in July. In May 2011 and in June and July 2012, the surveys were conducted from North Obhur. Surveys were all done on

SCUBA starting 30 minutes after sunset for 2 hours in 2011 and for 4 hours in 2012. All scleractinian corals were carefully examined while following a random diving path between the reef edge and the depth of 10m, looking for setting behaviour of polyps which immediately precedes the release of gametes, and spawning behaviour.

On Acropora spawning nights, all Acropora colonies were regularly monitored throughout the night, and spawning behaviour was described with the following categories: no activity, setting, defined by egg-sperm bundles appearing under the oral disk of the polyps (Babcock et al. 1986), or spawning, when gametes were being released. In 2012, 73 random Acropora colonies were tagged prior to the first night-time spawning surveys in order to check whether some colonies were spawning more than once and were surveyed throughout April and May. Coral colonies were identified in the field when possible, or from field photographs or from bleached samples observed under a stereo microscope. Porites colonies were identified after imaging with a scanning 67 electron microscope. Coral species were classified into clades following Huang (2012) and Arrigoni et al. (2012).

Comparisons of spawning patterns in the Red Sea to those in other regions of the world

In order to compare the spawning patterns in the Red Sea to other regions of the globe we identified a number of studies from the literature with comparable data: i.e. at least 3 months of in situ spawning observations in a given season. These data are available for surprising few locations and include observations in 1987, 1990 and 1991 from Okinawa

(Heyward et al. 1987, Hayashibara et al. 1993, Kinzie 1993), and partially published observations in 2007 from the Philippines (Vicentuan et al. 2008; Guest et al. unpublished data). We included Red Sea spawning observations for 2011 and 2012. For each year and each location, the number of species spawning per month and the families

(or genera in the case of abundant groups) involved were compared. Table 2 Stages of gametogenesis on various dates throughout the spawning season in 2011. Dates sampled were 2 April (April FM-16), 13 April (April FM-5), 11 May (May FM-6), 17 May (May FM), 31 May (May FM+14) 7 June (June FM-8), 25 June (June FM+10), 6 July (July FM-9), 4 October (Oct FM-8) and 26 November (Nov FM+16). Size of oocytes is in m. Sexuality is defined as hermaphroditic (H) or gonochoric (G). Field Stage of Size of Stage of Number Sexuality Species Date sampled Observations Oogenesis Oocytes Spermatogenesis Colonies [H/G] Clade III: Poritidae Goniopora columna May FM-6 - III 300 no 1♀ G Porites monticulosa* April FM-16 - no no 10 G May FM - no no 10 June FM-8 - no no 10 Oct FM-8 - III-IV 74 III-IV 4♀ 5♂ Nov FM+16 - no no 10 P. nodifera* May FM - IV 360 IV 4♀ 3♂ G June FM-8 - III-IV III 1♀ 1♂ Oct FM-8 - no - no 8 Nov FM+16 - no - no 8

Clade VI: Acroporidae Astreopora listeri May FM-6 - II-III 185 - 1 H A. myriophthalma May FM-6 - - - III 1 H May FM+14 big cream eggs IV 500 IV 2 June FM+10 - no - - 1 Montipora efflorescens May FM-6 - III 280 II 1 H M. tuberculosa May FM-6 - no - no 1 H June FM+10 - II 75 - 1 July FM-9 - no - no 1

Clade VII: Agariciidae Pavona varians April FM-5 - IV 250+ no 1♀ G

Clade XI Coscinaraea monile* April FM-5 - IV 240 - 1 n/a July FM-9 - III-IV 180 - 1

Clade XIV 69

Blastomussa loyae* June FM+10 - III-IV 320 - 1 n/a

Clade XV: Diploastreidae Diploastrea heliopora May FM-6 - III 250 - 1 G May FM+14 - no - no 1 June FM+10 - no - no 1

Clade XVII: Merulinidae Cyphastrea chalcidicum May FM+14 - IV 265 IV 1 H July FM-9 - II 110 - 1 C. kausti* May FM-6 - III 230 II-III 1 H C. serailia May FM-6 - III 200 - 1 H Echinopora forskaliana April FM-5 - II 140 - 1 H May FM-6 - III-IV 230 - 1 June FM+10 - III 200 - 1 E. gemmacea April FM-5 - IV 400 - 1 H May FM+14 - III 200 II-III 2 July FM-9 - III 160 II-III 1 E. hirsutissima* June FM+10 - IV 250 IV 1 H Favia albidus* July FM-9 - III 200 - 1 n/a F. helianthoides May FM-6 - III 305 III 1 H F. maritima* June FM+10 - III 300 - 1 n/a F. rotundata* May FM+14 - III 320 - 1 n/a July FM-9 - no - no 1 Favites abdita May FM-6 red eggs IV 300 IV 2 H July FM-9 - no - no 1 F. paraflexuosa* May FM-6 - II-III 200 - 1 n/a Goniastrea aspera June FM+10 - II-III 160 - 1 H G. pectinata May FM+14 red eggs III 190 III 1 H July FM-9 - III-IV 280 III 1 Favia stelligera June FM+10 - III 200 - 1 H Leptoria phrygia* April FM-5 light green eggs III 215 III 1 H Montastrea cf. curta* May FM-6 - III-IV 200+ - 1 n/a Oulophyllia bennettae April FM-5 - II-III 110 - 1 H Oulophyllia crispa June FM+10 - IV 250 IV 1 H July FM-9 - no** - no 1 Platygyra acuta May FM+14 pink eggs III-IV 235 - 1 H 70

Platygyra crosslandi April FM-5 pink eggs - - - 1 H Platygyra sinensis May FM-6 - II 80+ 1 H Hydnophora microconos April FM-5 - III 200+ II 1 H May FM+14 - no 1 no 1 June FM+10 - IV 200+ IV 1

Clade XIX: Lobophylliidae Lobophyllia corymbosa April FM-5 red eggs - - - 1 H Symphyllia erythraea* May FM+14 - no - no 1 n/a July FM-9 - III-IV 220 - 1 Echinophyllia aspera June FM+10 - III 200 III 1 H *new records **stretched empty mesenteries

Table 3 List of corals recorded to spawn in 2011, 2012 and 2013. All species were observed to spawn directly in the field. Spawning day: day(s) before (-) or after the full moon; NM, new moon. Number colonies: number of colonies observed to participate in the spawning per day. Setting duration: h:min; no, no setting observed before gamete release; yes, setting observed but duration not recorded; n/a, data not available. Form of release of gametes: B, egg-sperm bundles; S, sperm; E, eggs. Sexuality: H, hermaphroditic; G, gonochoric; GD, gynodioecious; PD, polygamodioecious. Form of Spawning Spawning Setting Species Hour of spawning Number colonies release of Sexuality Month Day duration gametes Dendrophylliidae Turbinaria stellulata* June 2012 5 20:20 1 no E G

Poritidae Porites columnaris* May 2012 0, 1 21:45-22:00 (S) 1♂, 3♂ no S+E G P. lobata May 2012 1 21:35-23:15 (S) 1 no S+E G P. lutea May 2012 0, 1, 2, 3 21:15-23:15 (S) 4♂2♀, 12♂6♀, 10♂11♀, no S+E G 22:05-23:20 (E) 3♂4♀ April 2013 2, 3 4♂, 2♂ P. solida* June 2012 4, 5 20:40-21:00 2, 2 no B G

Euphylliidae Galaxea fascicularis June 2012 6, 7 20:05-21:20 2, 2 no S+B GD July 2012 -1 1 Acroporidae Acropora downingi April 2012 0 22:30-22:35 7 2:30 B H April 2013 4 2 A. anthocercis April 2011 -2 22:40-23:10 1 3:00 B H April 2012 0, 1 2, 1 A. eurystoma May 2012 3 19:20 2 yes B H A. gemmifera April 2011 -2 22:45-23:30 12 2:45 B H April 2012 0, 1 3, 1 A. hemprichii April 2012 0, 1 22:40 2, 1 2:10 B H May 2012 3 1 April 2013 4 2 A. humilis June 2011 2 21:10-21:40 21 1:40 B H May 2012 3 4 June 2012 2, 3, 5, 7 2, 1, 1, 1 A. hyacinthus April 2011 -2 22:45-23:45 12 1:40 B H A. lamarcki April 2011 -2 22:40-23:10 5 2:40 B H April 2012 0 6 April 2013 0 2 A. loripes April 2013 4 22:30-22:50 3 1:10+ B H A. lutkeni May 2011 0, 1 21:00-22:15 7, 4 2:00 B H June 2011 2 1 April 2012 0, 1 1, 1 May 2012 3 5 A. maryae April 2012 0 (between 19:50 1 yes B H and 20:40) A. microclados April 2011 -2 22:50-23:00 2 B H May 2011 0 1 A. parapharaonis April 2011 -2 22:30 1 0:45+ B H April 2013 4 1 A. pharaonis April 2011 -2 22:25-23:00 2 2:25 B H April 2012 0 3 April 2013 0, 4 1, 2 A. plantaginea April 2011 -2 22:30-22:45 3 2:50 B H May 2011 0 2 72

April 2012 0, 1, 3 5, 1, 1 April 2013 -3, 0, 4 1, 2, 1 A. polystoma April 2011 -2 (after 22:15) 3 1:30+ B H A. samoensis April 2011 -2 22:45-22:55 2 2:15 B H April 2012 0, 1 4, 1 A. secale May 2012 3 21:50 1 B H A. selago April 2011 -2 (after 22:10) 1 B H A. squarrosa April 2012 0 22:40-22:50 2 2:00+ B H A. valida April 2011 -2 22:00-22:40 1 2:20 B H May 2011 0, 1 12, 2 April 2012 0, 1 4, 1 May 2012 3 4 April 2013 0 2 A. variolosa April 2011 -2 22:45 1 2:00 B H April 2012 0, 1 3, 1 April 2013 0, 4 2, 6 A. verweyi May 2011 0 20:20-20:30 1 yes B H Montipora efflorescens June 2012 1, 2 20:45-21:25 2, 1 0:30 B H July 2012 -3, 0, 1, 2 1, 1, 3, 6 M. tuberculosa June 2011 2 20:20-20:25 1 n/a B H M. turgescens June 2012 1 20:05 1 yes B H

Pocilloporidae Pocillopora verrucosa May 2011 NM-1 8:40-9:15 20+ no S+E G May 2012 NM-1 8:55-9:20 15+ Diploastreidae Diploastrea heliopora May 2011 4 22:20-22:35 (S) + 4♂1♀ no S+E PD June 2012 3 23:00 (E) 1 Merulinidae Cyphastrea May 2012 1, 2 20:50-21:05 3, 2 0:10 B H microphthalma Echinopora May 2011 5 22:25-23:30 1 no B H hirsutissima* April 2012 2 15+ June 2012 2, 3, 4 1, 1, 1 April 2013 2, 3 3, 2 Favia matthaii July 2012 2 22:20 1 no B H F. speciosa June 2012 2, 3 20:40-20:45 1, 1 no B H 73

F. veroni June 2012 2 22:30 1 no B H Favites abdita April 2013 1, 2 23:00-23:10 1, 1 no B H F. paraflexuosa* June 2012 2 22:55 1 no B H F. pentagona June 2012 5 20:35 1 no B H F. spinosa* June 2012 3 21:30 1 no B H Goniastrea edwardsi May 2011, 4, 5 21:30-23:15 10+, 10+ 0:30 B H May 2012 4 1 June 2012 4, 5 5+, 1 July 2012 -3 1 G. retiformis April 2012 2, 3 22:15-23:20 30+, 7+ 0:20 B H May 2012 1, 2, 4 22:15-22:50 22+, 1, 3 May 2013 1, 2 10+, 10+ Favia stelligera May 2011 5 22:00-23:00 5+ 0:30 B H April 2012 6 5+ June 2012 4, 5, 6, 7 2, 10+, 10+, 4 Leptoria phrygia June 2012 4 22:20 1 no B H Platygyra sinensis April 2012 6 20:20 1 n/a B H

Lobophylliidae Acanthastrea brevis* June 2012 5, 6, 7 21:05-21:35 5+, 1, 1 no B H Echinophyllia aspera June 2012 5 21:10 1 no B H Oxypora crassipinosa* June 2012 5 21:15 1 no B H * new records

Results

Spawning times based on histology and in situ examination of gamete maturity

Histological and in situ examinations of the stage of maturity of gametes of 37 coral species from 8 clades in spring 2011 suggest a reproductive season that extends from

April to July (Table 2). Inferred from the histological observation of stage IV gametes

(Table 1 and Fig. 3, 4) or in situ observation of pigmented gametes in the field, Pavona varians, Platygyra crosslandi, Lobophyllia corymbosa as well as some colonies of

Coscinarea monile and Echinopora gemmacea would have spawned on around the April full moon. Similarly, Echinopora forskaliana, Favites abdita, Montastrea cf. curta and most colonies of Porites nodifera (71%) would have spawned on or around the May full moon. Astreopora myriophthalma, Cyphastrea chalcidicum, Platygyra acuta, some colonies of Goniastrea pectinata, and a small proportion of Porites nodifera (29%) would have spawned on or around the June full moon. Blastomussa loyae, Echinopora hirsutissima, Oulophyllia crispa, Symphyllia erythraea as well as some colonies of

Coscinarea monile, Goniastrea pectinata and Hydnophora microconos, would have spawned on or around the July full moon. Finally, the absence of gametes in ten colonies of Porites monticulosa monitored monthly from April to June indicates this species spawns in a different season. Immature gametes were observed in October and were absent after the November full moon, suggesting P. monticulosa would have spawned on or around the November full moon.

a a

b d

i j

Figure 3 Stages of gametogenesis of some scleractinian corals throughout the reproductive season from April to July 2011 a Stage III-IV oogenesis in Coscinarea monile on 6th July b Stage

III-IV oogenesis of Blastomussa loyae on 25th June c-d Oulophyllia crispa on 25th June showing c stage IV gametogenesis and d detail of a spermary with visible spermatozoa tails e Stage II-III 76 oogenesis in Favites paraflexuosa on 11th May f stage II-III gametogenesis of Cyphastrea kausti on 11th May g-h stage IV gametogenesis in Cyphastrea chalcidicum on 31st May showing g a tight aggregation of oocytes and spermaries already forming bundles with stage II oocytes also present and h detail of a spermary with visible spermatozoa tails i-j Echinopora hirsutissima on

25th June showing i stage IV gametogenesis with a tight aggregation of mature gonads already forming bundles and j detail of spermary with visible spermatozoa tails

Inferred from the observation of earlier stages of gamete maturity, stages II and III (Table

1) followed in some cases by the absence of gametes, Diploastrea heliopora, Oulophyllia benettae, Leptoria phrygia as well as some colonies of Hydnophora micronos would have spawned on around the May full moon. Favia helianthoides would have spawned on or around the May or June full moon. Astreopora listeri, Cyphastrea kausti, C. serailia,

Favia rotundata, Favites paraflexuosa as well as some colonies of Montipora tuberculosa and Echinopora gemmacea would have spawned on around the June full moon. Platygyra sinensis and Montipora efflorescens would have spawned on around the

June or the July full moon. Cyphastrea chalcidicum, Echinophyllia aspera, Echinopora forskaliana, Favia albidus, F. maritima, Goniastrea aspera, Favia stelligera as well as some colonies of Montipora tuberculosa and Echinopora gemmacea would have spawned on around the July full moon.

a b

c d

Figure 4 a-b Porites monticulosa on 4th October showing a stage II oogenesis and b stage II-III spermatogenesis; c-d P. nodifera on 17th May showing c stage III-IV oogenesis and d stage IV spermatogenesis with visible spermatozoa tails

Spawning times based on direct observation of gamete release

Over the three reproductive seasons, 22 species from 4 clades spawned in 2011, 42 species from 8 clades were observed to spawn in the field in 2012, and 13 species from 3 clades spawned in 2013 (April only) totaling 52 species from 8 clades and including 8 species not previously observed to spawn (Fig. 5; Table 3). Spawning was observed in each month from April to June in 2011 and from April to July in 2012, and over a period of one to seven consecutive days within each month (Fig. 2). Numbers of species spawning at the same time on any one night ranged from one species on the 31st May

2011 and the 4th July 2012, to a maximum of 13 species on the 16th April 2011 and the 6th

April 2012. 78

In 2011, 13 Acropora species spawned together two nights before the full moon in April

2011, 5 Acropora species, 3 Merulinidae and Diploastrea heliopora spawned individually during a four-night period in May and finally 2 Acropora species and

Montipora tuberculosa spawned together two nights after the full moon in June.

Pocillopora verrucosa spawned in the morning on the day before the June new moon.

In 2012, 13 Acropora species and 4 Merulinidae spawned over five nights in April, 6

Acropora species, 3 Merulinidae and 3 Porites species spawned during a period of five consecutive nights in May. Acropora humilis, 2 Montipora species, Diploastrea heliopora, 9 Merulinidae, Turbinaria stellulata, Porites solida, Galaxea fascicularis and

3 Lobophylliidae spawned over a period of seven consecutive nights in June. Montipora efflorescens, 2 Merulinidae and Galaxea fascicularis spawned over 5 nights in July.

Pocillopora verrucosa spawned again in the morning of the day preceding the May new moon. In April 2013, 10 Acropora species, 3 Merulinidae and Porites lutea spawned over a period of four nights.

a b c

e d

h i

Figure 5 a setting of Acropora hemprichii b setting and c spawning of Goniastrea retiformis d spawning of Acropora plantaginea e Diploastrea heliopora releasing sperm f Galaxea fascicularis releasing a white egg-sperm bundle g Porites cf poritipora releasing egg-sperm 80 bundles h Turbinaria stellulata releasing yellow eggs I Acanthastrea brevis releasing pink egg- sperm bundles

Day in the lunar month of spawning

Spawning was observed as early as three nights before the full moon (e.g. July 2012) until at the latest seven nights after the full moon (e.g. June 2012), and Pocillopora verrucosa spawned one day before the new moon in 2011 as well as in 2012. In 2011 and

2012 the night with the most species spawning, was on the first night Acropora species spawned, 2 nights before the April full moon in 2011 and on the night of the April full moon in 2012. In 2013, the peak of Acropora spawning was over 2 nights with 6 species on the night of the full moon and 8 species four nights after the full moon. In 2011, only

6 species other than Acropora species were observed to spawn. However reproductive data from histology and from in situ examinations of gamete maturity suggested that many species spawn between April and July. Therefore, in 2012, we doubled the hours spent each night and increased the number of nights around each full moon, surveying corals for spawning behaviour and observed the spawning of an additional 23 species.

In May 2012, the peak of spawning was on the second and third nights after the full moon, with mostly 5 Porites species and 2 Merulinidae spawning on the second night followed the next night by 6 Acropora species and Porites lutea spawning. In June 2012 the peak of spawning was five nights after the full moon, with Acropora humilis , 3

Merulinidae, Turbinaria stellulata, Porites solida and 3 Lobophylliidae spawning at different times. In July 2012 no more than two species spawned on each night. 81

Day and hour of spawning of Acropora species

Throughout the reproductive season, 15 Acropora species spawned in 2011 and 16 species spawned in 2012. All Acropora species released egg/sperm bundles. Spawning times were consistent in 2011 and 2012. In 2011, on the 16th April (2 nights before full moon), 13 out of 16 Acropora species (57 colonies) spawned between 22:00 and 23:45 after a 2-3 hour setting period, except for Acropora verweyi which spawned earlier between 20:20 and 20:30. In 2012, on the 6th April (full moon night), 13 out of 20

Acropora species (64 colonies) spawned synchronously again except for Acropora eurystoma and A. maryae which spawned respectively around 19:20 and between 19:50 and 20:40 (sunset: 19:00). Both years, a few colonies from species which had already spawned in April (i.e. A. hemprichii, A. lutkeni, A. microclados, A. plantaginea, A. samoensis and A. valida) as well as from additional species (i.e. A. eurystoma, A. humilis, A. secale and A. verweyi) spawned in May between 21:00 and 22:00 and finally only 1 colony of A. lutkeni and A. humilis (21 colonies in 2011 and 2 colonies in 2012) spawned in June between 21:00 and 22:00.

Hour of spawning and spawning behaviour of other species from the coral assemblage

Except for Pocillopora verrucosa which spawned around 9:00 in the morning, all the coral colonies spawned in the five hours following sunset, between 19:20 and 23:45

(Table 3). Turbinaria stellulata released eggs for a short period around 20:20 in June.

Male and female Porites colonies slowly released sperm or eggs starting between 21:35 82 and 23:25. Sperm clouds were thick and dense and visible for up to 2 to 3 hours for most massive colonies. In June 2012, two neighbouring colonies of Porites solida were observed to release bundles between 20:40 and 21:00. Five Galaxea fascicularis colonies independently released white bundles at a slow pace of one bundle every 5-10 minutes, including one colony which also intermittently released separate jets of sperm, from

20:05-21:20 on three different nights in June and July 2012. Montipora species spawned in June 2011 and 2012 as well as in July 2012 from 20:05 to 21:25 after setting for at least 30 minutes. Diploastrea heliopora colonies released sperm from 22:20-22:35 and another colony released eggs around 23:00. Favia stelligera, Goniastrea edwardsi and G. retiformis all had a 20-30 minute setting period and spawned between 21:30 and 23:20, with most colonies within each species spawning synchronously (e.g. up to 30 colonies of

G. retiformis in April 2012). Three colonies of Cyphastrea microphthalma spawned in

May 2012 from 20:50 to 21:05 after at least ten minutes of setting. No other Merulinidae were observed to set before releasing gametes and typically the egg/sperm bundles were released simultaneously within a given colony. Favia species (except for F. stelligera) spawned in June and July: F. speciosa released light green bundles on consecutive nights around 20:45, and F. matthaii spawned at 22:20 followed by F. veroni at 22:30. Except for Favites abdita which spawned at 23:00 in April 2013, all Favites species all spawned in June: F. pentagona at 20:35 F spinosa at 21:30 and finally F. paraflexuosa at 22:55.

Finally Platygyra sinensis spawned at 20:20, Leptoria phrygia at 22:20 and over 15 colonies of Echinopora hirsutissima spawned from 22:25 to 23:30 although for each colony gametes were released simultaneously and no setting was observed. Acanthastrea brevis released pink egg/sperm bundles from 21:05 to 21:35. Echinophyllia aspera and 83

Oxypora crassipinosa were both released yellow egg-sperm bundles on the same night between 21:10 and 21:15.

Multiple spawning of single colonies

Some tagged colonies of Acropora and Montipora spawned more than once. One colony of Acropora gemmifera and one colony of A. variolosa spawned on 2 consecutive nights in April 2012. Two colonies of A. lutkeni and A. valida spawned 3 times, on 2 consecutive nights in April 2012 and again a month later in May 2012. One colony of

Montipora efflorescens spawned on 2 consecutive nights in July 2012. Two colonies of

Porites solida spawned over two consecutive nights in June 2012.

New spawning records

The following species were observed to spawn for the first time: Turbinaria stellulata was observed with one colony releasing small yellow eggs (Fig. 2G); Porites columnaris, released sperm and P. solida released purple-pink bundles; Acropora variolosa released pink egg/sperm bundles; the merulinids, Echinopora hirsutissima, Favites paraflexuosa and F. spinosa all released pink egg/sperm bundles; finally, two species from the family

Lobophylliidae, Acanthastrea brevis released pink egg/sperm bundles, and Oxypora crassipinosa released yellow egg/sperm bundles.

Comparison between the central Red Sea and other locations

Spawning observations were made over three to four months in the central Red Sea, three months in the Philippines and three to four months in Okinawa (Fig. 6). In all three 84 locations, spawning in the first months of the reproductive season was dominated by

Acropora species. Furthermore it is on the first spawning night of the reproductive season that most of the Acropora species spawning, as observed in the central Red Sea and in

Okinawa. In the Philippines only Galaxea fascicularis was seen to spawn in the first month; however reproductive surveys on the synchrony of maturation of gametes suggest that most Acropora species would have also spawned then (Guest et al., unpublished data). Montipora species spawn towards the end of the reproductive season in the Red

Sea and in the Philippines but throughout the entire reproductive season in Okinawa

(except for 1990). Porites species all spawn in May in the central Red Sea (except for two neighbouring colonies of Porites solida spawning in June), and spawn either in June

(1991) or in July (1990) in Okinawa. Merulinidae spawn throughout the entire reproductive season in all locations. Lobophylliidae spawn at the end of the reproductive season in the central Red Sea and in the Philippines but at the beginning of the reproductive season in Okinawa and in south-west Japan. Euphylliidae, represented in each location by Galaxea fascicularis, seem to spawn several times in the reproductive season in the central Red Sea, the Philippines and Okinawa. Within each lunar month, the length of the spawning period was shortest in the Philippines where species spawned over a maximum of four consecutive nights. In the central Red Sea, spawning lasted seven consecutive nights within a month and in Okinawa ten consecutive nights.

85 a. Central Red Sea 2011 b. Central Red Sea 2012 c. Philippines 2007 30 30 30 25 25 25 20 20 20 15 15 15 10 10 10 5 5 5 0 0 0 April May June April May June July April May June d. Okinawa 1987 e. Okinawa 1990 f. Okinawa 1991 30 30 30 25 25 25 20 20 20 15 15 15 10 10 10 5 5 5 0 0 0 June July August

Figure 6 Number of species observed to spawn per month for

different genera or clades. a Central Red Sea 2011 b Central Red

Sea 2012 c Philippines 2007 d Okinawa 1987 e Okinawa 1990 f

Okinawa 1991

Spawning of other marine organisms

In May 2012, a hermit crab and two different polychaete worms released gametes three nights after the full moon. In June 2012, five nights after the full moon, one octocoral

Heteroxenia sp released white planulae, six bivalves released clouds of eggs or sperm and a Diadema sea urchin released a cloud of sperm Another Diadema released a cloud of sperm seven nights after the full moon. Finally in July 2012, two nights before the full moon, the basket star Astroboa nuda released sperm. 86

Sea surface temperatures

Throughout this study (except March to November 2011 when no data was recorded), water temperatures in central Red Sea varied between 24.7°C and 33.6°C. Most of the water temperature increase happens from April to July, when waters warmed from 24.7 to 33.5 in 2012 and from 26.1 to 32.3 in 2013 (Fig. 7).

Figure 7 Hourly water temperatures from April to July in 2011, 2012 and 2013.

Discussion

Over three reproductive seasons, 52 species from 8 clades were directly observed to spawn on the reef. Additionally, the month of spawning could be inferred from the presence of mature gametes for an additional 17 species. These observations include novel data on reproductive biology of 20 species from 6 clades. The spawning season in the central Red Sea happens in the spring from April to July, and coincides with rapidly rising sea surface temperatures.

Some phylogenetic trends were evident in the time of spawning. Porites colonies all spawned in May in 2012 (4th lunar month) and in April in 2013 (4th lunar month), most

Acropora colonies spawned in April, Montipora colonies spawned in June and July, species from the family Lobophylliidae all spawned in June, whereas Merulinidae spawned throughout the entire reproductive season. Most coral species spawn at a predictable time each year (Willis et al. 1985). The month of spawning in the central Red

Sea was consistent between 2011, 2012 and 2013, although some taxa appeared shifted by one solar month. For example Diploastrea heliopora spawned in May in 2011 (5th lunar month) and in June in 2012 (5th lunar months). In contrast, Acropora humilis, which spawned in June in 2011 (6th lunar month), spawned in May as well as in June in 2012 (4-

5th lunar month). Additionally Acropora lutkeni spawned in May and June in 2011 (5-6th lunar month) but in April and May in 2012 (3-4th lunar month). In some years, depending on when the moon cycle falls, some spawning can be shifted to the next lunar cycle or split over two consecutive moon cycles (Baird et al. 2009). There is a possibility that the

Red Sea species observed to spawn in different months in 2011 and 2012 spawned more 88 than once and that the second spawning was not observed. In 2012, the full moon was eleven to twelve days earlier in the month and could have caused some species to spawn around the following full moon. This explains why Diploastrea heliopora appears to have spawned a month later in 2012. In April 2013, spawning observations indicated that on different reefs, corals seem to spawn within a few days of each other. Furthermore, some common Acropora species (e.g. Acropora gemmifera) did not contain any eggs on different reefs, four days before the full moon (pers. obs.) indicating they had either spawned earlier that month or with the March full moon. Some species that had spawned in May 2012 such as Porites species and some Merulinidae, spawned in April in 2013.

Possibly in the central Red Sea, more factors than just the timing of the moon cycle are responsible for determining the spawning months of the coral assemblage. Similarly, the nights of spawning in the central Red Sea might not be as predictable from one year to the next as indicated by the spawning observations of Acropora in 2013. For example,

Acropora humilis spawned between two and seven nights after the full moon depending on the month. None of the other species observed to spawn over at least two months or observed over consecutive years (e.g. Goniastrea edwardsii, Echinopora hirsutissima,

Favia stelligera) spawned on the same night before or after the full moon.

Traditionally, coral species have been described according to differences in the morphology of the colony and macromorphological structure of the corallite (e.g. Wells

1956; Chevalier and Beauvais 1987; Wallace 1999; Veron 2000). However, scleractinian corals show high plasticity in their morphology in reaction to environmental conditions such as wave exposure and depth (Hoeksema and Moka 1989; Todd 2008). Since the use 89 of molecular phylogenetics tools in the past decade, recent taxonomic work has completely changed the phylogeny of scleractinian corals. Genera have since then been assigned to new clades grouping different genera and species than were found in the former scleractinian families (Fukami et al. 2008; Huang et al. 2011; Arrigoni et al.

2012). Recent taxonomy revisions of the new Merulinidae family have placed Favia stelligera in a subclade with Goniastrea species and suggest that the genus should be changed to Goniastrea (Huang et al. 2011, Budd et al. 2012). The spawning behaviour of

Favia stelligera in this study supports the suggestion as, except for Cyphastrea that had a ten minute setting period, it was the only non-Goniastrea species of the Merulinidae that underwent a long setting behaviour, while none of the three other Favia species set before they spawned. Goniastrea edwardsii, G. retiformis and Favia stelligera also spawned in the same hours of the night, during a 1:30 window during which they were observed to spawn at the same time on several occasions.

The vast majority of Porites species are gonochoric, although a few populations have low levels of hermaphroditic colonies (Baird et al. 2009). The Porites species observed to spawn in the central Red Sea released either sperm or eggs, suggesting that they are gonochoric. However, one Porites species, P. solida, spawned a month later and was observed to release bundles. This is the first record of a species in the family Poritidae releasing bundles. It is not clear whether the bundles contained only eggs or contained sperm as well. Further observations or histological work are necessary to identify the sexuality of this species in the Red Sea. Porites monticulosa spawns around November.

This is the only species in the Red Sea to date which spawn out of the main spawning season from April to July. This is also the first records of spawning for the species 90 although it is widely distributed in the Indo-Pacific (Veron 2000). Some Porites species in other parts of the world spawn out of the main breeding season such as in the

Caribbean, on the GBR, in Western Australia and in Zanzibar (Bronstein and Loya 2011,

Stoddart et al. 2012).

Many species from additional clades still remain to be observed releasing their gametes.

In shallow environments we are still missing observations from Agariciidae such as

Pavona, Siderastreidae, Fungiidae and other species from small clades such as Plerogyra.

Some shallow water Pocilloporidae species such as Stylophora and Seriatopora species are brooders, but four spawning Pocillopora species on the Great Barrier Reef and

Pocillopora verrucosa in the central Red Sea were observed to spawn in the morning and around the new moon (Bouwmeester et al. 2011b / see annexe 1, p. 138; Schmidt-Roach et al. 2012). Possibly some of the missing observations from the coral assemblage are due to the fact that the spawning happens later at night or during the daytime as has been observed for Pavona gigantea in the Galapagos Islands, which have been observed to release their gametes in the late-afternoon (Glynn et al. 1996) or for fungiids in Okinawa,

Japan which have been observed to release gametes several hours after sunset, starting around 22:00 with species spawning mainly between 22:30 and 23:30 or between 01:00 and 02:00 or for other fungiid species around 03:00 or 05:00 (Loya et al. 2009).

Furthermore, surveying different reef habitats such as inshore reefs with different coral assemblages might uncover the spawning behaviour for species which are uncommon on most offshore reefs, such as agariciids and lobophylliids.

Observations in the central Red Sea are consistent with the few observations in other regions of the Red Sea apart from the Gulf of Aqaba. Surveys conducted in Hurghada and

Marsa Alam in the north-western Red Sea suggest that the peak spawning event for

Acropora species occurs between mid-April and the first week of May (Hanafy et al.

2010). Spawning of scleractinian corals in Eilat follows a very different pattern with no synchrony between species which release gametes throughout May to December

(Shlesinger and Loya 1985; Shlesinger et al. 1998). It is expected that the patterns observed in the central Red Sea might extend to the southern Red Sea. Indeed, in the

Farasan Islands (16°40’N, 42°00’E), coral spawn slicks have been reported near the time of the full moon, once in March but usually in April (Gladstone 1996). In the Gulf region, located approximately 1500km east of the Red Sea, spawning observations and reproductive surveys suggest a reproductive period of two months, in April and May

(Fadlallah 1996, Bauman et al. 2011). The Red Sea and the Gulf display similar reproductive patterns with spawning happening in the same months although the length of the reproductive season might be longer in the central Red Sea.

Corals on reefs in Eilat in the north of the Gulf of Aqaba, Red Sea, seem to show temporal reproductive isolation (Shlesinger and Loya 1985, Shlesinger et al. 1998).

Spawning patterns in the central Red Sea however show the opposite: corals reproduce during a few consecutive nights around the full moon during which up to nineteen species are involved in a single month, over a period of four months. These patterns are similar to those of other locations in the world such as the Philippines or Okinawa, Japan.

Comparisons with spawning observations in the above locations have shown that in each 92 of the localities a variety of species from different clades spawned during the surveyed months. However the order in genera or clades spawning was not consistent from one region to the next, except perhaps for the Acropora species which dominate the first spawning day(s) as well as month(s) in all localities. Although the first mass spawning event or GBR multi-specific spawning event has been described as the spawning of several colonies from several species in a single night (Willis et al. 1985), spawning over three to four months is a common spawning pattern worldwide which has already been shown in numerous regions such as French Polynesia, Papua New Guinea, Singapore and the GBR in Australia (Baird et al. 2009a, 2009b). Many regions worldwide have spawning data for one or two lunar cycles but most probably, in some locations, further observations or reproductive surveys might reveal a longer reproductive season than originally thought.

The present study provides the first records of multi-specific spawning in the Red Sea and presents spawning data for over fifty species from several clades. High synchrony in the spawning of scleractinian corals and knowing the timing of the spawning are important advantages to promote management and conservation measures (Guest et al.

2008; Hanafy et al. 2010). Indeed, restrictions on potential anthropogenic impacts can be put in place during the spawning periods to ensure successful fertilization of gametes and successful dispersal and settlement of the larvae once they have developed motility. The reproductive season of scleractinian corals in the central Red Sea goes from April to July, with most spawning centered on around the full moon days. This timing should be taken into consideration when planning coral reef protection measures on reefs as well as when 93 planning development projects on the coast of Saudi Arabia or of other neighbouring countries which may share similar coral reproduction patterns.

Acknowledgements

Thank you to all the volunteers from the KAUST community and the Red Sea Research

Center who joined in the night surveys and helped collect data during spawning events.

Thanks to F Benzoni and D Huang for assisting in some of the identification of scleractinian corals and Y Benayahu for confirming the planulae release and for identifying Heteroxenia sp. Spawning data from localities outside the central Red Sea was collected from the literature and compiled in a worldwide spawning database by JR

Guest. This work was conducted as part of the PhD thesis of J Bouwmeester in the King

Abdullah University of Science and Technology (KAUST).

References

Arrigoni R, Stefani F, Pichon M, Galli P, Benzoni F (2012) Molecular phylogeny of the Robust clade (Faviidae, Mussidae, Merulinidae, and Pectiniidae): An Indian Ocean perspective. Molecular Phylogenetics and Evolution 65:183-193 Babcock RC, Wills BL, Simpson CJ (1994) Mass spawning of corals on a high latitude coral reef. Coral Reefs 13:161-169 Babcock RC, Bull GD, Harrison PL, Heyward AJ, Oliver JK, Wallace CC, Willis BL (1986) Synchronous spawnings of 105 scleractinian coral species on the Great Barrier Reef. Mar Biol 90:379-394 Baird AH, Marshall PA, Wolstenholme JK (2002) Latitudinal variation in the reproduction of Acropora in the Coral Sea. 9th International Coral Reef Symposium 1:385-389 Baird AH, Guest JR, Willis BL (2009a) Systematic and Biogeographical Patterns in the Reproductive Biology of Scleractinian Corals. Annual Review of Ecology, Evolution, and Systematics 40:551-571 Baird AH, Kospartov MC, Purcell S (2010) Reproductive Synchrony in Acropora Assemblages on Reefs of New Caledonia1. Pacific Science 64:405-412 94

Baird AH, Blakeway DR, Hurley TJ, Stoddart JA (2011) Seasonality of coral reproduction in the Dampier Archipelago, northern Western Australia. Mar Biol 158:275-285 Baird AH, Birrel CL, Hughes TP, McDonald A, Nojima S, Page CA, Prachett MS, Yamasaki H (2009b) Latitudinal variation in reproductive synchrony in Acropora assemblages: Japan vs. Australia. Galaxea, Journal of Coral Reef Studies 11:101- 108 Baria MVB, Villanueva RD, Guest JR (2012) Spawning of three-year-old Acropora millepora corals reared from larvae in northwestern Philippines. Bulletin of marine science 88:61-62 Bastidas C, Cróquer A, Zubillaga AL, Ramos R, Kortnik V, Weinberger C, Márquez LM (2005) Coral mass- and split-spawning at a coastal and an offshore Venezuelan reefs, southern Caribbean. Hydrobiologia 541:101-106 Bauman A, Baird A, Cavalcante G (2011) Coral reproduction in the world’s warmest reefs: southern Persian Gulf (Dubai, United Arab Emirates). Coral Reefs 30:405- 413 Berumen ML, Hoey AS, Bass WH, Bouwmeester J, Catania D, Cochran JEM, Khalil MT, Miyake S, Mughal MR, Spaet JLY, Saenz-Agudelo P (2013) The status of coral reef ecology research in the Red Sea. Coral reefs:1-12 Bouwmeester J, Khalil MT, De La Torre P, Berumen ML (2011a) Synchronous spawning of Acropora in the Red Sea. Coral Reefs 30:1011-1011 Bouwmeester J, Berumen ML, Baird AH (2011b) Daytime broadcast spawning of Pocillopora verrucosa on coral reefs of the central Red Sea. Galaxea, Journal of Coral Reef Studies 13:23-24 Bronstein O, Loya Y (2011) Daytime spawning of Porites rus on the coral reefs of Chumbe Island in Zanzibar, Western Indian Ocean (WIO). Coral Reefs:1-1 Chevalier JP, Beauvais L (1987) Classifications en sous-ordres des Scléractiniaires. Traité de Zoologie, 3:679-764 Connolly SR, Baird AH (2010) Estimating dispersal potential for marine larvae: dynamic models applied to scleractinian corals. Ecology 91:3572-3583 Dai CF, Soong K, Fan TY (1992) Sexual reproduction of corals in northern and southern Taiwan. Proceedings of the Seventh International Coral Reef Symposium 1:22-27 Fadlallah YH (1996) Synchronous spawning of Acropora clathrata coral colonies from the western Arabian Gulf (Saudi Arabia). Bulletin of marine science 59:209-216 Figueiredo J, Baird AH, Connolly SR (2012) Synthesizing larval competence dynamics and reef-scale retention reveals a high potential for self-recruitment in corals. Ecology 94:650-659 Fukami H, Omori M, Shimoike K, Hayashibara T, Hatta M (2003) Ecological and genetic aspects of reproductive isolation by different spawning times in Acropora corals. Mar Biol 142:679-684 Fukami H, Chen CA, Budd AF, Collins A, Wallace C, Chuang Y-Y, Chen C, Dai C-F, Iwao K, Sheppard C, Knowlton N (2008) Mitochondrial and Nuclear Genes Suggest that Stony Corals Are Monophyletic but Most Families of Stony Corals Are Not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS ONE 3:e3222 95

Gilmour J (1999) Experimental investigation into the effects of suspended sediment on fertilisation, larval survival and settlement in a scleractinian coral. Mar Biol 135:451-462 Gladstone W (1996) Unique annual aggregation of longnose parrotfish (Hipposcarus harid) at Farasan Island (Saudi Arabia, Red Sea). Copeia 1996:483-485 Glynn PW, Colley S, Gassman N, Black K, Cortés J, Maté J (1996) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama, and Galapagos Islands (Ecuador). III. Agariciidae (Pavona gigantea and Gardineroseris planulata). Mar Biol 125:579-601 Graham EM, Baird AH, Connolly SR, Sewell MA, Willis BL (2013) Rapid declines in metabolism explain extended coral larval longevity. Coral reefs 32:539-549 Guest JR, Baird AH, Goh BPL, Chou LM (2005) Reproductive seasonality in an equatorial assemblage of scleractinian corals. Coral Reefs 24:112-116 Guest JR, Baird AH, Clifton KE, Heyward AJ (2008) From molecules to moonbeams: Spawning synchrony in coral reef organisms. Invertebrate Reproduction & Development 51:145-149 Guest JR, Dizon RM, Edwards AJ, Franco C, Gomez ED (2011) How Quickly do Fragments of Coral “Self-Attach” after Transplantation? Restoration Ecology 19:234-242 Hanafy MH, Aamer MA, Habib M, Rouphael AB, Baird AH (2010) Synchronous reproduction of corals in the Red Sea. Coral Reefs 29:119-124 Harrison PL (2011) Sexual Reproduction of Scleractinian Corals. In: Dubinsky Z, Stambler N (eds) Coral Reefs: An Ecosystem in Transition. Springer Netherlands, pp59-85 Harrison P, Wallace C (1990) Reproduction, dispersal and recruitment of scleractinian corals. Dubinsky Z (ed) Ecosystems of the world: coral reefs. Elsevier, Amsterdam, pp133–207 Harrison PL, Babcock RC, Bull GD, Oliver JK, Wallace CC, Willis BL (1984) Mass Spawning in Tropical Reef Corals. Science 223:1186-1189 Hayashibara T, Shimoike K, Kimura T, Hosaka S (1993) Patterns of coral spawning at Akajima Island, Okinawa, Japan. Marine Ecology Progress Series 101:253-253 Heyward A, Yamazato K, Yeemin T, Minei M (1987) Sexual reproduction of corals in Okinawa. Galaxea 6:331-343 Hoeksema BW, Moka W (1989) Species assemblages and phenotypes of mushroom corals (Fungiidae) related to coral reef habitats in the Flores Sea. Netherlands Journal of Sea Research 23:149-160 Huang D, Licuanan WY, Baird AH, Fukami H (2011) Cleaning up the'Bigmessidae': Molecular phylogeny of scleractinian corals from Faviidae, Merulinidae, Pectiniidae and Trachyphylliidae. BMC evolutionary biology 11:37 Hughes TP, Bellwood DR, Connolly SR (2002) Biodiversity hotspots, centres of endemicity, and the conservation of coral reefs. Ecology Letters 5:775-784 Kinzie III RA (1993) Spawning in the reef corals Pocillopora verrucosa and P. eydouxi at Sesoko Island, Okinawa. Galaxea 11:93-105 Kongjandtre N, Ridgway T, Ward S, Hoegh-Guldberg O (2010) Broadcast spawning patterns of Favia species on the inshore reefs of Thailand. Coral reefs 29:227-234 96

Loya Y, Sakai K, Heyward A (2009) Reproductive patterns of fungiid corals in Okinawa, Japan. Galaxea, Journal of Coral Reef Studies 11:119-129 Mangubhai S, Harrison PL (2008) Asynchronous coral spawning patterns on equatorial reefs in Kenya. Marine Ecology Progress Series 360:85-96 Mundy C, Green A (1999) Spawning observations of corals and other invertebrates in American Samoa. Report prepared for Department of Marine and Wildlife Resources, American Samoa Government:1-12 Negri AP, Marshall PA, Heyward AJ (2007) Differing effects of thermal stress on coral fertilization and early embryogenesis in four Indo Pacific species. Coral Reefs 26:759-763 Palmer CV, Graham E, Baird AH (2012) Immunity through early development of coral larvae. Developmental & Comparative Immunology 38:395-399 Raitsos DE, Hoteit I, Prihartato PK, Chronis T, Triantafyllou G, Abualnaja Y (2011) Abrupt warming of the Red Sea. Geophys Res Lett 38:L14601 Richmond RH, Hunter CL (1990) Reproduction and recruitment of corals: Comparisons among the Caribbean, the Tropical Pacific, and the Red Sea. Marine ecology progress series 60:185-203 Roberts CM, McClean CJ, Veron JEN, Hawkins JP, Allen GR, McAllister DE, Mittermeier CG, Schueler FW, Spalding M, Wells F (2002) Marine biodiversity hotspots and conservation priorities for tropical reefs. Science 295:1280-1284 Schmidt-Roach S, Miller KJ, Woolsey E, Gerlach G, Baird AH (2012) Broadcast Spawning by Pocillopora Species on the Great Barrier Reef. PLoS ONE 7:e50847 Shanks AL, Grantham BA, Carr MH (2003) Propagule dispersal distance and the size and spacing of marine reserves. Ecological Applications 13:159-169 Shlesinger Y, Loya Y (1985) Coral community reproductive patterns: Red Sea versus the Great Barrier Reef. Science 228:1333-1335 Shlesinger Y, Goulet T, Loya Y (1998) Reproductive patterns of scleractinian corals in the northern Red Sea. Mar Biol 132:691-701 Steneck R, Paris C, Arnold S, Ablan-Lagman M, Alcala A, Butler M, McCook L, Russ G, Sale P (2009) Thinking and managing outside the box: coalescing connectivity networks to build region-wide resilience in coral reef ecosystems. Coral Reefs 28:367-378 Stoddart CW, Stoddart JA, Blakeway DR (2012) Summer spawning of Porites lutea from north-western Australia. Coral reefs 31:787-792 Tay YC, Guest JR, Chou LM, Todd PA (2011) Vertical distribution and settlement competencies in broadcast spawning coral larvae: Implications for dispersal models. Journal of Experimental Marine Biology and Ecology 409:324-330 Todd PA (2008) Morphological plasticity in scleractinian corals. Biological Reviews 83:315-337 Van Woesik R, Lacharmoise F, Köksal S (2006) Annual cycles of solar insolation predict spawning times of Caribbean corals. Ecology Letters 9:390-398 Veron JEN (2000) Corals of the world. Australian Institute of Marine Science, Townsville Vicentuan K, Guest JR, Baria M, Cabaitan P, Dizon R, Villanueva RD, Aliño P, Edwards A, Gomez E, Heyward AJ (2008) Multi-species spawning of corals in north- western Philippines. Coral Reefs 27:83-83 97

Wallace CC (1999) Staghorn corals of the world: A Revision of the genus Acropora. CSIRO Publishing Wells JW (1956) Scleractinia. Treatise on invertebrate paleontology:328-444 Willis BL, Babcock RC, Harrison PL, Oliver JK, Wallace CC (1985) Patterns in the mass spawning of corals on the Great Barrier Reef from 1981 to 1984. 5th International Coral Reef Symposium 4:343-348 Wilson JR, Harrison PL (2003) Spawning patterns of scleractinian corals at the Solitary Islands-a high latitude coral community in eastern Australia. Marine Ecology Progress Series 260:115-123

SUPPLEMENTARY DATA

Plates of Porites species observed to spawn in the field.

In each plate:

a. field image: colony in its habitat

b. field image: macro of the colony

c. SEM image: corallite arrangement at 50x magnification

d. SEM image: arrangement of two corallites showing detail of the corallite wall at

80x magnification

e. SEM image: detail of a single corallite at 100x magnification with the directive septum at the top of the corallite

f. SEM image: detail inside a corallite at 200x magnification with the directive septum at the top of the corallite

Figure S1 Porites columnaris 100

Figure S2 Porites lobata 101

Figure S3 Porites lutea 102

Figure S4 Porites solida 103

CHAPTER 4: CORAL RECOVERY AND SETTLEMENT AFTER A BLEACHING

EVENT IN THE CENTRAL RED SEA

Abstract

Over the past decades, periods of abnormally high sea surface temperatures have caused coral bleaching: the result of the loss of the coral’s symbiotic algae. During the summer of 2010, high sea surface temperatures on reefs in the central Red Sea, resulting in a mass bleaching event that was followed by high coral mortality rates, especially on inshore reefs. On average at 5-10m depth, 84-100% corals had bleached, resulting in 67-93 % mortality of colonies 7 months later. This study describes the early post-bleaching recovery patterns in the adult coral assemblage, juvenile corals, and recently settled recruits on an inshore reef in the waters of Thuwal, Saudi Arabia from 0.5-3 years after the bleaching. Adult coral cover remained constant at a low base of 4 % on the exposed side and 16 % on the sheltered side throughout the 2.5 years of this study, but numbers of juvenile corals started to increase 18 months after the bleaching event. Although the density of settlers on experimental substrata was low, there was a strong seasonal pattern with a peak in settlement corresponding to the annual reproductive season of scleractinian corals in the region. Recovery on the inshore reefs of Thuwal is still at an early stage, but if coral settlement remains successful over the coming years, and assuming no other disturbances occur, coral assemblages should return to a state similar to that present before the bleaching event.

104

Keywords Scleractinia – Drupella – disturbance – juvenile corals – recovery – recruits – resilience – Saudi Arabia – settlement plates

Introduction

Coral reefs are currently under serious threat from climate change due to ocean acidification and rising sea surface temperatures (Barnett et al. 2001; Hoegh-Guldberg

2007). Coral bleaching events, caused mostly by periods of unusually high sea temperatures, have already had severe impacts on some coral reef systems which, in a few cases, have not returned to their pre-disturbance states (Gardner et al. 2003; Hughes and Tanner 2000; Hughes et al. 2007 but see also Bruno et al. 2009).

Not all bleaching results in the mortality of the affected colonies, because different species have different susceptibility to bleaching (Obura 2005). When mortality does occur, different species have shown varying resilience and recovery patterns (e.g. Loya et al. 2001; Baird and Marshall 2002; van Woesik et al. 2011), but, unfortunately, we know much more about the causes of coral reef collapse than we do about the factors which contribute to their recovery (Vermeij 2006). One of the key recovery processes is coral recruitment which requires the availability of settlement-competent larvae and suitable substrate (Arnold et al. 2010). Bleaching of adult coral colonies, even when not followed by mortality, can reduce their reproductive output for several years with smaller and fewer eggs being produced in fewer corallites, and with lower fertilisation rates (Omori et al. 2001; Baird and Marshall 2002; Ward et al. 2002).

105

The Red Sea is thought to be better adapted to extremes of temperature and salinity than other seas worldwide (Cantin et al. 2010). While reports of bleaching within the Red Sea are rare, the Red Sea is, however, not immune to bleaching events (Berumen et al. 2013 / see annexe 1, p. 155; Fine et al. 2013). Fishelson (1973) reported coral bleaching on reef flats following extreme low tides in Eilat, Gulf of Aqaba. The same phenomenon was reported again in 2007 on reefs on the Egyptian coast of the Gulf of Aqaba (C. Alter, unpublished data). Mass bleaching was reported worldwide in 1997-1998 following the major ENSO climate event (Wilkinson 1999; Goreau et al. 2000; Lough 2000), and, although bleaching was not as extensive as reported in other locations in the world, patchily distributed bleaching was recorded on the Saudi Arabian coast from the central

Red Sea to the Gulf of Aqaba and in the Farasan Islands, as well as on reefs in Sudan and

Yemen (DeVantier and Pilcher 2000, Kotb et al. 2004). High temperature stress over extended periods can be quantified with a “degree heating week” unit (Gleeson and

Strong 1995). A sharp increase in degree heating weeks was recorded off the coast of

Thuwal in late August 2010, and reached 10-11 degree heating weeks by late September, before decreasing in November, resulting in a local bleaching event in the central Red

Sea (Furby et al. 2013 / see annexe 1, p.167). Inshore reefs were the most heavily affected. Prior to the bleaching, the coral cover on two inshore reefs was 43-50% at 5 m,

24-36% at 10 m and 22-26% at 15 m. Following the hot temperatures in the summer

2010, 98-100% corals bleached at 5 m, 84-96% corals bleached at 10 m and 9-50% corals bleached at 15 m. Six months after the bleaching, the coral cover had dropped to 3-6% at

5m and 8-10% at 10m but had recovered at 15m (Furby et al. 2013 / see annexe 1, p.167).

106

The aim of this study was to monitor the first years of recovery of Abu Shosha Reef, one of the inshore reefs which was heavily impacted by the 2010 bleaching, by monitoring changes in adult coral cover, numbers of juvenile corals, and coral settlement between

March 2011 and July 2013.

Methods

Study Site

This study was conducted on Abu Shosha Reef (22°18'12.43"N, 39° 2'52.05"E), a small inshore reef (130 x 340 m) located 4 km from the coast of Thuwal, in the central Red Sea

(Fig. 1), 3km north of Tahla, one of the reefs surveyed in Furby et al. (2013 / see annexe

1, p.167). Our first survey was conducted six months after the bleaching event. Surveys were conducted both on the exposed (to the prevalent wind and wave conditions) and the sheltered side of the reef.

a b

107

Figure 1 Maps indicating a Thuwal, on the Saudi Arabian coast in the central Red Sea and b

Abu Shosha on the reefs in front of Thuwal

Adult scleractinian corals and other benthos

Benthic cover was quantified in March 2011, September 2011, March 2012, September

2012 and August 2013 using 20 m line intercept transects at three depths, 1 m, 5 m, and

10 m (only 1 m and 5 m in March 2011), on the western (exposed) and eastern (sheltered) sides of the reef. Three replicate transects separated by 5 m were completed at each depth and each side of the reef, resulting in a total of 18 transects per survey. Transects were placed haphazardly at each depth and exposure. On each transect, the benthic category was identified and measured to the nearest centimeter. The following main categories were recorded: hard coral, soft coral, hydroid, ascidian, sponge, algae, suitable substrate for coral settlement, rubble (defined here as 1-10 cm loose fragments) and sand. Hard corals were identified to the genus level and grouped per clade following Huang et al.

(2011); soft corals were identified to the genus level when possible; algae were separated into macroalgae and turf algae; and suitable substrate for coral settlement was separated into crustose coralline algae (CCA) and rock.

Juvenile corals

Juvenile corals were surveyed on the same days, depths and exposure as the adult communities using 1x1 m quadrats, placed every 5 m along each 70 m line intercept tape until reaching 15 replicates. All reefs were surveyed for juvenile corals, ranging from 0.1 108 to 4 cm in diameter (only 1-4 cm juveniles recorded in March 2011). The maximum diameter was chosen to be 4 cm, as within this size range corals resulting from sexual reproduction are easily distinguishable from corals resulting from asexual reproduction

(Bak and Engel 1979). To assist in finding the smaller juveniles, we used a fluorescence- exciting light (NightSea FL-1) on blinking mode. A yellow barrier filter was used over the diving mask to block out the excitation light and easily detect juveniles which would blink conspicuously, even in daytime (Roth and Knowlton 2009). Juvenile corals were categorised into one of three sizes: 0.1-1 cm, 1-2 cm and 2-4 cm, and identified to the genus level when possible. Very small recruits were only identified to the family level.

Coral settlement

Settlement was quantified using pairs of terracotta tiles (10 x 10 cm) superimposed to expose the unglazed sides. Each pair of tiles was individually attached with a bolt drilled onto the reef. Sixty pairs of tiles were installed at different angles in March 2011, a month before the beginning of the spawning season, which starts in April in the central

Red Sea (Bouwmeester et al. 2011 / see annexe 1, p. 137). Equal numbers of tiles were installed at 1m and 5m. However due to difficult access to the exposed side during the installation weeks, 49 pairs of tiles were installed on the sheltered side and 11 on the exposed side. The tiles were collected and replaced every four months until July 2013. In

March 2012, 5 new sets of tiles were added to the exposed side but 2 pairs were lost on the sheltered side by July 2012. By July 2013, 15 more pairs were lost on the sheltered side. Following retrieval, tiles were systematically examined using a Zeiss Stemi 2000 dissecting microscope. Macroalgae and other fast growing organisms were consistently 109 removed to expose all recruits after which recruits were easily detected. Living recruits were counted on each tile and photographed with a Zeiss AxioCam Icc1 camera mounted on a Zeiss Discovery.V20 stereomicroscope. Each recruit was then identified as belonging to the coral families Acroporidae, Pocilloporidae or Poritidae based on

Babcock et al. (2003) or were classified as “other” when assigned to another family or when they could not be identified.

Statistical analyses

All statistical analyses were conducted with R (R Core Team 2013). Difference in the mean cover of hard coral and numbers of juvenile corals among depth and exposure were explored using a one-way ANOVA and were followed by Tukey’s HSD post-hoc tests where applicable. Data was log (x + 1) transformed where necessary to meet the ANOVA assumptions. Shapiro’s test was used to test normality and Bartlett’s test was used to verify homogeneity of variance. Difference in coral cover over time was tested with a one-way ANOVA. Kruskal-Wallis’ test was used to test the difference in numbers of juveniles per square meter over time and was followed by a post-hoc test using Mann-

Whitney tests with Bonferroni correction where applicable, using the “coin” package

(Hothorn 2008). Variability of numbers of recruits according to exposure was explored using a Wilcoxon test.

Results

Adult assemblages 110

In total, 27 genera of scleractinian corals from 10 clades were counted during this study.

The adult assemblages were composed mainly of Porites (66.9%), Echinopora (6.3%),

Goniastrea (6.0%), and Acropora (5.6%) (Fig. 2a). There was a significant difference in coral cover between the exposed and the sheltered side (F(1,82) = 66.34, p-value <

0.001) throughout the survey with an average cover of 3.9 ± 0.5 (% ± SE) on the exposed side and 15.6 ± 1.3 (% ± SE) on the sheltered side (Fig. 3, 4). The sheltered side was composed mainly of massive Porites colonies which had resisted the 2010 bleaching event (Fig. 2b). Depth had no effect on coral cover except in March 2011 when coral cover was higher at 5m than at 1m (F(1,8) = 5.91, p-value = 0.04). No significant increase in coral cover was detected on the exposed side (F(1,40) = 1.32, p-value = 0.26) or the sheltered side (F(1,40) = 0.05, p-value = 0.83)between March 2011 and August

2013.

20 Relative Relative %covercoral 10

111

100% 80%

60%

40% exposed 20%

exposed sheltered

0% %Distribution sheltered vs

Figure 2 a Relative abundance (out of all coral cover) of the 15 most abundant genera during the entire survey b Relative distribution of each genus on the sheltered versus the exposed side

± 15 sheltered

10 exposed %covercoral

0 2011 2012 2013

Figure 3 Average coral cover (% ± SE) during this study on the exposed and sheltered side of

Abu Shosha. Each point represents a survey date 112

Figure 4 Exposed side of Abu Shosha Reef at 1-5m, showing a very low live coral cover and mostly dead branching coral colonies, which probably died following the 2010 bleaching event.

Abundance of juvenile corals

A total of 2034 juveniles from 30 genera and 11 clades were counted in 420 m2. The most abundant genera were Porites (12.3 %), Stylophora (8.1 %), Leptastrea (6.6 %), and

Acropora (5.7 %) (Fig. 5a). There was no significant difference in numbers of juveniles between depths or exposures at the beginning of the study (depth: F(1,56) = 5.86, p-value

= 0.68; exposure: F(1,56) = 3.69, p-value = 0.06), but from September 2011 onwards (12 months after the bleaching event), there were significantly more recruits in deeper waters than in shallower waters (Sept 2011: F(2,84) = 23.97, p-value < 0.001; Mar 2012: F(2,84)

= 30.02, p-value < 0.001; Sept 2012: F(2,84) = 7.98, p-value < 0.001; Aug 2013: F(2,84)

= 17.34, p-value < 0.001;) (exception: September 2012 where number of juveniles at 5 113 and 10 m were not significantly different), and, from September 2012 onwards, there were significantly more juveniles on the exposed side than on the sheltered side (F(1,84)

= 28.99 (2012) and 29.55 (2013), p-value <0.001). March 2011 (bleaching + 6 months) was dominated by Stylophora juveniles, September 2011 and March 2012 by Stylophora and Porites juveniles, and September 2012 and August 2013 by Porites juveniles (Fig.

5b). Numbers of juvenile corals were stable for the first year but started to increase in

March 2012 (Fig. 6). Time had a significant effect on the number of juveniles (X2(3) =

25.78, p-value < 0.001); by August 2013, three years after the bleaching event, numbers were significantly higher than in all earlier surveys (p-value < 0.001).

juveniles juveniles [%] 20 Relative Relative abundanceof 10

114

100% 90% 80% 70% 60% Aug 2013 50% Sept 2012 40% 30% Mar 2012 20% Sept 2011 10% Mar 2011 0%

Figure 5 Relative abundance of the 16 most abundant genera of juvenile corals during the entire survey

Coral settlement

All coral recruits were found on the bottom tile of each superimposed pair, except for 1 recruit, found on a top plate at 5 m depth on the sheltered side in November 2012.

Recruits were patchily distributed on the plates, ranging from 0 to 17 recruits per tile, making up a total of 421 recruits for all surveys. The number of recruits revealed a seasonal pattern with a peak in July (Fig. 6a). There was no significant difference in settlement between 1 m and 5 m depth throughout the 2.5 years of this study (Jul 2011: W

= 467.5, p-value = 0.76; Nov 2011: W = 512.5, p-value = 0.21; Mar 2012: W = 465.5, p- value = 0.72; Jul 2012: W = 502.5, p-value = 0.73; Nov 2012: W = 530, p-value = 0.58;

Mar 2013: W = 472.5, p-value = 0.69; Jul 2013: W = 481, p-value = 0.62). No recruits were found on the exposed side in November and March (all years). However, in July, recruits were distributed on both sides of the reef with significantly more recruits on the 115 sheltered side than the exposed side in 2012 and 2013 (W = 518.5, p-value = 0.008 and W

= 562, p-value < 0.001 respectively). Each 4-month period was dominated by pocilloporid recruits (Fig. 6b). Acroporids were only found on tiles collected in July.

Poritids were mostly present in July and November, and one recruit was found in March both in 2012 and 2013. When comparing the peak seasons, the abundance of recruits was not significantly different between 2011, 2012 and 2013 although recruitment appeared to have dropped from 2011 to 2012, before increasing again (Fig. 6a).

± 8

4 Juveniles Juveniles perm

0 2011 2012 2013

Figure 6 Average number of juveniles per m2 (± SE) during this study. Depths and exposures are pooled together. Each point represents a survey date

116

2.5

± 1.5

Recruitsper tile 0.5

100% 80% 60% 40% 20% 0% Jul Nov Mar Jul Nov Mar Jul 2011 2011 2012 2012 2012 2013 2013

Acroporidae Pocilloporidae Poritidae other

Figure 7 a Average number of recruits per 100 cm2 tile (± SE) during this study. Depths and expositions are pooled together b Proportion of recruits from Acroporidae, Pocilloporidae,

Poritidae and other families present on the settlement tiles after each 4-month period

Availability of suitable substrate for recruitment and algal cover

There was no significant difference in the availability of suitable recruitment substrate or algal cover over time (algal cover: F(4,76) = 0.48, p-value = 0.75; suitable substrate:

F(4,76) = 0.79, p-value = 0.54). There was on average significantly more available substrate on the sheltered side than on the exposed side (F(1,76) = 10.5, p-value =

0.0018) and also significantly more available substrate at 1 m and 5 m than at 10 m depth

(F(2,76) = 27.26, p-value < 0.001). There was on average significantly more algae on the 117 exposed side than on the sheltered side (F(1,76) = 10.89, p-value = 0.0015) and significantly more algae in shallower waters than deeper waters (F(2,76) = 23.12, p-value

< 0.001).

Discussion

The adult coral cover in this study did not change in 2.5 years. The abundance of juvenile corals however started to increase three years after the bleaching, and coral settlement on artificial substrata remained low but showed seasonal variability. Recovery of coral systems after a disturbance depends heavily on the supply of coral larvae to replenish reefs, which in turns depends on the survival of sufficient adult colonies able to release fertile gametes (Hughes 2000, Lukoschek et al. 2013).

The adult coral cover remained constant on each side of Abu Shosha Reef, with a high coral cover on the sheltered side, mostly due to the presence of massive Porites colonies which survived the bleaching event in 2013. Prior to the bleaching, Abu Shosha Reef seems to have been dominated by branching corals such as Acropora species, assumed from the extensive coverage of dead branching corals covered with algae at the beginning of this study, notably on the exposed side of the reef (Fig. 4). The shift of an Acropora- dominated assemblage to a Porites-dominated assemblage is characteristic of many mass bleaching events (Loya et al. 2001; Baker et al. 2008). The recovery process of Abu

Shosha Reef is still in its early years, and, three years after the bleaching event, the adult coral assemblage has not shown any sign of returning to a healthier state. Slow recovery of reefs in other parts of the world indicate this is not unusual (Baker et al. 2008); a 118 similar pattern was observed on Scott’s Reef in Western Australia where coral cover started to slowly increase two years after a mass bleaching event. Seven years passed before coral cover had doubled, still rising towards its pre-bleaching state (Gilmour et al.

2013). Thus, more time is needed before changes in adult coral cover can be observed in

Abu Shosha.

An increase in the numbers of juvenile corals and continuing coral settlement indicates successful replenishment processes. While coral cover had not shown any signs of recovery between March 2011 and August 2013, the number of juvenile corals significantly increased (Fig. 6). The composition of the juvenile corals also matches closely the adult genera composition, indicating that juveniles could be originating from local adult coral colonies. This is supported by observations of gravid Acropora corals on

Abu Shosha Reef (pers. obs.) in March 2011 and 2012, at 10 m depth, where less bleaching had occurred. Coral larvae are capable of dispersal over hundreds of kilometers, but a strong link between adult abundance and larval supply is often evident

(Hughes et al. 2000). It can, however, not be excluded that coral planulae which settled on Abu Shosha Reef originated from further offshore reefs, where coral communities suffered less from bleaching.

The abundance of settling recruits observed on Abu Shosha Reef was within a range of

0.3 - 1.7 recruits per tile depending on the period and the year, with settlement values above one individual per tile during the spawning season (Fig. 6a). A large-scale study on coral recruitment at different latitudes of the Great Barrier Reef was conducted by 119

Hughes et al. (2002) and showed the variation in recruitment along a latitudinal gradient from 10 - 32°S. No such study is available for the Red Sea and no baselines exist yet to assess whether coral settlement is low for the region. Hughes et al. (2002) however found a general decrease in numbers of recruits with increasing latitudes. They also found an increase in the proportion of brooders versus spawners with increasing latitudes. Thuwal is located at 22°N, a latitude at which coral settlement would be reduced and the proportion of brooders versus spawning would be increased, if on the Great Barrier Reef.

Although we cannot assume abundance would be comparable on the two reef systems, without significantly more data from the Red Sea, the high proportions of settling brooders (Pocilloporidae in this study) compared to spawners can be in part explained by the high latitude of the central Red Sea region. Due to natural temporal variations of settlement in corals and due to the seasonal reproduction patterns of scleractinian corals,

2.5 years were however not enough to determine increase in settlement numbers.

Spatial patterns of recruit distribution on settlement tiles in terms of exposure did not match those of juveniles. Coral recruits on artificial substrata were more abundant on the sheltered side and juveniles on natural substrata were more abundant on the side exposed to wave action, wind, and currents. This suggests potential differences in early stage mortality leading to a juvenile and subsequently adult community different than the one suggested by the early recruits as measured by recruitment tiles. This also underlines the importance of coupling tile-based recruitment studies with in situ juvenile surveys, as conclusions made from recruitment studies alone could provide an incomplete understanding of recovery processes (sensu Penin and Adjeroud 2013). 120

Recruitment on artificial substrata on Abu Shosha Reef showed a seasonal pattern with most recruitment occurring on settlement plates deployed between March and July. The recruits from the peak period most likely originate from spawning corals whose reproductive season in the central Red Sea starts in April and continues for a period of 3-

4 months (Bouwmeester et al. 2011 / see annexe 1, p. 137; Bouwmeester et al. unpublished data). This is further confirmed by the fact that recruits from the

Acroporidae family (which in the Red Sea are all spawners except for one Alveopora species (Baird et al. 2009)) were only present on settlement plates collected after this period. Massive Porites colonies spawn in April or May in the central Red Sea

(Bouwmeester et al. unpublished data). However, reproductive data on Porites monticulosa suggest that this species spawn in November. This would explain why some recruits of Poritidae were found not only on the March-July plates but also on the July-

November plates and even one each year on the November-March plates, possibly after having stayed in the water column a little longer than most other planulae. Recruits of

Pocilloporidae dominated all settlement plates and were found in all three periods, although most were found in the March-July period. Additionally to the proportions of brooders being possibly favoured over spawners in higher latitude regions, their numbers could be enhanced by the fact that they appear to be settle more easily on artificial substrates than other families (Babcock et al. 2003; Nozawa et al. 2011).

Low numbers of settling recruits, similar to those observed in this study, have been observed in post-bleaching studies. For example after the 1998 mass bleaching event, 121 recruitment numbers in Scott’s Reef in Western Australia did not exceed 1 recruit per 100 cm2 tile until 5 years after the bleaching event and did not return to their pre-bleaching levels of 25-50 recruits per tile until 10 years after the bleaching (Gilmour et al. 2013). It is not yet know how far away from “normal” coral settlement is in the central Red Sea.

Only long-term monitoring of settlement patterns on the local reefs will reveal the trends of the region.

At the current rate of increasing sea surface temperatures, bleaching events in the Red

Sea are more and more likely to occur on a more frequent basis (Cantin et al. 2010). The resilience of corals in the Red Sea is, however, not yet fully understood (Riegl et al.

2013). Protective measures such as pollution prevention and limiting fishing of herbivorous fish are crucial during post-disturbance years to allow for the affected reefs to process their recovery (Hughes et al. 2007; Ledlie et al. 2007). Herbivores are present on Abu Shosha Reef, but low densities compared to other locations in the world highlight the need for local management (Khalil et al. 2013).

After three years since the bleaching event which resulted in high mortality on inshore reefs off the coast of Thuwal, it may be slightly too early to see full signs of recovery at

Abu Shosha Reef, given that recovery trajectories often take decades or more in most coral reef systems (Berumen and Pratchett 2006; Pratchett et al. 2011). However, the increasing numbers of juveniles suggest that the next years will probably see a strong increase in both coral cover and recruitment to finally return to an assemblage similar to that present before the disturbance, pending the absence of further disturbances in the 122 region. Nearby offshore reefs, which are exposed to deeper waters less likely to warm extensively during high degree heating weeks on inshore reefs, might also act as refugia for species heavily affected by bleaching, and consequently, protective measures should be applied to them on a long-term basis to ensure the presence of healthy corals able to export larvae to the less resistant inshore reefs which may be repeatedly affected in the future.

Acknowledgements

We would like to thank MM Noble for early discussions in the design of this project.

This work was conducted as part of the PhD thesis of J Bouwmeester in the King

Abdullah University of Science and Technology (KAUST). The work was funded by baseline research funding from KAUST to MLB.

References

Arnold SN, Steneck RS, Mumby PJ (2010) Running the gauntlet: inhibitory effects of algal turfs on the processes of coral recruitment. Marine Ecology Progress Series 414:91-105 Babcock RC, Baird AH, Piromvaragorn S, Thomson DP, Willis BL (2003) Identification of scleractinian coral recruits from Indo-Pacific reefs. Zool Stud 42:211-226 Bak RPM, Engel MS (1979) Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Mar Biol 54:341-352 Baird AH, Marshall PA (2002) Mortality, growth and reproduction in scleractinian corals following bleaching on the Great Barrier Reef. Mar Ecol Prog Ser 237:133-141 Baird AH, Guest JR, Willis BL (2009) Systematic and Biogeographical Patterns in the Reproductive Biology of Scleractinian Corals. Annual Review of Ecology Evolution and Systematics 40:551- 571 Baker AC, Glynn PW, Riegl B (2008) Climate change and coral reef bleaching: An ecological assessment of long-term impacts, recovery trends and future outlook. Estuarine, Coastal and Shelf Science 80:435-471 Barnett TP, Pierce DW, Schnur R (2001) Detection of Anthropogenic Climate Change in the World's Oceans. Science 292:270-274 Berumen ML, Pratchett MS (2006) Recovery without resilience: persistent disturbance and long-term shifts in the structure of fish and coral communities at Tiahura Reef, Moorea. Coral reefs 25:647-653 123

Berumen ML, Hoey AS, Bass WH, Bouwmeester J, Catania D, Cochran JEM, Khalil MT, Miyake S, Mughal MR, Spaet JLY, Saenz-Agudelo P (2013) The status of coral reef ecology research in the Red Sea. Coral reefs:1-12 Bouwmeester J, Khalil MT, De La Torre P, Berumen ML (2011) Synchronous spawning of Acropora in the Red Sea. Coral Reefs 30:1011-1011 Bruno JF, Sweatman H, Precht WF, Selig ER, Schutte VGW (2009) Assessing evidence of phase shifts from coral to macroalgal dominance on coral reefs. Ecology 90:1478-1484 DeVantier L, Pilcher N (2000) The Status of Coral Reefs in Saudi Arabia–2000. PERSGA Technical Series Report, Jeddah 1-45 Dunstan PK, Johnson CR (1998) Spatio-temporal variation in coral recruitment at different scales on Heron Reef, southern Great Barrier Reef. Coral reefs 17:71-81 Fine M, Gildor H, Genin A (2013) A coral reef refuge in the Red Sea. Global Change Biology 19:3640- 3647 Fishelson L (1973) Ecological and biological phenomena influencing coral-species composition on the reef tables at Eilat (Gulf of Aqaba, Red Sea). Mar Biol 19:183-196 Furby KA, Bouwmeester J, Berumen ML (2013) Susceptibility of central Red Sea corals during a major bleaching event. Coral reefs 32:505-513 Gardner TA, Côté IM, Gill JA, Grant A, Watkinson AR (2003) Long-Term Region-Wide Declines in Caribbean Corals. Science 301:958-960 Gilmour JP, Smith LD, Heyward AJ, Baird AH, Pratchett MS (2013) Recovery of an Isolated Coral Reef System Following Severe Disturbance. Science 340:69-71 Gleeson MW, Strong AE (1995) Applying MCSST to coral reef bleaching. Advances in Space Research 16:151-154 Goreau T, McClanahan T, Hayes R, Strong A (2000) Conservation of Coral Reefs after the 1998 Global Bleaching Event. Conserv Biol 14:5-15 Hoegh-Guldberg O, Mumby PJ, Hooten AJ, Steneck RS, Greenfield P, Gomez E, Harvell CD, Sale PF, Edwards AJ, Caldeira K, Knowlton N, Eakin CM, Iglesias-Prieto R, Muthiga N, Bradbury RH, Dubi A, Hatziolos ME (2007) Coral Reefs Under Rapid Climate Change and Ocean Acidification. Science 318:1737-1742 Hothorn T, Hornik K, van de Wiel MA, Zeileis A (2008). Implementing a class of permutation tests: the coin package. Journal of Statistical Software 28(8), 1-23. URL http://www.jstatsoft.org/v28/i08/R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/ Huang D, Licuanan WY, Baird AH, Fukami H (2011) Cleaning up the'Bigmessidae': Molecular phylogeny of scleractinian corals from Faviidae, Merulinidae, Pectiniidae and Trachyphylliidae. BMC evolutionary biology 11:37 Hughes TP, Tanner JE (2000a) Recruitment failure, life histories, and long-term decline of Caribbean corals. Ecology 81:2250-2263 Hughes TP, Baird AH, Dinsdale EA, Moltschaniwskyj NA, Pratchett MS, Tanner JE, Willis BL (2000b) Supply-side ecology works both ways: the link between benthic adults, fecundity, and larval recruits. Ecology 81:2241-2249 Hughes TP, Baird AH, Dinsdale EA, Harriott VJ, Moltschaniwskyj NA, Pratchett MS, Tanner JE, Willis BL (2002) Detecting regional variation using meta-analysis and large-scale sampling: Latitudinal patterns in recruitment. Ecology 83:436-451 Hughes TP, Rodrigues MJ, Bellwood DR, Ceccarelli D, Hoegh-Guldberg O, McCook L, Moltschaniwskyj N, Pratchett MS, Steneck RS, Willis B (2007) Phase Shifts, Herbivory, and the Resilience of Coral Reefs to Climate Change. Current Biology 17:360-365 Khalil M, Cochran JM, Berumen M (2013) The abundance of herbivorous fish on an inshore Red Sea reef following a mass coral bleaching event. Environmental Biology of Fishes 96:1065-1072 Kotb M, Abdulaziz M, Al-Agwan Z, Alshaikh K, Al-Yami H, Banajah A, Devantier L, Eisinger M, Eltayeb M, Hassan M (2004) Status of coral reefs in the Red Sea and Gulf of Aden in 2004. In: Wilkinson CR (ed.), Status of Coral Reefs of the World. Australian Institute of Marine Science Ledlie MH, Graham NAJ, Bythell JC, Wilson SK, Jennings S, Polunin NVC, Hardcastle J (2007) Phase shifts and the role of herbivory in the resilience of coral reefs. Coral reefs 26:641-653 124

Lough JM (2000) 1997–98: Unprecedented thermal stress to coral reefs? Geophysical Research Letters 27:3901-3904 Loya Y, Sakai K, Yamazato K, Nakano Y, Sambali H, van Woesik R (2001) Coral bleaching: the winners and the losers. Ecology Letters 4:122-131 Lukoschek V, Cross P, Torda G, Zimmerman R, Willis BL (2013) The Importance of Coral Larval Recruitment for the Recovery of Reefs Impacted by Cyclone Yasi in the Central Great Barrier Reef. PLoS ONE 8:e65363 Obura DO (2005) Resilience and climate change: lessons from coral reefs and bleaching in the Western Indian Ocean. Estuarine, Coastal and Shelf Science 63:353-372 Omori M, Fukami H, Kobinata H, Hatta M (2001) Significant drop of fertilization of Acropora corals in 1999: An after-effect of heavy coral bleaching? Limnol Oceanogr 46:704-706 Nozawa Y, Tanaka K, Reimer JD (2011) Reconsideration of the Surface Structure of Settlement Plates Used in Coral Recruitment Studies. Zool Stud 50:53-60 Penin L, Michonneau F, Baird AH, Connolly SR, Pratchett MS, Kayal M, Adjeroud M (2010) Early post- settlement mortality and the structure of coral assemblages. Marine Ecology-Progress Series 408:55-64 Pratchett MS, Trapon M, Berumen ML, Chong-Seng K (2011) Recent disturbances augment community shifts in coral assemblages in Moorea, French Polynesia. Coral reefs 30:183-193 Riegl B, Berumen M, Bruckner A (2013) Coral population trajectories, increased disturbance and management intervention: a sensitivity analysis. Ecology and evolution 3:1050-1064 Roth MS, Knowlton N (2009) Distribution, abundance, and microhabitat characterization of small juvenile corals at Palmyra Atoll. Marine Ecology Progress Series 376:133-142 van Woesik R, Sakai K, Ganase A, Loya Y (2011) Revisiting the winners and the losers a decade after coral bleaching. Marine Ecology Progress Series 434:67-76 Vermeij MJA (2006) Early life-history dynamics of Caribbean coral species on artificial substratum: the importance of competition, growth and variation in life-history strategy. Coral reefs 25:59-71 Ward S, Harrison P, Hoegh-Guldberg O (2002) Coral bleaching reduces reproduction of scleractinian corals and increases susceptibility to future stress. Proceedings of the Ninth International Coral Reef Symposium, Bali, 23-27 October 2000 2:1123-1128 Wilkinson CR (1999) The 1997-1998 mass bleaching event around the world. Compilation of Internet Reports. Global Coral Reef Monitoring Network, Australian Institute of Marine Science special publication

125

CHAPTER 5: GENERAL DISCUSSION

Understanding how and when corals reproduce is critical to be able to protect them effectively. In this thesis I significantly increased the knowledge on reproduction of scleractinian corals in the Red Sea. I verified the presence of multi-specific spawning in the Red Sea involving many species from different families. I identified reproductive traits for species which had not been observed to spawn previously. I suggested the use of a plant flowering synchrony index to quantify the level of spawning synchrony of corals in the Red Sea and to compare the Red Sea with other locations in the world. Finally, I monitored coral settlement, juvenile coral abundance and coral cover over 2.5 years on a reef recovering from high mortality after a bleaching event that occurred in the summer

2010.

Chapters two and three specifically address the controversial issue of coral spawning synchrony. In particular, early work from the Red Sea suggested that corals in this region spawn asynchronously, in contrast to corals on the Great Barrier Reef. This thesis demonstrates unequivocally that this early work from Eilat is not representative of the

Red Sea in general and adds to the growing body of evidence to support the hypothesis that multi-specific spawning synchrony is a feature of all specious coral assemblages. I present here an index from the plant biology literature that allows the level of synchrony among populations and assemblages to be quantified and compared. This index suggests that synchrony does vary among coral assemblages. Shlesinger and Loya (1985) and

Shlesinger et al. (1998), conclude from their work that spawning patterns in Eilat are 126 asynchronous and that Red Sea corals exhibit temporal reproductive isolation. However, strong similarities exist between coral assemblages from Eilat and from the central Red

Sea. In both regions the reproductive season lasts four months. In both regions Acropora species spawn early in the season, and most of the species spawn within a week of the full moon. Furthermore the conclusions from Shlesinger and Loya (1985) and Shlesinger et al. (1998) are based on incomplete sampling of broadcasting coral assemblages, with a small proportion of species sampled from distantly related families (Guest et al. 2005). I support the assertion that spawning in Eilat is not asynchronous and that a distinct reproductive season seems to exist. The authors might have been drawn to the conclusion of asynchronous spawning due to the fact that they included brooding corals in their analysis. The first description of multi-specific spawning (from Australia) implied a single period of spawning during which most species spawned over a few days, with a few exceptional years when spawning was split over two consecutive months (Harrison et al. 1994; Willis et al. 1995; Babcock et al. 1996). Following the initial Australian description, spawning patterns in Eilat appeared significantly less synchronous. However, later studies on reefs of the Whitsunday Islands on the GBR showed a reproductive season of 5 months in the Acropora assemblage, showing that the GBR is also capable of a longer reproductive season (Baird et al. 2009b). Discussions on the definition of asynchronous, multi-specific and mass spawning (Mangubhai and Harrison 2008; Baird and Guest 2009a; Mangubhai and Harrison 2009) have led to the need to quantify synchrony so that the right conclusions can be made on the spawning synchrony of a coral assemblage in regions worldwide (Baird and Guest 2009a). Baird and Guest

(2009a) suggested a synchrony index which calculated the probability that two colonies 127 would spawn in the same month. In the Red Sea, some colonies were observed to spawn in two consecutive months and thus the above index could not be used as the probability values were heavily biased in the case of those colonies. I suggested in this thesis to use

Marquis’ synchrony index (Marquis 1988) for quantifying spawning synchrony and show that it is effective in determining the level of synchrony. With this index, further studies on the spawning synchrony and reproductive seasonality of corals should not encounter the situation where spawning would be classified as asynchronous when it is not.

The level of synchrony in the Acropora assemblage in the central Red Sea is lower than recorded in the northern Red Sea on reefs off the Egyptian cities of Hurghada and Marsa

Alam, where a high proportion of colonies from within the coral assemblage mature synchronously in a single month (Hanafy et al. 2010). Subsequent night-time observations in Hurghada revealed spawning of 12 Acropora species over 2 consecutive nights around the full moon of May 2012 (Kotb and Hanafy, unpublished data), which is similar to the proportions recorded in the central Red Sea in 2011 and 2012 during this thesis. This represents approximately half of the number of Acropora species in the region. Further night-time observations in Hurghada around the full moon of April, May,

June and July might reveal the spawning of the other Acropora species as well as additional species from other families, and determine whether the reproductive season is indeed limited to one month only in April or May or whether the reproductive season is longer and more similar to that in the central Red Sea.

128

A well-established spawning season, as it appears to be in the Red Sea, offers easier management options for the relevant environmental authorities. For example in Western

Australia, dredging activities are required to stop during significant multi-specific spawning events (Styan and Rosser 2012). Such a measure could easily be put in place in the Red Sea, possibly also temporarily restricting construction activities on the coastline during that time as many Red Sea reefs are fringing reefs and thus are directly in front of the shore.

Regional differences in the timing of spawning in corals could be due to an evolutionary legacy or due to adaptations to local conditions (Rosser 2013). On reefs around Australia, synchronous multi-specific spawning of corals typically occurs during the austral spring on the east coast and during autumn on the west coast, while there is a secondary multi- specific spawning period in the opposite season on both coasts (Stobart et al. 1992;

Wolstenholme 2004; Rosser and Gilmour 2008). In Western Australia, nine species of

Acropora spawn in both reproductive seasons but in individual colonies, mature eggs are only observed once a year, and in the same season over consecutive years (Rosser and

Gilmour 2008). The existence of two reproductive groups within which spawning is synchronous has also been observed in populations of Mycedium elephantotus and

Echinophyllia aspera in Southern Taiwan (Fan 1996; Dai et al. 2000). The two reproductive groups of Mycedium elephantotus were also clearly subdivided genetically

(Dai et al. 2000). The existence of two reproductive groups within a species may result from either recent divergence in sympatric populations or secondary contact of populations with different origins (Dai et al. 2000; Rosser and Gilmour 2008). Both 129 hypotheses have been suggested to explain the differences between Eastern and Western

Australia. The reproductive seasonality of corals in Australia could be the result of an inherited rhythm from northern ancestors (Simpson 1991). Alternatively, many species could have independently but similarly responded to strong selective pressures promoting reproductive success (Oliver et al. 1988). In Western Australia, Rosser (2013) suggests that adaptation to local environmental conditions, rather than genetic legacy, explains the consistently high proportion of species that spawn during the main reproductive season regardless of latitude. Rosser (2013) also suggests that the same mechanism (local adaptation) explains the trend that a decreasing proportion of species spawn in spring as latitude increases. However, no simple correlation between any single environmental variable and the timing of spawning has been observed (Rosser 2013). Results from this thesis do not offer yet the possibility to determine the exact factors controlling the timing of spawning and/or the differences in timing between regions in the Red Sea. More thorough surveys over multiple years in different regions of the Red Sea would be necessary to eliminate the differences from annual variations within each region and accurately compare patterns between regions. Multivariate analysis on the interactions among the proportion of species spawning, the latitude and the local environmental variables would help understand the drivers of spawning patterns in the Red Sea. The combination of multiple environmental variables to determine the reproductive seasonality of corals has not been fully explored in any study worldwide but future research may provide an insight on the factors which influence the timing of reproduction in scleractinian corals (Rosser 2013).

130

Three years after the bleaching event that happened in 2010 in the central Red Sea, adult coral cover had not increased, however, juvenile abundances were increasing and successful settlement, while comparatively low, was occurring. The reef is still at an early stage of its recovery but providing no further disturbances occur in the next years, the reef is likely to return to a coral cover and species community similar to the one present before the bleaching. Long-term models in the Red Sea predict that shifts in the coral community are unlikely to happen if the frequency of disturbances is maintained at a rate of 20 years or more between disturbance events (Riegl et al. 2013). There are few records of bleaching events or other disturbances in the Central Red Sea, although additional disturbances could have occurred while remaining unreported due to limited observer effort in the region (Berumen et al. 2013 / see annexe 1, p. 155)

Although reproductive data and/or spawning observations have been collected for a wide range of species in the central Red Sea, still many species remain to be studied. One of the main limiting factors with in situ night-time observations is the proportion of species which can be observed within a single site. For this thesis, we selected sites with a relatively high coral cover, high Acropora species richness but with also other species such as merulinids and poritids. Unfortunately some species of the region would have been rare in those sites and observations in a reef with different species composition such as one of the inshore reefs may lead to many additional spawning observations.

Veron (2000) described 292 coral species for the Red Sea. Many of the described genera have since then been revised or are currently under revision and more species have been 131 described since then (e.g. Stefani et al. 2008; Benzoni et al. 2010; Huang et al. 2011;

Benzoni et al. 2012a; Keshavmurthy et al. 2013). As the Red Sea is an isolated basin, many endemic species occur in the Red Sea and although many Red Sea endemics have been identified, it is possible that many remain undescribed. This is naturally also a possibility for species with a wider distribution. One species encountered during this thesis, a Cyphastrea (Merulinidae) with eight septa, clearly did not fit any of the descriptions available for the genus and is consequently now being described as a new species (unpublished data). Other species in this thesis had to be identified to the closest description available following the published literature, pending published results from the revisions of each family or genus. Additionally, recent but still ongoing molecular work suggests some species have been allocated the wrong genus or the wrong family

(e.g. Budd et al. 2012; Benzoni et al. 2012b). In this period of taxonomic change, accurately identifying the 93 species described in this thesis has been a challenging task and could only be completed by supplementing field observations with lab observations of coral skeletons with a dissecting microscope and in some cases with scanning electron microscopy. In parallel, for a speciose group such as Acropora where 38 species were identified in this thesis, a field guide of the different species was built along the course of this thesis, for the sake of consistency in identification, and for future reference, knowing that future revisions of the genus will most probably affect some of the current species designations.

More reproductive data from different locations in the Red Sea are still necessary to accurately identify the differences in spawning patterns and synchrony within the Red 132

Sea. Multivariate analyses on environmental variables in those locations would also help understand the factors which might drive those differences. Reproductive data is also still missing on an important proportion of species in the region and conducting reproductive surveys as well as night-time observations in different habitats might help increase the numbers species observed to spawn. Acropora species are a very useful coral assemblage to study in order to determine the level of synchrony in the region and compare with other locations in the world. However, identifying Acropora species in understudied regions such as the Red Sea (Berumen et al. 2013 / see annexe 1, p. 155) often results in tedious work, with insufficient literature available to confidently identify each species. A revision of the genus with modern taxonomy tools (such as e.g in Stefani et al. 2008;

Benzoni et al. 2012) is therefore needed in the region, as well as in other places in the world. Finally, although coral recruitment and juvenile abundances have been quantified in this thesis, the numbers collected were of those in a recovering system. Recruit and juvenile abundances in normal conditions in a healthy reef are not known yet for the central Red Sea but would be useful to be able to estimate the recovery progress of reefs after a high mortality event.

References

Babcock RC, Bull GD, Harrison PL, Heyward AJ, Oliver JK, Wallace CC, Willis BL (1986) Synchronous spawnings of 105 scleractinian coral species on the Great Barrier Reef. Mar Biol 90:379-394 Baird AH, Guest JR (2009a) Spawning synchrony in scleractinian corals: Comment on Mangubhai & Harrison (2008). Marine Ecology Progress Series 374:301-304 133

Baird AH, Birrel CL, Hughes TP, McDonald A, Nojima S, Page CA, Prachett MS, Yamasaki H (2009b) Latitudinal variation in reproductive synchrony in Acropora assemblages: Japan vs. Australia. Galaxea, Journal of Coral Reef Studies 11:101-108 Benzoni F, Arrigoni R, Stefani F, Stolarski J (2012) Systematics of the coral genus Craterastrea (Cnidaria, Anthozoa, Scleractinia) and description of a new family through combined morphological and molecular analyses. Systematics and Biodiversity 10:417-433 Benzoni F, Arrigoni R, Stefani F, Reijnen BT, Montano S, Hoeksema BW (2012b) Phylogenetic position and taxonomy of Cycloseris explanulata and C. wellsi (Scleractinia: Fungiidae): lost mushroom corals find their way home. Contrib Zool 81:125-146 Berumen ML, Hoey AS, Bass WH, Bouwmeester J, Catania D, Cochran JEM, Khalil MT, Miyake S, Mughal MR, Spaet JLY, Saenz-Agudelo P (2013) The status of coral reef ecology research in the Red Sea. Coral reefs:737-748 Budd AF, Fukami H, Smith ND, Knowlton N (2012) Taxonomic classification of the reef coral family Mussidae (Cnidaria: Anthozoa: Scleractinia). Zoological Journal of the Linnean Society 166:465-529 Dai C, Fan T, Yu J (2000) Reproductive isolation and genetic differentiation of a scleractinian coral Mycedium elephantotus. Marine Ecology Progress Series 201:179-187 Fan T-Y (1996) Life histories and population dynamics of foliaceous corals in northern and southern Taiwan. Ph. D. thesis, National Taiwan University, Taipei, Taiwan, Guest JR, Baird AH, Goh BPL, Chou LM (2005) Seasonal reproduction in equatorial reef corals. Invertebrate Reproduction & Development 48:207-218 Hanafy MH, Aamer MA, Habib M, Rouphael AB, Baird AH (2010) Synchronous reproduction of corals in the Red Sea. Coral Reefs 29:119-124 Harrison PL, Babcock RC, Bull GD, Oliver JK, Wallace CC, Willis BL (1984) Mass Spawning in Tropical Reef Corals. Science 223:1186-1189 Huang D, Licuanan WY, Baird AH, Fukami H (2011) Cleaning up the'Bigmessidae': Molecular phylogeny of scleractinian corals from Faviidae, Merulinidae, Pectiniidae and Trachyphylliidae. BMC evolutionary biology 11:37 Keshavmurthy S, Yang S-Y, Alamaru A, Chuang Y-Y, Pichon M, Obura D, Fontana S, De Palmas S, Stefani F, Benzoni F, MacDonald A, Noreen AME, Chen C, Wallace CC, Pillay RM, Denis V, Amri AY, Reimer JD, Mezaki T, Sheppard C, Loya Y, Abelson A, Mohammed MS, Baker AC, Mostafavi PG, Suharsono BA, Chen CA (2013) DNA barcoding reveals the coral “laboratory-rat”, Stylophora pistillata encompasses multiple identities. Sci Rep 3 134

Mangubhai S, Harrison PL (2008) Asynchronous coral spawning patterns on equatorial reefs in Kenya. Marine Ecology Progress Series 360:85-96 Mangubhai S, Harrison PL (2009) Extended breeding seasons and asynchronous spawning among equatorial reef corals in Kenya Marquis RJ (1988) Phenological variation in the neotropical understory shrub Piper arielanum: causes and consequences. Ecology 69:1552-1565 Oliver J, Babcock R, Harrison P, Willis B (1988) Geographic extent of mass coral spawning: clues to ultimate causal factors. Proc 6th int coral Reef Symp 2:803-810 Riegl B, Berumen M, Bruckner A (2013) Coral population trajectories, increased disturbance and management intervention: a sensitivity analysis. Ecology and evolution 3:1050-1064 Rosser NL (2013) Biannual coral spawning decreases at higher latitudes on Western Australian reefs. Coral reefs 32:455-460 Rosser N, Gilmour J (2008) New insights into patterns of coral spawning on Western Australian reefs. Coral reefs 27:345-349 Shlesinger Y, Loya Y (1985) Coral community reproductive patterns: Red Sea versus the Great Barrier Reef. Science 228:1333-1335 Shlesinger Y, Goulet T, Loya Y (1998) Reproductive patterns of scleractinian corals in the northern Red Sea. Mar Biol 132:691-701 Simpson C (1991) Mass spawning of corals on Western Australian reefs and comparisons with the Great Barrier Reef. Journal of the Royal Society of Western Australia 74:85-91 Stefani F, Benzoni F, Pichon M, Cancelliere C, Galli P (2008) A multidisciplinary approach to the definition of species boundaries in branching species of the coral genus Psammocora (Cnidaria, Scleractinia). Zoologica Scripta 37:71-91 Stefani F, Benzoni F, Yang SY, Pichon M, Galli P, Chen CA (2011) Comparison of morphological and genetic analyses reveals cryptic divergence and morphological plasticity in Stylophora (Cnidaria, Scleractinia). Coral reefs 30:1033-1049 Stobart B, Babcock R, Willis B (1992) Biannual spawning of three species of scleractinian coral from the Great Barrier Reef. Proc 7th int coral reef Symp 1:494-499 Styan CA, Rosser NL (2012) Is monitoring for mass spawning events in coral assemblages in north Western Australia likely to detect spawning? Marine Pollution Bulletin 64:2523- 2527 Veron JEN (2000) Corals of the world. Australian Institute of Marine Science, Townsville 135

Willis BL, Babcock RC, Harrison PL, Oliver JK, Wallace CC (1985) Patterns in the mass spawning of corals on the Great Barrier Reef from 1981 to 1984. 5th International Coral Reef Symposium 4:343-348 Wolstenholme JK (2004) Temporal reproductive isolation and gametic compatibility are evolutionary mechanisms in the Acropora humilis species group (Cnidaria; Scleractinia). Mar Biol 144:567-582

136

APPENDIX 1: LIST OF PUBLICATIONS

Publications arising from this thesis

1. Bouwmeester J, Khalil MT, De La Torre P, Berumen ML (2011) Synchronous spawning of Acropora in the Red Sea. Coral Reefs 30:1011-1011

This paper resulted from chapter 3 and describes the first observation of synchronous multi-specific spawning in the Red Sea. Contribution: JB collected the data in the field and conducted the data analysis.

2. Bouwmeester J, Berumen ML, Baird AH (2011) Daytime broadcast spawning of Pocillopora verrucosa on coral reefs of the central Red Sea. Galaxea, Journal of Coral Reef Studies 13:23-24

This paper resulted from chapter 3 and describes the first in situ observation of day-time spawning of Pocillopora verrucosa. Contribution: JB observed the spawning in the field and wrote the description.

Reef sites

Synchronous spawning of Acropora in the Red Sea

Multi-specific synchronous spawning is a reproductive strategy used by scleractinian corals that has now been described from coral reefs in 23 locations globally (Baird et al. 2009). While high multi-specific synchrony in the reproductive condition of Acropora colonies has been documented in the Red Sea in April and/or May (Hanafy et al. 2010), multi-specific synchronous spawning has not been directly observed. In April 2011, mature oocytes were found in a high proportion of colonies in numerous species of Acropora on reefs near Thuwal, Saudi Arabia (2218¢19.26†N, 3857¢56.66†E). On the night of April 16 2011, two nights before the full moon, egg-sperm bundles were first observed in Acropora polyps at 20:30 h (Fig. 1a). Between 22:30 and 23:45 h, 43 colonies from 10 out of 13 surveyed Acropora species released egg/sperm bundles (Fig. 1b), including three species that had not been observed to spawn previously (A. plantaginea, A. parapharaonis, and A. lamarcki). This is the first documented multi-specific synchronous spawning event in the Red Sea, demonstrating that the asynchronous spawning pattern at Eilat in the Gulf of Aqaba (Shlesinger and Loya 1985) is not representative of the Red Sea, and providing further support for the prediction that these events are characteristic of all speciose coral assemblages (Guest et al. 2005).

Acknowledgments The authors would like to thank AH Baird for his assistance in Acropora taxonomy.

References

Baird AH, Guest JR, Willis BL (2009) Systematic and biogeographical patterns in the reproductive biology of scleractinian corals. Ann Rev Ecol Evol Sys 40:551–571 Guest JR, Baird AH, Goh BPL, Chou LM (2005) Seasonal reproduction in equatorial reef corals. Invertebr Reprod Dev 48:207–218 Hanafy MH, Aamer MA, Habib AB, Baird AH (2010) Synchronous reproduction of corals in the Red Sea. Coral Reefs 29:119–124 Shlesinger Y, Loya Y (1985) Coral community reproductive patterns: Red Sea versus the Great Barrier Reef. Science 228:1333–1335

J. Bouwmeester (&) Á M. T. Khalil Á P. De La Torre Á M. L. Berumen Red Sea Research Center, King Abdullah University of Science and Technology, Thuwal, Kingdom of Saudi Arabia e-mail: [email protected] Fig. 1 a Setting of gamete bundles in Acropora gemmifera, b Chaetodon austriacus M. L. Berumen feeding on egg-sperm bundles released from Acropora lamarcki Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA

Received: 3 May 2011 / Accepted: 13 June 2011 / Published online: 9 July 2011 Coral Reefs (2011) 30:1011 Springer-Verlag 2011 DOI 10.1007/s00338-011-0796-5

123

140

3. Bouwmeester J, Baird AH, Chen C-J, Guest JR, Vicentuan KC, Berumen ML (in preparation) Spawning synchrony of scleractinian corals in the Red Sea. Coral Reefs. Contribution: JB collected the data in the field and conducted the data analysis.

This paper is presented in chapter 3 and describes reproductive data from a wide variety of species as well as spawning observations of scleractinian corals during three consecutive reproductive seasons.

4. Bouwmeester J, Baird AH, Edwards AJ, Guest JR, Bauman A, Berumen ML (in preparation) Reproductive seasonality of the Acropora assemblage in the central Red Sea. Proc B.

This paper is presented in chapter 2 and presents two years of reproductive data in the central Red Sea. A flowering synchrony index was adapted to quantify spawning synchrony and was used to compare synchrony in various locations in the Indo Pacific. Contribution: JB collected the data in the field, collected coral samples, performed histological analyses and conducted the data analysis.

5. Bouwmeester J, Khalil MT, AH Baird, Berumen ML (in preparation) Coral recovery and recruitment after a bleaching event in the central Red Sea. Coral Reefs

This paper is presented in chapter 4 and describes the early post-bleaching recovery patterns in the adult coral community, juvenile corals and coral recruits on an inshore reef which suffered high mortality after a bleaching event. Contribution: JB conducted the surveys in the field, installed and collected the settlement plates, quantified coral recruits at the lab, and conducted the data analysis.

141

Additional publications

6. Roder C, Berumen ML, Bouwmeester J, Papathanassiou E, Al-Suwailem A, Voolstra CR (2013) First biological measurements of deep-sea corals from the Red Sea. Sci Rep 3

This paper resulted from the deep sea leg (leg 5) of the KAUST Red Sea Expedition 2011 on board the R/V Aegaeo from 30 November to 14 December 2011. It describes observations of several deep sea corals found on sea mounts in the central Red Sea as well as biological measurements on some of the species collected. Contribution: JB was chief scientist of the deep sea leg, did the morphological description for the coral taxonomy, created GIS maps of the sites, and analysed the ROV tracking data.

7. Berumen ML, Hoey AS, Bass WH, Bouwmeester J, Catania D, Cochran JEM, Khalil MT, Miyake S, Mughal MR, Spaet JLY, Saenz-Agudelo P (2013) The status of coral reef ecology research in the Red Sea. Coral reefs:1-12

This review underlines the little coral reef ecology research which has been conducted in the Red Sea, compared to other locations such as the Caribbean and the Great Barrier Reef. Contribution: JB analysed and wrote the coral reproductive biology section

8. Furby KA, Bouwmeester J, Berumen ML (2013) Susceptibility of central Red Sea corals during a major bleaching event. Coral reefs 32:505-513

This paper documents for the first time the susceptibility to bleaching of corals in the Red Sea during unusually high temperatures in the summer 2010 and the subsequent survival or mortality of those corals. Contribution: JB participated in the data collection in the field

OPEN First biological measurements of

SUBJECT AREAS: deep-sea corals from the Red Sea ECOSYSTEM ECOLOGY C. Roder1, M. L. Berumen1,2, J. Bouwmeester1, E. Papathanassiou3, A. Al-Suwailem4 & C. R. Voolstra1 GENETICS

METABOLISM 1 2 Red Sea Research Center, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia, Biology Department, BIODIVERSITY Woods Hole Oceanographic Institution, Woods Hole, USA, 3Hellenic Centre for Marine Research, Anavissos, Greece, 4Coastal and Marine Resources Core Lab, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia.

Received 11 March 2013 It is usually assumed that metabolic constraints restrict deep-sea corals to cold-water habitats, with ‘deep-sea’ and ‘cold-water’ corals often used as synonymous. Here we report on the first measurements of Accepted biological characters of deep-sea corals from the central Red Sea, where they occur at temperatures 11 September 2013 exceeding 206C in highly oligotrophic and oxygen-limited waters. Low respiration rates, low calcification rates, and minimized tissue cover indicate that a reduced metabolism is one of the key adaptations to Published prevailing environmental conditions. We investigated four sites and encountered six species of which at least 3 October 2013 two appear to be undescribed. One species is previously reported from the Red Sea but occurs in deep cold waters outside the Red Sea raising interesting questions about presumed environmental constraints for other deep-sea corals. Our findings suggest that the present understanding of deep-sea coral persistence and resilience needs to be revisited. Correspondence and requests for materials should be addressed to lthough first observed more than a century ago1, coral ecosystems of the deeper aphotic zones of the 2 C.R. (cornelia.roder@ oceans have only recently attracted broad interest . Deep-sea corals primarily inhabit the dysphotic and kaust.edu.sa) or C.R.V. aphotic zones of continental shelves, mainly in areas of distinct topography such as slopes, canyons, and A 3 4 seamounts . Relatively strong currents in such areas maximize the food supply and also facilitate sediment (christian.voolstra@ removal from either sessile organisms or the seafloor to provide suitable settlement substrata for coral larvae5. kaust.edu.sa) In addition to nutrient availability, the main factors determining coral settlement include temperatures not exceeding 12uC2,3,6, aragonite saturation7, and sufficient oxygen levels8. While the latter two are vital to maintaining calcification and aerobic metabolism in corals, the low temperature regimes decelerate food decay and reduce metabolic demands. Aphotic deep-sea corals are thus synonymously referred to as ‘cold-water corals’e.g. 2–4,6,7. Deep-sea coral reefs are largely valued for their importance as biodiversity hotspots2, climate archives9, habitats for commercial fish species10, and promising sources of bioactive compounds11. To date, most research in the area has focused on describing the diversity and distribution of deep-sea coral reefs. While responses to local disturbances such as deep-sea trawling, seafloor drilling, mining, or oil spills have been investigated12, the impacts of global climate change, such as ocean acidification and warming, remain less well understood, especially because they are difficult to measure in deep-water habitats. An excess of dissolved carbon dioxide in the oceans has been predicted to shoal the aragonite saturation horizon (ASH), limiting the availability of suitable habitat for deep-sea corals13. By 2099 more than two thirds of the known deep-water coral habitats are predicted to be located in waters below the ASH potentially altering calcification abilities of corals, and possibly impeding planktonic growth, and hence affecting food availability for the deep-sea benthos14. Rising sea surface temperatures, a well-described threat to shallow-water coral reefs resulting in bleaching and mass mortalities15, have been documented at great depths16. Although deep-sea corals can tolerate substantial fluctuations in temperature17, a rise of 2uC for only a few hours can cause significant increases in energy demands18 with potentially serious consequences for organisms thriving in nutrient-deprived environments. A detailed understanding of the impacts of global climate change on the biology of slow-growing, long-living deep-sea corals is therefore critical. The Red Sea is one of the warmest, most saline, and most oligotrophic marine ecosystems on Earth. It is unique in sustaining year-round temperatures exceeding 20uC throughout the water column (.2000 m)19. Unfortunately, the ecology of the Red Sea is under-studied, particularly outside of the Gulf of Aqaba20. Nonetheless, it is known for its extensive shallow water coral reefs of which some species are known to occur in depths outside the photic zone, in areas where available light is below 1% of that from the surface21. The presence of azooxanthellate corals in deep waters (between 400 – 1000 m) of the Red Sea was first documented in the late nineteenth century22, but it was not possible to collect biological measurements at that time. Later

Table 1 |Summary of sites, locations, species observed, environmental characteristics and other measurements at four study sites in the central Saudi Arabian Red Sea. Measurements are listed with (6SE). GenBank accession numbers are included with each species. TSM: Total Suspended Matter, OC: Organic Carbon. Note that at all sites bathymetric data, video recordings, and CTD data were collected

# obeserved Nitrite 1 Site Geographic along ROV Oxygen TSM OC in Nitrate metabolic water 21 21 21 # Location Species GenBank ID Depth [m] Habitat track Temp (uC) (mg l ) Varag (mg l ) TSM (wt.%) (mmol l ) sampled measurements samples 122u46.0259N/ Sp A (Caryophyllidae) [JX629249] 600–700 on hard . 14 21.53 6 1.18 3.48 1.47 75.65 16.32 YY 38u03.5649E substrate, (0.0001) (60.002) (60.02) (60.08) (611.76) (60.13) sometimes on sandy bottom and close to rocky structures Dendrophyllia sp. [JX629248] 570–640 associated 8 21.53 1.54 3.44 2.07 70.04 15.47 YYY with ridges, (60.00006) (60.00004) (60.03) (60.43) (67.38) (60.56) mainly hanging Red polyp (solitary) 680–740 5 NN White polyp (solitary) 600–740 7 NN 222u14.8629N/ Sp A (Caryophyllidae) 240–550 7 YY 38u53.5989E Rhizotrochus typus 280–290 1 NNY Eguchipsammia fistula [JX629250] 230–270 forming . 100 21.60 1.02 3.61 1.52 75.16 16.58 YY monospecific (60.0006) (60.004) (60.06) (60.09) (613.16) (60.21) coral gardens

on seamount www.nature.com/ tops 322u16.3289N/ Sp A (Caryophyllidae) 560–590 1 NNN 38u49.0719E 422u17.8359N/ Eguchipsammia fistula 320 . 50 YNN 38u53.8159E scientificreports 2 www.nature.com/scientificreports work utilized extensive submersible deployments in the Gulf of m depth), and at depths around 300 m declined even below 1 mg l21 Aqaba21,23,24 to investigate extreme boundaries of the mesophotic reef (Fig. 1). With a pH between 8–8.1 below the thermocline and a total system (to approximately 200 m depth). To date, however, no mech- alkalinity (TA) between 2400–2500 mequiv kg21, aragonite saturation anism has been provided to explain the persistence of azooxanthel- (Varag) in the sampling sites was on average 3.5 (Fig. 1). Nitrate plus late corals at depths between 200 – 1000 m in the Red Sea under nitrite concentrations rose from almost 0 in surface to more than relatively warm water temperatures, nor has their physiology and 15 mM in deep water. While amounts of total suspended matter metabolism been studied. Because the existing body of knowledge (TSM) were relatively consistent over the water column, C:N ratios of deep-sea coral is based on studies of cold-water corals, the iden- increased with depth (Fig. 1). Specific measurements of environmental tification and study of deep-sea corals and their underlying biology in parameters were taken at two sites where coral samples were collected the Red Sea could open the door to new studies of the mechanisms (Table 1); these measurements were consistent with the results from used to adapt to these unique conditions. our profiling of the water column (Fig. 1). The overall aim of this study is to present a report of opportunistic sampling of deep-sea corals in the Red Sea. We give some description Taxonomy. At least six deep-sea coral species were observed in the of the distribution and habitat of various species of the observed Red Sea of which the following three were successfully collected and deep-sea corals. Measurement of some basic biological characters morphologically examined. provides some insight to how these corals function and survive in Eguchipsammia fistula (Dendrophyllidae) (Fig. 2): Corallum colo- such a unique environment. These results are particularly interesting nial, corallum and septa porous, branching extra-tentacular, septa in light of current theories of deep-sea coral biology, which assume arranged in Pourtale`s plan, columella spongy, theca porous, no pali. much cooler water conditions. Colonies formed monospecific coral gardens on seamount tops. In some cases, gardens of dead skeleton were also found on the sea- Results mounts, usually a few meters below the top. Study sites. All four sites investigated by ROV or submersible were Dendrophyllia sp. (Dendrophyllidae) (Fig. 3): Corallum colonial, found to have corals present (Table 1). Two sites were found to have branching extra-tentacular, septal symmetry hexameral, columella multiple species of corals, while two sites had only single species spongy, theca rough, single cycle of septa observed on sample, no present (although sometimes in large numbers). All four sites pali. Colonies were found around ridges, mainly hanging under a showed high topographic relief (seamount-like or steep-walled ridge but also in compacted sand just below or above a ridge. habitats) (Supplement Fig. 1). Undetermined species of Caryophyllidae (hereafter ‘‘species A’’) (Fig. 4): Corallum colonial, branching extra-tentacular, septal sym- Environment of the deep Red Sea. While temperature in the aphotic metry hexameral, septa arranged in 5 cycles, columella present, theca zone of the Red Sea barely varies from 21.5uC and salinity below the smooth with single row of bumps on ridge of costae, septa edges thermocline increases to more than 40 PSU, oxygen concentrations smooth, septa walls granular, pali present before septa of 3rd cycle. werefoundtobelowerthan2mgl21 in subthermocline water (. 200 Colonies were mostly observed on the bottom of rocky structures or

Figure 1 | (A) Depth profiles (y-axes) of physical and biogeochemical water properties of the Red Sea. Left graph: temperature, salinity, oxygen concentration. Middle graph: pH, total alkalinity (TA), aragonite saturation (Varag). Right graph: nitrate plus nitrite concentrations, carbon-to-nitrogen ratios, total suspended matter (TSM) contents. Solid black lines in each graph indicate the margin of the photic zone (1% surface radiation) for all wavelengths within the spectral photosynthetically active radiation (PAR) range. Differently dashed lines represent distribution depths of the three sampled coral species: the style of the dashed line indicates the respective coral species in its preferred habitat ((B) Eguchipsammia fistula; (C) Species A; (D) Dendrophyllia sp.). Images obtained with ROV camera.

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 3 www.nature.com/scientificreports

Figure 2 | Eguchipsammia fistula, Dendrophylliidae. (a) Habitat view. (b) Lateral view of corallum. (c) Colony view (scalebar in cm). (d) and (e) Calicular view of corallum. (f) and (g) Close-up (SEM) of the theca half way up the corallum. Image a obtained with ROV camera. Image c by Roder/ Voolstra. Images b, d, e, f, g by Bouwmeester.

ridges just above the sand, with a few colonies growing directly on the hanging from the top of the ridge. Although HD video images were compacted sand. available, their resolution was still too low to identify either to family The other three deep-sea coral species observed (Supplement Fig. level or further. 2) were not collected. One species, a single white solitary coral, Phylogenetic analyses (Fig. 5) of partial sequences of the mitochon- observed within a field of dead skeletons of Eguchipsammia fistula drial 16S ribosomal RNA encoding gene were conducted to confirm at site 2 (Supplement Fig. 2a) could be determined as Rhizotrochus the taxonomic classification of our deep-sea coral specimens22,25. typus due to its funnel-shaped corallite, approximately 1 cm thick at Morphological classifications coincided with molecular phylogenetic the base and 5 cm wide at the top. On ridges at site 1, more solitary comparisons with published scleractinian coral sequences26–28.Inthe polyps were observed, attached to the hard substrate. An orange phylogenetic tree, all three species clustered with known cold- taxon was found (Supplement Fig. 2b–c) growing on the wall of water and/or azooxanthellate corals. Furthermore, E. fistula and the ridge and a white taxon was observed (Supplement Fig. 2d) Dendrophyllia sp. could not be individually resolved but clustered with

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 4 www.nature.com/scientificreports

Figure 3 | Dendrophyllia sp, Dendrophylliidae. (a) Habitat view. (b) Colony view (scalebar in cm). (c) Lateral view of corallum. (d) and (e) Calicular view of corallum. (f) Close-up (SEM) of the theca at the top of the corallum. (g) Close-up (SEM) of the microstructure of the theca. Image a obtained with ROV camera. Image b by Roder/Voolstra. Images c, d, e, f, g by Bouwmeester. high bootstrap support with other members of the family carbon ratios were much lower in POM compared to corals Dendrophylliidae within the complex corals. Species A grouped with high (nearly 5%). bootstrap support with other members of the family Caryophyllidae within the robust corals. Discussion In accordance with what is known from the presence of deep-sea Metabolism. Shipboard-based incubations of the corals displayed an scleractinian corals elsewhere2,8,29, deep-sea corals in the Red Sea overall low metabolic activity (Fig. 6A). Respiration rates were not were exclusively encountered in areas displaying strong topography detectable for species A (Caryophyllidae) and lower for E. fistula than such as steep walls or seamounts and attached to or associated with for Dendrophyllia sp. Similarly, calcification rates were also reduced hard substrate, i.e. overhangs, rocks, edges, or faults. However, there in species A and E. fistula. Subsequent analyses of isotopic carbon are also striking differences between described habitats of deep-sea and nitrogen ratios in the coral tissues (Fig. 6B) revealed similar corals elsewhere and those where deep-sea corals occur in the Red ratios for all three species and both elements. While nitrogen Sea. The physical and biogeochemical properties of Red Sea habitats values of particulate organic matter (POM . 0.7 mm) collected fall well outside established ecosystem boundaries for deep-sea corals from 1 L of water in reef vicinity were similar to those of corals, in regard to temperature, oxygen availability, and salinity (Fig. 1,

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 5 www.nature.com/scientificreports

Figure 4 | Species A (undetermined), Caryophylliidae. (a) Habitat view. (b) Colony view (scalebar in cm). (c) Lateral view of corallum. (d) and (e) Calicular view of corallum. (f) and (g) Close-up (SEM) of the theca half way up the corallum and at the top of the corallum respectively. Image a obtained with ROV camera. Image b by Roder/Voolstra. Images c, d, e, f, g by Bouwmeester.

Table 1). Temperatures in the deep Red Sea are as high as in some water is rich in dissolved inorganic carbon compounds (resulting tropical shallow reef communities and nearly 10uC higher than the from remineralization processes13) cold-water corals typically exist next-warmest known habitat of deep-sea corals2. Additionally, salin- in locations that are near the lower limits of aragonite saturation ity levels in deep-sea reefs of the Red Sea are higher than the highest tolerance for calcifying organisms (undersaturation is defined as V previously recorded salinity for deep-sea coral communities (38.8 5 1.0)14. In contrast, alkalinity and aragonite saturation levels in the PSU)6. Studies of reefs built by Lophelia pertusa, perhaps the most Red Sea are high, even at greater depths and potentially support coral well-studied deep-sea coral, have documented oxygen concentra- calcification. Changes in aragonite saturation can have profound tions 3 to 5 times higher than what we recorded (ranging between impacts on marine calcifiers (including corals)14 and might be as 5.7 – 10.3 mg l21)29. Most importantly, oxygen concentrations below important as temperature, especially in regard to global climate 4mgl21 (3 ml l21) have been shown to inhibit aerobic metabolism in change. Factors influencing carbonate chemistry in the Red Sea deep-sea corals, and accordingly deep-sea corals in the Red Sea must might be the high salinity30, the warm temperature13, or a combina- have evolved adaptations to cope with these low levels, e.g. a highly tion of the two and possibly other parameters. Because mechanisms reduced gross metabolic rate18. As atmosphere-segregated deep-sea driving aragonite saturation are complex especially in continental

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 6 www.nature.com/scientificreports

Figure 5 | Molecular phylogram of relations among 24 species of corals and Hydra vulgaris (outgroup) as determined by Bayesian, Maximum Likelihood, and Neighbor-Joining analyses. This phylogram was generated on the basis of sequences from the mitochondrial 16S ribosomal gene region26–28. The numbers on the branches represent values from 1,000 bootstrap replicates presented in the order of Bayesian/Maximum Likelihood/ Neighbor-Joining. Azooxanthellate species are denoted by $, Red Sea species by ***. NJ: Neighbor-Joining. Images by Roder/Voolstra. shelf or marginal seas31, it remains difficult to speculate whether the Red Sea are lower than in temperate oceans, the latter with unexplored habitats with similar conditions exist outside the Red decreasing quality over depth (i.e., nitrogen becomes quickly Sea that might harbor as-yet undetected deep-sea corals. depleted compared to organic carbon) as would be expected for Supporting the hypothesis that food availability is low, we decaying organic material32. Moreover, the relatively low primary observed very low abundances of other deep-water filter feeders production in Red Sea waters19 seems at odds with the assumption common in temperate waters: Only few specimens of anemones that deep-sea coral nutrition is highly dependent on food transport and sponges (in E. fistula coral gardens), or sponges and octocorals from productive surface waters to less productive deep waters4. (attached to rocks in the habitats of Species A and Dendrophyllia sp.) Of the three collected coral species only one (Eguchipsammia could be observed at our study sites. Indeed, concentrations of inor- fistula) could be identified to the species level. For the two other ganic nutrients and suspended matter in subthermocline waters of species (Dendrophyllia sp. and species A [Caryophyllidae]) we were

Figure 6 | (A) Calcification (G) and respiration (R) rates (6 SE) of three coral species found in the deep Red Sea (Eguchipsammia fistula, species A (Caryophyllidae), Dendrophyllia sp.). (B) Stable isotope measurements of d13C and d15N in the particulate organic matter (POM) fraction of the total suspended material in reef water from the Red Sea and of the tissue of the three coral species found in the Red Sea.

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 7 www.nature.com/scientificreports not able to determine the specimens to the species level. We In summary, the existence of deep-sea coral communities in the consistently had exclusive encounters of all species in a distinct micro- warm, saline, oxygen- and food-deprived environment of the Red Sea habitat indicative of niche partitioning. E. fistula forms monospecific extends ecosystem-boundaries defined by current knowledge. We coral gardens on seamount tops. Species A is found on hard or sandy propose that the corals’ survival is a combined result of maximizing substrates close to rocky structures. We cannot exclude that corals were available resources while minimizing metabolic demands. They originally attached to nearby rocks and continued to grow on the soft appear to maintain minimal amounts of tissue to survive and have bottom after breaking off, or that the hard substrate originally used for strongly reduced respiration rates, which may be some of the key settling became gradually covered with a fine layer of sediment and adaptations to meet the metabolic challenges of low oxygen and high therefore was not visible. Dendrophyllia sp. was found mainly attached temperature. Calcification is facilitated under conditions of high upside down from hanging walls. It remains to be determined whether temperature and aragonite saturation, but it may be reduced in or how these distinct niches relate to adaptations for living in the deep-sea corals in the Red Sea as a consequence of low oxygen levels challenging conditions in the deep Red Sea. and scarce food availability. E. fistula, among the first deep-sea corals found in the Red Sea22,is Our findings suggest that existing theories on deep-sea coral eco- also known to occur in the Indo-Pacific, Australia, and New logy need to be re-examined. Clearly, not all species are dependent on Zealand33. This indicates that larval connectivity between the Red cold-water biotopes. Rather, the deep-sea corals of the Red Sea Sea and the Indian Ocean exists (or existed), and that E. fistula is or expand the known physical and chemical boundaries of deep-sea was connected via the Bab-el-Mandeb34. Genetic connectivity studies corals and provide new insights into understanding environmental between specimens from the Indo-Pacific and Red Sea should help to constraints and controls in these ecosystems. The Red Sea constitutes verify this observation. a marginal habitat that allows for the study of adaptation of corals Deep-sea corals in the Red Sea show specific adaptations with very that thrive in cold, deep waters (e.g., the Indo-Pacific) as well as the low respiration and calcification rates in comparison to, e.g., L. per- warm and deep waters of the Red Sea. Understanding the ability of tusa, which has a respiration rate of about 0.5 mmol g21 h2118and a these ecosystems to adapt and function is paramount given that calcification rate of up to 1% of its entire weight per day35. global climate change is significantly altering natural environments Interestingly, in all species analyzed, only polyps were alive (i.e., at an unprecedented rate. the distal end of the corals), whereas the rest of the skeleton was not covered with tissue. This might be one of the adaptations to living Methods in warm and deep waters by minimizing metabolic needs. Higher Study sites. During an expedition (December 1 - 13, 2011) on the R/V Aegaeo (operated by the Hellenic Center for Marine Research, Greece) in the central Red Sea, respiration rates of Dendrophyllia sp. compared to the two other we explored four sites (Table 1, Figure 1). The sites were characterized using various species might be explained with regard to their habitat. It was mainly combinations of approaches (detailed in Table 1): bathymetric seafloor mapping encountered hanging from rocks into the open water column, (Seabeam 2100), conductivity-temperature-depth (CTD) profiling, water sampling, a whereas species A and E. fistula were found to be bottom dwellers remotely operated vehicle (ROV) (Max Rover), and a manned submersible (THETIS, Comex S.A.) with likely less exposure to resources, and hence with greater food Multibeam sonar was used to identify specific sites with suitable topographic relief and oxygen depletion. Similarly, calcification rates were low in all (i.e., seamount-like or steep-walled habitats) as candidates for further investigation species despite the high aragonite saturation levels, but even lower in using the ROV and submersible. species A and E. fistula in comparison to Dendrophyllia sp. This ROV video transects were recorded using a SEA MAX universal underwater video camera with pan and tilt and three modules (two wide angle and one zoom color video indicates that other factors, such as oxygen supply, might limit the features). This camera was deployed to record continuously. We also used two High- growth of deep-sea corals in the Red Sea. It will be interesting for Definition Ocean ProHD Undersea cameras with pan and tilt (with wide angle and future studies to determine age and growth rate of deep-sea corals of zoom features), which were utilized only at sites of interest. The manned submersible the Red Sea. Our initial results suggest that deep-sea corals in the Red was similarly equipped with a high-resolution video camera (XC 999 P). Subsequent video analyses identified characteristics of deep-sea coral habitats. A total of 33 hours Sea seem to compensate for the lack of nutrients and oxygen through of video footage was analyzed. greatly reduced calcification and respiration rates. Interestingly, isotopic signatures of coral tissue and particulate Environment of the deep Red Sea. A CTD (SBE 37-SMP) attached to the ROV organic matter (POM) indicate that deep-sea corals in the Red monitored temperature in the sampling areas at all times, while a Niskin was entrained three times at each site to the ROV to sample water directly from the Sea do not exclusively feed on POM present in the vicinity of the benthic boundary layer for measurements of total alkalinity (TA) and aragonite reef. Nitrogen signatures have been shown to be enriched by 2.5– saturation (Varag), inorganic nutrient (nitrate 1 nitrite), and total suspended matter 3.5% in L. pertusa and Madrepora oculata compared to local (TSM) contents, as well as stable isotope ratios (d13C and d15N) of the organic fraction particulate organic matter (POM)36, whereas nitrogen signatures of the suspended material (details on methodology below). in our coral specimens were only slightly enriched (by about For each Niskin, 50 ml of water were carefully drained into sampling containers and immediate analyzed. Varag was calculated from TA, pH, temperature, and salinity 0.5%). It might be possible that deep-sea corals in the Red Sea using the CO2SYS Excel program41. Another 1 l of water from the Niskin was filtered utilize plankton not captured in the POM fraction as a food (200 mm Hg) on pre-combusted and pre-weighed filters (Whatman GF/F), which source (Red Sea euphotic zone plankton have a nitrogen signature were stored at 220uC until being lyophilized. TSM was determined (to 0.01 mg) on a 37 microbalance (Mettler Toledo). The filters were cut in half; while one half was of about 4.5% ) as the Red Sea is known to have extraordinary acidified (0.1 N HCl) to remove inorganic carbon for the determination of the stable zooplankton migrations with a subsurface zooplankton maximum isotope ratio of organic carbon (d13C), the other half was used for determination of the between 100–750 m38. However, carbon signatures recorded for nitrogen isotope ratio (d15N). the same Red Sea plankton ranged from 217 to 220%37 and as Isotope ratios were quantified with an isotope ratio mass spectrometer (Delta plus % XP, Thermo Finnigan) relative to Pee Dee Belemnite (PDB) standard and atmo- carbon levels are known to increase by about 0.5–1.0 per spheric nitrogen for carbon and nitrogen, respectively. As the exact weight of the 39 trophic level , the shift in our carbon data suggests that yet other sample was determined prior to analysis, the ratio of organic carbon and nitrogen food sources (or processes, e.g. microbial) might contribute to the (CN) could be determined via the coupled elemental analysis (yielding the weight corals’ diet. Moreover, a study on trophic food webs in the deep percent of carbon and nitrogen) during the isotope measurements. Inorganic nutrient concentrations of the filtrate (frozen at 220uC until analysis) were determined using sea found intricate and distinct trophic pathways, i.e. observed standard colorimetric tests in an autoanalyzer (QuickChem 8000, Zellweger Analytics large overlaps in isotopic signatures between the different food Inc.). web members were discussed as a result of competition for and In March 2012, we sampled the water column between sites 1 and 2 for physical and adaptation to limited food availability40. Considering the potential biogeochemical analyses. We repeatedly lowered a CTD (Ocean Seven 320 Plus, complexity of such deep-sea food webs, further work in the Red Idronaut) including probes for pH and oxygen and recorded measures of depth, temperature, salinity, oxygen, and pH between 0 and 800 m in one-second intervals. Sea system will be needed to unravel the nature of the trophic Additionally, water samples were collected over the same depth range via Niskin pathways these corals utilize. deployment (3 times) for subsequent analyses (all methods as above). TSM and

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 8 www.nature.com/scientificreports

corresponding CN ratios were measured for each two replicates at 1, 5, 20, 50 (only 3. Davies, A. J. & Guinotte, J. M. Global habitat suitability for framework-forming one replicate), 100, 200 (only one replicate), 300, 500, 600, 700, 800 m. Inorganic cold-water corals. PLoS One 6, e18483 (2011). nutrients (nitrate 1 nitrite) were measured in triplicates at 5, 20, 250, 500 (only two 4. Davies, A. et al. Downwelling and deep-water bottom currents as food supply replicates), 600 (only two replicates) m. TA and Varag were measured in triplicates at mechanisms to the cold-water coral Lophelia pertusa (Scleractinia) at the 5, 50, 100, 200, 300, 400, 500, 600 (only 2 replicates), 700, 800 m. Mingulay Reef complex. Limnol. Oceanogr. 54, 620–629 (2009). In May 2012, light penetration of the upper 300 m of the water column was 5. Rogers, A. D. The biology of seamounts. Adv. Mar. Biol. 30, 305–350 (1994). measured at the same locations using optical sensors (RAMSES ACC-VIS, TriOS) to 6. Freiwald, A., Fossa, J. H., Grehan, A. J., Koslow, T. & Roberts, J. M. Cold-water determine the maximum boundaries for light-dependent communities. In total, three coral reefs: out of sight - no longer out of mind. UNEP-WCMC Cambridge, UK casts were conducted and light spectra between 270 and 1000 nm wavelengths were (2004). measured at every 5 m. The boundary of the photic zone was determined as the depth 7. Tittensor, D. P., Baco, A. R., Hall-Spencer, J. M., Orr, J. C. & Rogers, A. D. beyond which the respective wavelength was less than 1% of the sunlight measured at Seamounts as refugia from ocean acidification for cold-water stony corals. Mar. the surface. Ecol. 31, 212–225 (2010). 8. Tittensor, D. P. et al. Predicting global habitat suitability for stony corals on Species observed. Deep-sea corals that were observed were classified to the highest seamounts. J. of Biogeogr. 36, 1111–1128 (2009). taxonomic resolution possible. Some samples were collected using the manipulator 9. Smith, J. E., Schwarcz, H. P., Risk, M. J., McConnaughey, T. A. & Keller, N. arm of the ROV and a sampling basket (Table 1) and were subsequently Paleotemperatures from deep-sea corals: overcoming ’vital effects’. Palaios 15, morphologically examined. One specimen from each collected species was bleached 25–32 (2000). in sodium hypochlorite for 12 hours to remove organic matter. Samples were further 10. Husebø, A., Nøttestad, L., Fossa˚, J. H., Furevik, D. M. & Jørgensen, S. B. assessed using traditional microscopy and scanning electron microscopy (SEM) and Distribution and abundance of fish in deep-sea coral habitats. Hydrobiologia 471, identified to the closest taxon. The skeleton samples were left uncoated for analysis by 91–99 (2002). SEM on a Quanta 200 FEG SEM at low vacuum. Eguchipsammia fistula was 11. Rocha, J., Peixe, L., Gomes, N. C. M. & Calado, R. Cnidarians as a source of new identified according to descriptions from Alcock42 and Marenzeller22, whose marine bioactive compounds - an overview of the last decade and future steps for taxonomy was later updated by Cairns43. Dendrophyllia sp. was identified with the bioprospecting. Marine Drugs 9, 1860–1886 (2011). help of Cairns25 and species A (Caryophylliidae) was run through an identification 12. Glover, A. G. & Smith, C. R. The deep-sea floor ecosystem: current status and key to the genera of azooxanthellate scleractinia44, but could not be identified further prospects of anthropogenic change by the year 2025. Environ. Conserv. 30, 219– than the family level as morphological characters did not further match any of those 241 (2003). of the key, and no species or genera descriptions from various earlier expeditions 13. Orr, J. et al. Anthropogenic ocean acidification over the twenty-first century and matched our samples. Rhizotrochus typus was identified according to Scheer and its impact on calcifying organisms. Nature 437, 681–686 (2005). Pillai45. For other non-collected corals we can only report limited taxonomic 14. Guinotte, J. M. et al. Will human-induced changes in seawater chemistry alter the 4 resolution based on video imagery. distribution of deep-sea scleractinian corals? Front. Ecol. Environ. , 141–146 (2006). We also utilized molecular approaches to determine taxonomic relationships 15. Hughes, T. P. et al. Climate change, human inmpacts, and the resilience of coral where possible (samples evaluated after incubations, details of sample handling in reefs. Science 301, 929–933 (2003). next paragraph). From the three species we collected, DNA was extracted using the 16. Levitus, S., Antonov, J. I., Boyer, T. P. & Stephens, C. Warming of the World Qiagen DNeasy Plant Mini Kit following instructions from the manufacturer. 10 – Ocean. Science 287, 2225–2229 (2000). 30 ng of DNA was used in a 30 ml PCR reaction using the Qiagen Multiplex PCR kit 17. Mienis, F. et al. The influence of near-bed hydrodynamic conditions on cold- and primers LP16SF/LP16SR27. The temperature cycling for amplification were 1 water corals in the Viosca Knoll area, Gulf of Mexico. Deep Sea Res., Part I 60,32– cycle at 95uC for 15 min; 30 cycles at 95uC for 30 sec, 55uC for 90 sec, and 72uCfor 45 (2012). 90 sec; and one final extension at 72uC for 30 min. PCR products were subsequently 18. Dodds, L. A., Roberts, J. M., Taylor, A. C. & Marubini, F. Metabolic tolerance of purified using the Qiagen MinElute PCR purification kit according to manufacturer’s the cold-water coral Lophelia pertusa (Scleractinia) to temperature and dissolved instructions, cloned into Qiagen pDrive, and sequenced on an ABI 3730 XL. oxygen change. J. Exp. Mar. Biol. Ecol. 349, 205–214 (2007). Sequences were clipped, trimmed, and assembled with CodonCode Aligner. All 19. Edwards, A. J. & Head, S. M. Key Environments: Red Sea. Pergamon Books Ltd. assemblies yielded single contigs that are accessible in GenBank: Eguchipsammia (1987). fistula [JX629250], Dendrophyllia sp. [JX629248], species A (Caryophylliidae) 20. Berumen, M. L. et al. The status of coral reef ecology research in the Red Sea. Coral [JX629249]. Alignment of partial mitochondrial 16S ribosomal RNA genes was Reefs dx.doi.org/10.1007/s00338-013-1055-8 (2013). generated with MUSCLE as implemented in MEGA546. All sites with missing data 21. Fricke, H. W. & Schuhmacher, H. The depth limits of Red Sea stony corals: an were deleted, which yielded 247 characters. Neighbor-Joining and Maximum ecophysiological problem (a deep diving survey by submersible). Mar. Ecol. 4, Likelihood trees were generated with MEGA using the T92 substitution model and a 163–194 (1983). discrete gamma distribution with 5 categories as determined from the model test 47 22. Marenzeller, E. in Expeditionen S.M. Schiff "Pola" in das Rote Meer, noerdliche und function implemented in MEGA. Bayesian analysis was performed with MrBayes suedliche Haelfte 1895-1898. Berichte der Komission fuer ozeanographische 5 and the nst 1 substitution model using a discrete gamma distribution and 5 Forschungen (1906). categories. MrBayes was set to run until convergence using the stop rule and a value of 23. Fricke, H. W. & Hottinger, L. Coral bioherms below the euphotic zone in the Red , 0.01 for the standard deviation of split frequencies. 1,000 bootstrap replicates were Sea. Mar. Ecol. Prog. Ser. 11, 113–117 (1983). conducted for all analyses. 24. Fricke, H. W. & Knauer, B. Diversity and spatial pattern of coral communities in the Red Sea upper twilight zone. Oecologia 71, 29–37 (1986). Metabolic measurements. We collected specimens of three species (Table 1). 5 25. Cairns, S. A Generic Revision And Phylogenetic Analysis Of The Dendrophylliidae samples per species were assessed in aquaria onboard the ship after collection. After (Cnidaria: Scleractinia). Smithonian Insitution Press (2001). acclimatization of a minimum 4 hours in opaque tanks filled with freshly collected 26. Romano, S. & Palumbi, S. Evolution of scleractinian corals inferred from seawater, we incubated each specimen in 1 l beakers for 60 – 80 minutes. The running molecular systematics. Science 271, 640–642 (1996). of the vessel engine ensured a constant moderate stirring of the water in the tanks and 27. Le Goff-Vitry, M., Rogers, A. & Baglow, D. A deep-sea slant on the molecular beakers. The temperature in the tanks was monitored and remained constant at phylogeny of the Scleractinia. Mol. Phylogenet. Evol. 30, 167–177 (2004). around 21 6 0.5uC. Control incubations were run with beakers containing only water. 28. Stolarski, J. et al. The ancient evolutionary origins of Scleractinia revealed by Before and after incubation, the oxygen content of the beakers was measured using an azooxanthellate corals. Evol. Biol. 11, 361 (2011). oxygen probe (Multi 3500i, WTW). Water from the beakers was also sampled for total 29. Davies, A. J., Wisshak, M., Orr, J. C. & Murray Roberts, J. Predicting suitable alkalinity using the Gran approximation by determining the second endpoint48 via habitat for the cold-water coral Lophelia pertusa (Scleractinia). Deep Sea Res., Part automatic potentiometric titration (using 0.01 N HCl) on a titration apparatus I 55, 1048–1062 (2008). (Titrando 888, Metrohm). Calcification rates were derived via the alkalinity anomaly 30. Lee, K. et al. Global relationships of total alkalinity with salinity and temperature 49 technique based on the 251 stoichiometric relationship between TA and CaCO3 in surface waters of the world’s oceans. Geophys. Res. Lett. 33, http://dx.doi.org/ using automatic potentiometric titration. Changes in the control incubations were 10.1029/2006gl027207 (2006). less than 0.05% and therefore not included in subsequent calculations. Respiration 31. Blackford, J. C. Predicting the impacts of ocean acidification: Challenges from an and calcification rates were obtained by subtracting values obtained after incubation ecosystem perspective. J. Mar. Syst. 81, 12–18 (2010). from values obtained before incubation. 32. Knauer, G. H., Martin, J. H. & Bruland, K. W. Fluxes of particulate carbon, After incubation, coral samples were flash-frozen in liquid nitrogen and the polyps nitrogen, and phosphorus in the upper water column of the northeast Pacific. were ground to a fine powder for tissue analyses. An aliquot (approximately 3 mg) Deep Sea Res., Part I 26, 97–108 (1979). was acidified with 0.1 N HCl to remove inorganic carbon for the determination of the 33. van der Land, J. UNESCO-IOC Register of Marine Organisms (URMO) (2008). stable isotope ratio of organic carbon (d13C); another aliquot of the same amount was 34. Taviani, M. et al. Last glacial deep-water corals from the Red Sea. Bull. Mar. Sci. 81, directly used for the measurement of organic nitrogen (d15N). 361–370 (2008). 35. Maier, C., Hegeman, J., Weinbauer, M. G. & Gattuso, J. P. Calcification of the cold- water coral Lophelia pertusa under ambient and reduced pH. Biogeosciences 6, 1. Cairns, S. Species richness of recent scleractinia. Atoll Res. Bull. 459, 1–12 (1999). 1671–1680 (2009). 2. Roberts, J. M., Wheeler, A. J. & Freiwald, A. Reefs of the deep: the biology and 36. Kiriakoulakis, K. et al. Lipids and nitrogen isotopes of two deep-water corals from geology of cold-water coral ecosystems. Science 312, 543–547 (2006). the North-East Atlantic: initial results and implications for their nutrition. In

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 9 www.nature.com/scientificreports

Cold-water Corals And Ecosystems (eds Freiwald, A. & Roberts, J. M.) 715–729 49. Schneider, K. & Erez, J. The effect of carbonate chemistry on calcification and Springer Verlag (2005). photosynthesis in the hermatypic coral Acropora eurystoma. Limnol. Oceanogr. 37. McMahon, K. W., Berumen, M. L., Mateo, I., Elsdon, T. S. & Thorrold, S. R. 51, 1284–1293 (2006). Carbon isotopes in otolith amino acids identify residency of juvenile snapper (Family: Lutjanidae) in coastal nurseries. Coral Reefs 30, 1135–1145 (2011). 38. Weikert, H. The vertical distribution of zooplankton in relation to habitat zones in the area of the Atlantis II Deep, central Red Sea. Mar. Ecol. Prog. Ser. 8, 129–143 Acknowledgements (1982). We thank the crew of the R/V Aegaeo, especially the ROV and submersible team. We thank 39. McCutchan, J. H., Lewis, W. M., Kendall, C. & McGrath, C. C. Variation in trophic H. Zibrowius and F. Benzoni for assistance in taxonomic assignment of the coral specimens, shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102, 378–390 Ali R Behzad from the advanced nanofabrication imaging and characterization core lab at (2003). KAUST for assistance with SEM, and X. Irigoien, P. Magistretti, S. Catsicas, and J. Calvin for 40. Iken, K., Brey, T., Wand, U., Voigt, J. & Junghans, P. Food web structure of the helpful comments on the manuscript. The manuscript furthermore benefitted from benthic community at the Porcupine Abyssal Plain (NE Atlantic): a stable isotope comments of two anonymous reviewers. Sequences have been deposited in NCBI GenBank analysis. Progr. Oceanogr. 50, 383–405 (2001). under accession numbers [JX629250] (Eguchipsammia fistula), [JX629248] (Dendrophyllia sp.), [JX629249] (species A (Caryophylliidae)). 41. Lewis, E. & Wallace, D. W. R. Program Developed for CO2 System Calculations. ORNL/CDIAC-105. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, US Department of Energy, Oak Ridge, Tennessee (1998). Author contributions 42. Alcock, A. Report on the deep-sea Madreporaria of the Siboga-Expedition. M.L.B., C.R., C.R.V. conceived the research cruise. C.R. and C.R.V. designed and conducted Siboga-Expeditie 16a: 52pp., 5pls. (1902). the metabolic and isotopic measurements. C.R., C.R.V., J.B. generated data. C.R., C.R.V., 43. Cairns, S. Scleractinia Of The Temperate North Pacific. Smithsonian Institution J.B., M.L.B. wrote the article, analyzed and interpreted data. E.P. and A.A.S. provided Press, Washington D.C. (1994). materials, reagents, and logistical support. All authors critically read the article and 44. Cairns, S. D. & Kitahara, M. V. An illustrated key to the genera and subgenera of approved the final version. the recent azooxanthellate Scleractinia (Cnidaria, Anthozoa), with an attached glossary. ZooKeys 1–47 (2012). 45. Scheer, G. & Pillai, C. S. G. Report on the stony corals from the Red Sea. Zoologica Additional information 45, (1983). Supplementary information accompanies this paper at http://www.nature.com/ 46. Tamura, K. et al. MEGA5: Molecular Evolutionary Genetics Analysis Using scientificreports Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Competing financial interests: The authors declare no competing financial interests. Methods. Mol. Biol. and Evol. 28, 2731–2739 (2011). How to cite this article: Roder, C. et al. First biological measurements of deep-sea corals 47. Ronquist, F. & Huelsenbeck, J. P. MrBayes 3: Bayesian phylogenetic inference from the Red Sea. Sci. Rep. 3, 2802; DOI:10.1038/srep02802 (2013). under mixed models. Bioinformatics 19, 1572–1574 (2003). 48. Grasshoff, K., Ehrhardt, M. & Kremling, K. Methods Of Seawater Analysis. Verlag This work is licensed under a Creative Commons Attribution 3.0 Unported license. Chemie Weinheim, New York (1983). To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0

SCIENTIFIC REPORTS | 3 : 2802 | DOI: 10.1038/srep02802 10 152

SUPPLEMENTARY DATA

First biological measurements of deep-sea corals from the Red Sea

C. Roder1*, M. L. Berumen1,2, J. Bouwmeester1, E. Papathanassiou3, A. Al-Suwailem4, C. R. Voolstra1*

1Red Sea Research Center, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia 2Biology Department, Woods Hole Oceanographic Institution, Woods Hole, USA 3Hellenic Centre for Marine Research, Anavissos, Greece 4Coastal and Marine Resources Core Lab, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia

*Correspondence to: [email protected], [email protected].

Running title: Biology of Red Sea deep-sea corals

153

Supplement Fig. 1: Map of study area and specific study sites for deep-sea corals in the central Red Sea. Insets show multibeam bathymetry where further surveys were conducted by ROV and manned submersible. Depth (m) is indicated by the color gradient as shown in the key. The grey line indicates the route followed by a remotely operated vehicle (ROV) for detailed habitat surveys and sample collections. 154

The background map was created with ArcGIS 10.1 (Esri). Bathymetry maps were created with Fledermaus 7 Visualization and Analysis Software (Ageotec) and ArcGIS 10.1 (Esri). Editing was done using Adobe Illustrator CS3 13.0.0 and Adobe Photoshop CS3 10.0.

Supplement Fig. 2: Images (best resolution available) extracted from the ROV HD video recordings of other specimens observed. a Rhizotrochus typus in a field of dead skeletons of Eguchipsammia fistula (see Fig. 3), live specimen top-right and dead specimen bottom-left. b and c Orange solitary coral attached to a seamount wall. d White solitary corals hanging under a ridge.

Coral Reefs (2013) 32:737–748 DOI 10.1007/s00338-013-1055-8

REVIEW

The status of coral reef ecology research in the Red Sea

M. L. Berumen • A. S. Hoey • W. H. Bass • J. Bouwmeester • D. Catania • J. E. M. Cochran • M. T. Khalil • S. Miyake • M. R. Mughal • J. L. Y. Spaet • P. Saenz-Agudelo

Received: 23 February 2013 / Accepted: 6 June 2013 / Published online: 21 June 2013 Ó Springer-Verlag Berlin Heidelberg 2013

Abstract The Red Sea has long been recognized as a than 50 % of the published research from the Red Sea region of high biodiversity and endemism. Despite this originated from the Gulf of Aqaba, a small area (\2%of diversity and early history of scientific work, our under- the area of the Red Sea) in the far northern Red Sea. We standing of the ecology of coral reefs in the Red Sea has summarize the general state of knowledge in these eight lagged behind that of other large coral reef systems. We topics and highlight the areas of future research priorities carried out a quantitative assessment of ISI-listed research for the Red Sea region. Notably, data that could inform published from the Red Sea in eight specific topics (apex science-based management approaches are badly lacking in predators, connectivity, coral bleaching, coral reproductive most Red Sea countries. The Red Sea, as a geologically biology, herbivory, marine protected areas, non-coral ‘‘young’’ sea located in one of the warmest regions of the invertebrates and reef-associated bacteria) and compared world, has the potential to provide insight into pressing the amount of research conducted in the Red Sea to that topics such as speciation processes as well as the capacity from Australia’s Great Barrier Reef (GBR) and the of reef systems and organisms to adapt to global climate Caribbean. On average, for these eight topics, the Red Sea change. As one of the world’s most biodiverse coral reef had 1/6th the amount of research compared to the GBR and regions, the Red Sea may yet have a significant role to play about 1/8th the amount of the Caribbean. Further, more in our understanding of coral reef ecology at a global scale.

Keywords Apex predators Á Connectivity Á Communicated by Biology Editor Dr. Hugh Sweatman Coral bleaching Á Coral reproduction Á Herbivory Á Electronic supplementary material The online version of this Marine protected area Á Porifera Á Reef-associated bacteria article (doi:10.1007/s00338-013-1055-8) contains supplementary material, which is available to authorized users.

M. L. Berumen (&) Á A. S. Hoey Á W. H. Bass Á Introduction J. Bouwmeester Á D. Catania Á J. E. M. Cochran Á M. T. Khalil Á S. Miyake Á M. R. Mughal Á The Red Sea has long been recognized as a region of high J. L. Y. Spaet Á P. Saenz-Agudelo biodiversity (Stehli and Wells 1971) and endemism Red Sea Research Center, King Abdullah University of Science and Technology, Thuwal, Jeddah 23955, (Ormond and Edwards 1987), home to well over 1,000 Kingdom of Saudi Arabia species of fishes and over 50 genera of hermatypic corals. e-mail: [email protected] The Red Sea coral reefs were some of the first to be examined by early European taxonomists (e.g. Forsska˚ll, M. L. Berumen Biology Department, Woods Hole Oceanographic Institution, Ehrenberg, Ruppell, Klunzinger), but have become rela- 266 Woods Hole Rd, Woods Hole, MA 02543, USA tively inaccessible to Western scientists in recent decades due to complicated visa and permitting regulations, cou- A. S. Hoey pled with a lack of marine research infrastructure. Despite Australian Research Council Center of Excellence for Coral Reef Studies, James Cook University, Townsville, QLD 4811, the size of the Red Sea and diversity of its reef-associated Australia inhabitants, it remains a poorly studied system compared to 123 738 Coral Reefs (2013) 32:737–748 other large coral reef systems around the world (e.g. the making a push to invest heavily in education and research Great Barrier Reef or the Caribbean). The Red Sea, how- institutions in the region (e.g. Mervis 2009), creating an ever, is of increasing interest to scientists working on cli- opportunity for renewed and expanded research on coral mate change due to its relatively high and variable water reefs in the wider Red Sea. It is our hope that this review temperatures (from 20 °C in spring to 35 °C in summer) can highlight some areas in which these research efforts and high salinity (c. 40.0 psu in the northern Red Sea; could quickly address important gaps in our knowledge. Edwards 1987). Consequently, the Red Sea is an ideal model system for understanding how reefs may fare under predicted scenarios of global climate change. It is arguably General methodology the world’s warmest and most saline habitat in which extensive reef formation occurs. To quantify and compare the status of globally accessible Despite this biodiversity, early history of scientific work research in the Red Sea, we identified eight areas of and the potential importance for understanding the ability research focus of current interest to coral reef scientists of reefs to adapt and/or acclimate to changing environ- (hereafter we will refer to these as ‘‘topics’’), namely apex mental conditions, there is a current lack of understanding predators, connectivity, coral bleaching, coral reproduc- of the ecology of Red Sea coral reefs. Researchers wishing tion, herbivory, marine protected areas, non-coral inverte- to access information about the ecology of the Red Sea via brates (primarily sponges) and reef-associated bacteria. We modern channels (e.g. ISI or Google Scholar searches) will then quantified the number of published studies in each of find relatively little information, particularly in recent these topics using Thomson Reuter’s Web of Knowledge years. Further, the vast majority of the accessible published (WoK) and used various regional restrictors to compare the research originates from a *6 km stretch of coastline in number of studies conducted within the Red Sea to that of the far northern Red Sea, within the Gulf of Eilat/Aqaba the Great Barrier Reef and the Caribbean (Table 1; and see (hereafter Gulf of Aqaba). We estimate that over half of the Electronic Supplementary Material, ESM, for details). recent research in the Red Sea originates from this region although it represents\2 % of the Red Sea area. There are many reasons to suspect that the processes in this relatively Status of ecological research in the Red Sea small and isolated region may not be representative of the broader Red Sea ‘‘proper’’ (sensu Head 1987). Apex predators This review seeks to quantify the number of ecological studies conducted in the Red Sea and compares this to the It is well documented that top predators are the most number of studies conducted in two other large coral reef heavily exploited of marine organisms (Myers et al. 2007; systems, the Great Barrier Reef of Australia and the Carib- Collette et al. 2011). The collapse of fisheries targeting bean. Specifically, the objective of this review is to highlight large-bodied, long-lived, predatory teleosts and sharks, the lack of research in the Red Sea by examining eight topics however, suggests that it is unlikely that many of these of current ecological interest in other coral reef systems and species can be sustainably harvested at a commercially comparing the current state of knowledge in these topics viable level without strict management (Walker 1998; among three distinct biogeographic regions: the Red Sea, the Hilborn 2007). Effective management can only be achieved Caribbean and the Great Barrier Reef (GBR). The coral reef through a thorough understanding of the ecology and area within the Red Sea (c. 8,890 km2) is broadly comparable population biology of these species. Unfortunately, for to that of the Caribbean (c. 10,530 km2). Although the most Red Sea fisheries-targeted species, this information is reef area is only approximately half that of the GBR completely lacking. (c. 17,400 km2), these two reef systems are of similar length Our WoK search revealed only 39 Red Sea studies (Red Sea: c. 2,000 km; GBR: c. 2,300 km), making the Red ([60 % of which were from the Gulf of Aqaba), compared Sea one of the longest coral reef systems in the world. to 85 from the Great Barrier Reef and 206 from the We present below assessments of various disciplines Caribbean (Table 1, ESM 1). An analysis of the few studies falling broadly under the heading of ‘‘coral reef ecology’’. available on Red Sea top predators showed that their We do not seek to comprehensively review details of all objectives, with very few exceptions (e.g. Shpigel and Fi- relevant research in these various disciplines, but rather to shelson 1991; Belmaker et al. 2005; Clarke et al. 2011), quantify the relative magnitude of work in the Red Sea as focused on the physiology of a very limited number of compared to other regions. Where appropriate, brief sum- species (e.g. Fishelson and Baranes 1997; Karpestam et al. maries of the primary focus of the work within each topic 2007). In some cases, the principal focus was on other highlight areas where the Red Sea knowledge base is species, such as ospreys (Fisher et al. 2001), sponges particularly weak. A number of Middle East countries are (Burns and Ilan 2003), humans (Randall and Levy 1976)or 123 Coral Reefs (2013) 32:737–748 739

Table 1 Summary results of the number of publications found using standard Boolean operators ‘‘OR’’ and ‘‘AND’’ along with the Web of Knowledge search engine for various topics and ﬁltered for wildcard symbol ‘‘*’’ to include truncations of the search term speciﬁc comparative regions of interest. Search terms included Topic Red Sea Great barrier reef Caribbean All ESM Search terms All Excluding Eilat/Aqaba

Apex predators 40 15 85 206 2,014 1 ‘‘Predatory fish*’’ OR shark* Connectivity 77 41 471 670 26,840 2 [(Structure or connectivity) and (genetic* or population*) marine] or (phyloge*) and (marine) not ‘‘community structure’’ Coral bleaching 12 5 339 167 1,182 3 (‘‘Coral bleaching’’ or ‘‘bleaching event’’) Coral reproduction 11 2 197 108 603 4 Scleractinia* reproducti* Herbivory 34 19 322 255 1,066 5 (Herbiv* AND coral-reef OR Graz* AND coral reef) Marine protected areas 17 10 220 152 2,247 6 (‘‘Marine reserve*’’ or ‘‘marine protected area*’’ or ‘‘no take zone*’’ or ‘‘no entry zone*’’) and (‘‘fisheries*’’ or ‘‘spillover*’’ or ‘‘catch*’’ or ‘‘fishing*’’) Porifera 23 5 88 170 18,570 7 Sponge* OR porifer* AND ‘‘coral reef’’ Reef-associated bacteria 41 8 342 166 1,555 8 Reef bacteria*

Red Sea results are separated into all relevant Red Sea results and results exclusive of studies conducted only in the Gulf of Aqaba/Gulf of Eilat. Full details of the records for each search are included online as Electronic Supplementary Material (ESM) with the respective ESM file number indicated in the table parasites (Ivanov and Lipshitz 2006), and only relates Research priorities should include species verification and peripherally to the top predators. Interestingly, this is the areas of origin of all species taken directly or as bycatch in only topic of the eight examined for which the Red Sea fisheries. There is a clear need to identify spatially and actually had more published research than the GBR up temporally sensitive areas, such as nurseries or seasons of until the late 1990s, but has subsequently fallen behind high bycatch. Some of the larger, more mobile apex pre- (Fig. 1a). dators almost certainly cross national borders on a regular There is a current lack of data on the population ecology, basis and potentially even have connections with Indian reproductive biology, resource partitioning and migration Ocean populations. Any attempts at management for these patterns of apex predators in the Red Sea. This lack of species will have to secure multinational cooperation—a management-relevant data poses a major obstacle for suc- non-trivial undertaking in this region of the world. cessful conservation planning. The GBR, which is also lacking essential data about the population status of its top Connectivity predators, is nonetheless one of the most intensely managed reef systems in the world (McCook et al. 2010). While Connectivity is a broad term that in the marine realm declines of top predator populations are more likely to be usually refers to ‘‘the demographic linking of local popu- noticed in this type of well-managed system (Robbins et al. lations through the dispersal of individuals among them as 2006), they would most likely go unnoticed in the Red Sea. In larvae, juveniles or adults’’ (Sale et al. 2005). Coral reefs the Caribbean, the ecology of many predatory species has are naturally fragmented, discontinuous habitats, and the received considerable attention (e.g. Lutcavage et al. 1997; extent to which reef-associated species exchange individ- Pikitch et al. 2005) and many marine reserves have been uals among spatially discrete populations has implications established (Bond et al. 2012). In addition, apex predators are for their persistence, resilience and recovery (Jones et al. of high value to Caribbean and Australian tourism industries 2009). Connectivity is now recognized as an important and are valued socially and culturally. This does not appear parameter informing conservation and management deci- to be the case in most of the Red Sea countries. sions (e.g. Sale et al. 2005; Almany et al. 2009; McCook If Red Sea marine ecosystems, and the goods and ser- et al. 2009). Studies dealing with connectivity in coral reefs vices that humans derive from them, are to be sustained, have dramatically proliferated in the last decade (Jones future work will have to focus on diversity, abundance and et al. 2009), and have been fuelled by the development of life histories of top predator species within the Red Sea. new genetic techniques (Lowe and Allendorf 2010). Yet,

123 740 Coral Reefs (2013) 32:737–748

Fig. 1 Cumulative number of 250 abApex Predators 700 Connectivity Web of Knowledge listed 600 publications through time for 200 GBR 500 various topics in coral reef 150 ecology from three regions: Red Caribbean 400 Sea (red line), Australia’s Great 100 Red Sea 300 Barrier Reef (blue line) and 200 50 Caribbean (green line). Topics 100 were searched on 0 0 apps.webofknowledge.com 1940 1947 1954 1961 1968 1975 1982 1989 1996 2003 2010 2011 2011 using speciﬁc search terms 1955 1962 1969 1976 1983 1990 1997 2004 (see Table 1) and reﬁned by the respective region 350 cdCoral Bleaching 350 Reef-Associated Bacteria 300 300 250 250 200 200 150 150 100 100 50 50 0 0 2011 2011 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 1966 1973 1980 1987 1994 2001 2008

200 e Coral Reproduction 700 f Invertebrates / Sponges 600 150 500 400 Number of publications in Web Knowledge 100 300 50 200 100 0 0 2011 2011 1976 1983 1990 1997 2004 1935 1942 1949 1956 1963 1970 1977 1984 1991 1998 2005

350 ghHerbivory 200 Marine Protected Areas 300 250 150 200 100 150 100 50 50 0 0 1974 1981 1988 1995 2002 2009 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 the number of studies on connectivity within the Red Sea is majority of genetic studies were mostly phylogenies that minimal compared to the research that has been published included a few samples from the Red Sea as part of the from the GBR or the Caribbean (Fig. 1b). phylogenetic trees but that did not actually measure gene flow Our WoK search revealed only 77 Red Sea studies, com- or connectivity within this region. pared to 471 from the Great Barrier Reef and 670 from the Our search revealed only eight studies on fishes in the Caribbean (Table 1, ESM 2). We were primarily seeking Red Sea. Among these eight studies, two used otolith research that clearly studied connectivity or population chemistry to estimate connectivity among habitats, while genetics and explicitly excluded studies only focusing on the remaining six studies used genetic markers to discern community structure. While many studies utilized genetic population structure. Fish are perhaps one of the best- approaches for taxonomic purposes (especially for microbes), studied groups in coral reefs in terms of phylogeography studies related to population connectivity and population and population connectivity (Jones et al. 2009). Yet, in the genetics were rare. Of the 77 Red Sea connectivity studies, Red Sea, this field of study remains poorly addressed. only 13 papers measured gene flow among populations within The unique geologic history of the Red Sea and the high the Red Sea or between the Red Sea and adjacent regions. The number of endemic species that it harbours make it an ideal

123 Coral Reefs (2013) 32:737–748 741 system to study gene flow and speciation. In particular, thermal stress or ultraviolet radiation on zooxanthellae and there is an opportunity to investigate how biophysical coral regulatory pathways (Winters et al. 2006, 2009; gradients (from south to north), distance and historical Zeevi-Ben-Yosef and Benayahu 2008; Kvitt et al. 2011), fluctuations of the sea level have shaped the genetic two discussed oceanographic aspects of thermal stress diversity and species distribution in this region (DiBattista (Veal et al. 2010; Davis et al. 2011), and one used coral et al. 2013). There are likely to be some hidden genetic cores to quantify the deleterious effects of ocean warming breaks throughout its length, and in general, connectivity on coral growth rates (Cantin et al. 2010). Other relevant studies will definitely help to reveal some insight into the papers investigated the possibility of Red Sea corals acting processes and mechanisms at the origin of such high as refugia of biodiversity in scenarios of severe global endemism rates that are present today in the Red Sea. environmental stress (Riegl and Piller 2003), included the Connectivity studies are essential if effective spatial Red Sea in a comparative review of global trajectories of management is to be achieved in a region (McCook et al. the state of coral reefs (Pandolfi et al. 2003) or discussed 2009). Given a lack of such management in most of the the relationship between coral disease and bleaching Red Sea, future connectivity research should focus on (Rosenberg and Ben-Haim 2002; Danovaro et al. 2008). elucidating general patterns of larval dispersal, gene flow It is readily apparent that research relating to coral among populations and linkages among the many types of bleaching in the Red Sea has lagged well behind that of the coastal habitats found in the Red Sea (e.g. snapper GBR and Caribbean. Studies quantifying the extent and migration corridors from coastal wetlands to reefs further severity of bleaching events in the Red Sea, the effect of offshore; McMahon et al. 2012). Such data would be bleaching on reef biodiversity and community structure, essential to inform the design of marine protected areas, patterns of recovery, and predictions for the future are should there be a move to develop them within this region. either lacking or published in grey literature that is difficult to access (e.g. PERSGA/GEF 2010). Although we do not Coral bleaching know what tolerance thresholds Red Sea corals have, functional genomics work could make large advances in Red Sea reefs thrive in the warmest and most saline waters this field by addressing questions regarding the mecha- of any extensive coral reef system in the world, due largely nisms for tolerance to the local environmental conditions. to the limited freshwater inflow and restricted water This is an obvious area in which Red Sea research could exchange with the Indian Ocean (Sheppard et al. 1992). For make significant contributions to global coral reef ecology. this reason, Red Sea reefs have been assumed to be highly In contrast, research from both the Caribbean and the thermo-tolerant and thus predicted to be among the last GBR covers a more diverse variety of bleaching-related reefs to bleach as seawater temperatures increase (Grims- topics with a considerable number of publications con- ditch and Salm 2006). However, recent observations of cerned with each topic (Fig. 1c). Such topics include the severe bleaching on inshore reefs in the central Red Sea following: detailed, quantitative reports on bleaching (Furby et al. 2013) indicate that this prediction may be events (e.g. Baird and Marshall 1998; Jimenez 2001); inaccurate and raises questions as to whether bleaching investigations of potential causes of bleaching events events have been occurring in the Red Sea but were (Gleason and Wellington 1993; Shinn et al. 2000); the overlooked or not reported due to limited observer effort. effect of bleaching on different groups of organisms and Our WoK search yielded only 12 relevant coral various aspects of reef community structure and ecological bleaching papers from the Red Sea compared with 339 processes (e.g. Meesters and Bak 1993; Mumby 1999; from the GBR and 167 from the Caribbean (Table 1, ESM Baird and Marshall 2002); the tolerance thresholds of dif- 3). More than half (7 out of 12) of the relevant Red Sea ferent coral taxa and clades of zooxanthellae to thermal publications were conducted within the Gulf of Aqaba, stress (e.g. Marshall and Baird 2000; Lasker 2003); showing both a general lack of information on the topic of recovery patterns, long-term considerations and future coral bleaching in the Red Sea as a whole and a very strong predictions (e.g. Hughes 2003; Baker et al. 2008); and regional bias in the little information that is available. many other subtopics. Of the 12 ISI-listed publications found from the Red Sea, only one mentioned the actual occurrence of bleaching Coral reproductive biology events in the Red Sea, recalling three major events all of which occurred in the 1990s in the central and southern Over the past three decades, considerable research has Red Sea (Turak et al. 2007). The description of these focused on the reproductive strategies (timing and mode of bleaching events was based largely on qualitative obser- reproduction) of scleractinian corals (Baird et al. 2009). vations and contained very limited quantitative data. Of the Early research focused on brooding taxa that employ remaining publications, four examined the effects of internal fertilization and release fully developed planulae 123 742 Coral Reefs (2013) 32:737–748

(Harrison 2011), and led to the general misconception that Herbivory brooding was the main form of larval development of reef corals. However, the discovery of mass spawning events in Herbivory is widely accepted as a key process structuring the mid-1980s on the Great Barrier Reef (Harrison et al. benthic communities on coral reefs and consequently is one 1984; Babcock et al. 1986) challenged this view and of the most thoroughly studied aspects of coral reef ecol- established that hermaphroditic broadcast spawning is the ogy (e.g. Hay 1984; Hughes 1994). The importance of predominate reproductive strategy among scleractinian herbivory has long been recognized in coral reef ecosys- corals. This reproductive strategy has since been the focus tems worldwide (e.g. GBR: Stephenson and Searles 1960; of most studies in coral reproductive biology. Currently, Caribbean: Randall 1965; Red Sea: Vine 1974); however, information on the sexual reproductive biology of scle- the majority of these early studies were largely descriptive. ractinian corals is available for over 444 species worldwide The increasing prevalence of anthropogenic and climate- (Harrison 2011). induced disturbance and subsequent collapse of several Our WoK search returned 11 articles on coral repro- reef systems from coral- to macroalgal dominance high- ductive biology in the Red Sea (nine of which were from lighted the critical importance of herbivory and brought the Gulf of Aqaba). In comparison, 197 articles were found this research area to the fore (Hughes 1994; McClanahan from the Great Barrier Reef and 108 from the Caribbean, et al. 2001; Bellwood et al. 2004). This renewed emphasis which together represent over 50 % of the total number of lead to marked increases in herbivory-focused research on articles in this field (Table 1, ESM 4). Of the two articles the GBR and Caribbean reefs, but there was no corre- from the Red Sea basin that were not from the Gulf of sponding increase in such research in the Red Sea Aqaba, one study described the synchrony of reproductive (Fig. 1h). As a consequence, our current understanding of condition of Acropora species (Hanafy et al. 2010) and the herbivory in the Red Sea is limited and is based largely on other one found limited larval dispersal of a brooding coral inference from other regions. after studying the gene flow between 2 locations in the Gulf Despite the considerable attention herbivory has attrac- of Aqaba and a third location in the main Red Sea basin ted, our WoK search revealed that only 34 of the 1,066 located over 500 km away (Maier et al. 2005). herbivory papers (3.2 %) were conducted within the Red Interestingly, among the studies from the Gulf of Aqaba, Sea (Table 1, ESM 5). In marked contrast, herbivory has one study revealed that spawning in this particular region of attracted much greater attention on the Great Barrier Reef the Red Sea is asynchronous, with different species releasing (322 papers) and in the Caribbean (255 papers), accounting gametes at different times throughout the year (Shlesinger for 30 and 24 % of herbivory studies, respectively. et al. 1998). This seems to be in stark contrast with obser- Within Indo-Pacific and Caribbean reef systems, a vations from the Great Barrier Reef where synchronous coral wealth of studies have documented large variation in the spawning seems to be a general trend (e.g. Shlesinger et al. abundance and community structure of herbivores across a 1998). However, two independent studies conducted in the range of spatial scales (e.g. latitude: Floeter et al. 2005; main Red Sea basin have reported both the presence of shelf position: Hoey and Bellwood 2008; habitat: Hay synchronous spawning of Acropora species in the central 1981) and related this to variation in algal communities Red Sea (Bouwmeester et al. 2011) and a strong synchrony in across similar scales (e.g. Wismer et al. 2009). These cor- the gametogenesis of Acropora corals (Hanafy et al. 2010), relative relationships have been supported by experiments suggesting that reproduction patterns differ between the Gulf that have demonstrated the exclusion of herbivores leads to of Aqaba and the central Red Sea. The generality of these a proliferation of algal biomass and a shift toward larger patterns, however, needs to be investigated further. erect macroalgae (e.g. Hughes et al. 2007) and direct esti- To date, our knowledge of patterns of reproduction of mates of herbivory using macrophyte assays (e.g. McCook corals in the main Red Sea (i.e. excluding the Gulf of 1996). While this body of work clearly demonstrates the Aqaba) is limited to a few species of Acropora (Hanafy importance of herbivory per se, another suite of studies have et al. 2010), and our understanding of recruitment patterns focused on understanding the functional importance of is limited to a single study in the Gulf of Aqaba (Glassom individual taxa, the level of redundancy within functional et al. 2004). There is a clear need for future studies to groups, and the influence of habitat characteristics on for- document the reproductive modes of Red Sea corals, aging activities of herbivores (e.g. Bellwood et al. 2006; quantify the timing of spawning events and how they vary Mumby 2006; Burkepile and Hay 2006; Hoey and Bell- among species and locations, identify environmental cues wood 2011; Rasher et al. 2013). Collectively, these studies that are associated with spawning events, and quantify identified a functional dichotomy between those species pelagic larval durations and larval competency to better that have the capacity to prevent (i.e. grazers) or potentially understand the reproductive connectivity within the Red reverse (i.e. browsers) shifts to macroalgal dominance on Sea. coral reefs. 123 Coral Reefs (2013) 32:737–748 743

In marked contrast, studies examining herbivory within of this research on Egyptian reefs meant that, of the eight the Red Sea have been largely restricted to descriptions of topics reviewed, this topic had the lowest proportion of among-habitat variation in herbivore assemblages (e.g. studies conducted within the Gulf of Aqaba. Bouchon-Navaro and Harmelin-Vivien 1981; Brokovich Although very few quantitative data are available, it et al. 2010). Apart from a couple of recent papers that have appears that many of the Red Sea reefs experience heavy quantified variation in the role of parrotfishes on Red Sea fishing pressure (Jin et al. 2012) with little or no effective reefs (Alwany et al. 2009; Afeworki et al. 2011), there is a management. With the exception of perhaps Egypt, the lack of quantitative information on the role of herbivores in need for management strategies and implementation is this region. The Red Sea contains many endemics for readily apparent. There have been several, and relatively which basic dietary data are completely lacking. Future ambitious, plans for MPAs in some regions of the Red Sea. research should firstly focus on identifying the trophic and For example, in 1988, the IUCN/UNEP (1988) proposed 40 functional affinities of individual taxa and quantifying MPAs along the Saudi Arabian coast of the Red Sea. Most large-scale variation (i.e. cross-shelf and latitudinal) in of these recommendations, however, have not been herbivore communities and functions. The limited rainfall implemented and are generally hidden in the grey litera- and freshwater input into the Red Sea makes it an ideal ture. As several endemic species are targeted in the fish- system to examine the effects of herbivory in the absence eries (e.g. DesRosiers 2011; Jin et al. 2012), and the of land-based eutrophication. For example, cross-shelf region’s largest oil producer (Saudi Aramco) has recently variation in benthic communities on the GBR, in particular begun extraction operations in the Saudi Arabian Red Sea, the high macroalgal cover on inshore reefs, has been sug- there are compelling conservation motivations for creating gested to be related to increased nutrient input from ter- a plan to sustainably harvest the natural resources of the restrial activities (Wismer et al. 2009). However, Saudi Red Sea. Arabian reefs in the central Red Sea display similar cross- shelf variation in benthic communities (Hoey and Beru- Non-coral invertebrates (represented by sponges) men, pers obs). Finally, comparisons of rates and agents of herbivory between the Red Sea and species-rich regions Non-coral invertebrates provide the greatest biodiversity to such as the GBR (the Red Sea has approximately one-third coral reef environments, but are probably the least studied of the herbivore species richness of the GBR; Bellwood (Bouchet et al. 2002; Appeltans et al. 2012). The sheer and Wainwright 2002) present a unique opportunity to number of invertebrate species inhabiting coral reefs pre- examine how herbivore diversity influences ecosystem vented us from reviewing them collectively. Among all process on coral reefs. non-coral invertebrates associated with reef systems, sponges are an important structural and functional com- Marine protected areas ponent of coral reefs (Diaz and Tzler 2001) and are among the most commonly studied non-coral invertebrates in reef Marine protected areas (MPAs) are increasingly popular systems. As such, they were selected as an ‘‘indicator’’ tools for the management of coral reef systems (Mora et al. group to observe the trends and current state of research of 2006; Almany et al. 2009). Very few MPAs, unfortunately, invertebrates in the Red Sea relative to the GBR and the have the luxury of having the full spectrum of scientific Caribbean. data available to ensure optimal reserve design (McCook Our WoK search revealed 23 studies on sponges from the et al. 2009). However, as MPAs become more prevalent, Red Sea (18 from the Gulf of Aqaba), compared to 88 from scientific attention is likewise growing in attempts model the Great Barrier Reef and 170 from the Caribbean local characteristics to optimize reserve designs and to (Table 1, ESM 7). These 23 studies from the Red Sea do not assess the effectiveness of existing MPAs. seem to show any unifying theme. In fact, they span a broad As with other topics, our WoK search revealed only 17 spectrum of topics—from taxonomy to reproduction and Red Sea studies, compared to 220 from the Great Barrier anti-predatory defences (e.g. Burns et al. 2003; Gugel et al. Reef and 152 from the Caribbean (Table 1, ESM 6). 2011). Further, several of the papers were not specific to Within the Red Sea, Egypt has the largest number and sponges, instead focused more generally on benthic fauna greatest area of reefs within MPAs. Accordingly, the with only peripheral attention to sponges. Several articles majority of the results from our WoK search were con- detail the taxonomic identification of new species of spon- ducted in the Egyptian Red Sea (ESM 6). Of the studies ges; these primarily originate from the Gulf of Aqaba (ESM conducted in Egyptian waters, most investigated the 7). Of the five studies conducted outside the Gulf of Aqaba, effectiveness of established MPAs (Galal et al. 2002; two examined the community structure of sponges using Ashworth et al. 2006; Kochzius 2007; Marshall et al. 2010; genetic techniques (Wo¨rheide 2006;Wo¨rheide et al. 2008), Hannak et al. 2011; Samy et al. 2011). The concentration while the remaining three papers examined symbiotic 123 744 Coral Reefs (2013) 32:737–748 relationships between sponges and other organisms (Wil- Perhaps the most unusual studies are those regarding the kinson and Fay 1979; Ilan et al. 1999; Magnino et al. 1999). giant bacterium Epulopiscium fishelsoni, an intestinal tract As genetic tools and methods continue to be developed, symbiont of some Red Sea surgeonfishes (e.g. Clements sponges together with other non-coral invertebrates could and Bullivant 1991; Angert et al. 1993). This enigmatic prove to be ideal model organisms to address numerous bacterium was originally discovered in the Red Sea, and evolutionary and ecological questions. the paper that identified them as bacteria (as opposed to Non-coral invertebrates are understudied worldwide, eukaryotic protists) is amongst one of the most highly cited and the lack of research on this group appears to be papers from the region (Angert et al. 1993). accentuated within the Red Sea. Future research should Globally, coral reef-associated microbial research is prioritize describing the diversity and distribution of non- focusing on relationships between microbes and coral dis- coral invertebrates on Red Sea reefs. At the same time, new ease and more recently coral bleaching (e.g. Aronson and species may be screened for biomedical applications. Precht 2001; Rosenberg et al. 2007). Notably, many studies Sponges and the vast array of bioactive compounds they now recognize the importance of identifying microbiota harbour have long been of commercial interest for their associated with corals, sponges and other benthic organisms potential pharmacological value. Given that intensive tax- under normal, ‘‘healthy’’ conditions (e.g. Rohwer et al. onomic efforts have lapsed since the earliest European 2002; Taylor et al. 2007). In fact, most microbes are likely naturalists visited, the Red Sea likely holds many exciting non-infectious and many play key roles in nutrient recycling new discoveries. Given the number of endemic corals and of the reef ecosystem (Moriarty et al. 1985). Understanding fishes within the Red Sea, it is likely that a particularly long the role of these microbial communities will no doubt add to list of new species of non-coral invertebrates awaits our understanding of many interesting ecological phenom- description by modern taxonomy (with increasingly com- ena. Microbial research on reefs in the Red Sea should bined morphological and molecular approaches) in the Red continue to describe distribution, community composition Sea. and function of reef-associated microbes to establish baseline data. The environmental conditions of the Red Sea, Reef-associated bacteria however, provide a unique opportunity to examine the role of microbes under elevated seawater temperatures and Recent advances in meta-‘‘omic’’ approaches have opened salinity. This may provide critical insights into the potential up entirely new fields of work related to coral- and reef- role of microbes on reefs worldwide under continued cli- associated microbial communities (Rohwer and Youle mate change. 2010). Although the ecological role of most of these associated microbial communities remains enigmatic in most cases, this is an emerging field of work that will General discussion inform research in coral disease, coral health and resilience, and likely even thermal tolerances. Our WoK search Despite the extent of coral reefs in the Red Sea, the bio- was restricted to ‘‘bacteria’’ as ‘‘microbiology’’ proved to diversity and endemism these reefs contain, and the sem- be far too broad. We acknowledge that while our search inal work of early natural historians in the Red Sea, this may not capture all aspects of microbiology, it does pro- region is sorely understudied. On average across the topics vide a useful comparison of research efforts in this field we investigated, Australia’s Great Barrier Reef and the among the three geographic regions. Caribbean have 6 and 8 times the number of published Our WoK search revealed 41 Red Sea studies, compared studies, respectively, than the Red Sea. The lack of avail- to 342 from the Great Barrier Reef and 166 from the able information represents, in many cases, a significant Caribbean (Table 1, ESM 8). Of all the topics we investi- hurdle for conservation and management in the region. We gated, reef-associated bacteria are among the most poorly hope that, by highlighting gaps and areas of most urgent studied outside of the Gulf of Aqaba. More than 80 % of need for research attention, we can spur work in these the papers our search returned were from the Gulf of Aqaba areas. (33 out of 41 papers), with the vast majority of the Red Sea We do not intend to diminish the value of the work that very poorly covered. Of the 41 papers from the Red Sea has been done in the Red Sea region. Work from the Red (ESM 8), six focused on the relationships between coral Sea has undeniably been influential in coral reef ecology, disease and certain bacteria (e.g. Arotsker et al. 2009; as shown by the large number of results that arise due to the Zvuloni et al. 2009). Several publications focused on KeyWords Plus feature in WoK searches when using the bacteria isolated from sponges (e.g. Kelman et al. 2001; term ‘‘Red Sea’’. Specifically, this is an indication that Bergman et al. 2011) and from coral mucus (e.g. Kooper- many studies from other parts of the world heavily cite man et al. 2007; Shnit-Orland and Kushmaro 2009). work based in the Red Sea. Nonetheless, the current state 123 Coral Reefs (2013) 32:737–748 745 of knowledge of coral reef ecology in the Red Sea is the Red Sea for each of our topics but rather to highlight generally far behind that of comparable regions around the the knowledge gap. world. Given modern predictions of climate change, the The Red Sea has the potential to once again have an Red Sea may serve as an important natural laboratory to important influence on coral reef ecology worldwide. As help understand the near future for coral reefs elsewhere in reefs face increasing pressure from global climate change, the world. The increasing attention to research in the Red many authors have suggested that corals in the Red Sea Sea can thus have broader impact than simple acquisition may provide insight into mechanisms of adaptation or of region-specific knowledge. tolerance to elevated temperatures. Given the relatively While conducting this review, we have come across young age of the Red Sea and high endemism of its fauna, numerous relevant references that would traditionally be the system may help us to understand speciation processes considered ‘‘grey’’ literature. Many Red Sea studies are and other processes creating and maintaining biodiversity published in conference proceedings or journals of local on coral reefs. Given general consensus that we are expe- institutions that may be difficult for researchers outside of riencing a global biodiversity crisis, coupled with limited the region to discover. Of course, grey literature also exists resources for conservation, understanding these processes for the other regions to which we compare the Red Sea is more important than ever. The Red Sea represents a results, but the relative importance of this body of knowl- largely untapped scientific resource with great potential, edge is not clear. A substantial language barrier may exist and it is our hope that future work will strategically address for many researchers working in Red Sea countries, and in many of these important gaps in our knowledge. many cases, these researchers may not submit their work to WoK-listed journals. In the Caribbean and Great Barrier Acknowledgments Discussions with GR Almany, PL Munday and Reef, however, it could be argued that a higher proportion B Gladfelter improved the manuscript. We thank H Sweatman and an anonymous reviewer for feedback on previous versions of the man- of the authors are native English speakers, and many of uscript. We extend particular thanks to M Hassan for the discussion of these authors target WoK-listed journals. In these two regional relevance of the grey literature and her analysis of authorship regions, some hold a perception that grey literature repre- nationalities for Red Sea publications. sents work of insufficient quality to have been accepted in WoK-listed journals. On the other hand, many Red Sea researchers may view grey literature as a primary option for publication. A related observation is that the first authors for References the vast majority of the WoK papers from the Red Sea are not natives of a Red Sea country. In other words, it appears Afeworki Y, Bruggemann JH, Videler JJ (2011) Limited flexibility in that most of these papers are written by scientists visiting resource use in a coral reef grazer forgaing on seasonally changing algal communities. Coral Reefs 30:109–122 the region. This may be indicative of a fundamental dif- Almany G, Connolly S, Heath D, Hogan J, Jones G, McCook L, Mills ference in publication strategy between the Red Sea and the M, Pressey R, Williamson D (2009) Connectivity, biodiversity other two regions we assessed. Simply increasing the conservation and the design of marine reserve networks for coral amount of research funding (or number of institutions) in reefs. Coral Reefs 28:339–351 Alwany MA, Thaler E, Stachowitsch M (2009) Parrotfish bioerosion the region would not likely address this fundamental chal- on Egyptian Red Sea reefs. J Exp Mar Biol Ecol 371:170–176 lenge. Therefore, in parallel to increased capacity and Angert ER, Clements KD, Pace NR (1993) The largest bacterium. financial support, efforts for regional improvement should Nature 362:239–241 also focus on creative solutions to support local scientists to Appeltans W, Ahyong ST, Anderson G, Angel MV, Artois T, Bailly N, Bamber R, Barber A, Bartsch I, Berta A, Blazewicz- make a transition from targeting regional ‘‘grey’’ outlets to Paszkowycz M, Bock P, Boxshall G, Boyko CB, Brandaõ SN, internationally recognized journals. Bray RA, Bruce NL, Cairns SD, Chan T-Y, Cheng L, Collins While we acknowledge that our WoK searches do not AG, Cribb T, Curini-Galletti M, Dahdouh-Guebas F, Davie PJF, fully capture all possible known information about these Dawson MN, De Clerck O, Decock W, De Grave S, de Voogd NJ, Domning DP, Emig CC, Erseús C, Eschmeyer W, Fauchald areas of research, we maintain that these numbers reflect K, Fautin DG, Feist SW, Fransen CHJM, Furuya H, Garcia- what is readily accessible to most colleagues who might Alvarez O, Gerken S, Gibson D, Gittenberger A, Gofas S, take an interest in regional work. Although some databases Go´mez-Daglio L, Gordon DP, Guiry MD, Hernandez F, do exist for ‘‘grey’’ Red Sea work (e.g. PERSGA/GEF Hoeksema BW, Hopcroft RR, Jaume D, Kirk P, Koedam N, Koenemann S, Kolb JrB, Kristensen RM, Kroh A, Lambert G, 2010), in many cases, some prior knowledge of these Lazarus DB, Lemaitre R, Longshaw M, Lowry J, Macpherson E, databases must exist in order to locate the articles. There- Madin LP, Mah C, Mapstone G, McLaughlin PA, Mees J, fore, a standardized WoK search likely reveals the actual Meland K, Messing CG, Mills CE, Molodtsova TN, Mooi R, relative state of internationally accessible knowledge Neuhaus B, Ng PKL, Nielsen C, Norenburg J, Opresko DM, Osawa M, Paulay G, Perrin W, Pilger JF, Poore GCB, Pugh P, available for each region. Again, our purpose in this review Read GB, Reimer JD, Rius M, Rocha RM, Saiz-Salinas JI, is not to definitively assess all available information from Scarabino V, Schierwater B, Schmidt-Rhaesa A, Schnabel KE, 123 746 Coral Reefs (2013) 32:737–748

Schotte M, Schuchert P, Schwabe E, Segers H, Self-Sullivan C, Burns E, Ilan M (2003) Comparison of anti-predatory defenses of Red Shenkar N, Siegel V, Sterrer W, Stohr S, Swalla B, Tasker ML, Sea and Caribbean sponges. II. Physical defense. Mar Ecol Prog Thuesen EV, Timm T, Todaro MA, Turon X, Tyler S, Uetz P, Ser 252:115–123 van der Land J, Vanhoorne B, van Ofwegen LP, van Soest Burns E, Ifrach I, Carmeli S, Pawlik JR, Ilan M (2003) Comparison of RWM, Vanaverbeke J, Walker-Smith G, Walter TC, Warren A, anti-predatory defenses of Red Sea and Caribbean sponges. Williams GC, Wilson SP, Costello MJ (2012) The magnitude of I. Chemical defense. Mar Ecol Prog Ser 252:105–114 global marine species diversity. Curr Biol 22:2189-2202 Cantin NE, Cohen AL, Karnauskas KB, Tarrant AM, McCorkle DC Aronson RB, Precht WF (2001) White-band disease and the changing (2010) Ocean warming slows coral growth in the Central Red face of Caribbean coral reefs. Hydrobiologia 460:25–38 Sea. Science 329:322–325 Arotsker L, Siboni N, Ben-Dov E, Kramarsky-Winter E, Loya Y, Clarke C, Lea J, Ormond R (2011) Reef-use and residency patterns of Kushmaro A (2009) Vibrio sp as a potentially important member a baited population of silky sharks, Carcharhinus falciformis,in of the Black Band Disease (BBD) consortium in Favia sp corals. the Red Sea. Mar Freshw Res 62:668–675 FEMS Microbiol Ecol 70:515–524 Clements KD, Bullivant S (1991) An unusual symbiont from the gut Ashworth JS, Bruce OE, El Hellw M (2006) Fish assemblages of Red Of surgeonfishes may be the largest known prokaryote. J Bac- Sea backreef biotopes. Aquat Conserv: Mar Freshw Ecosyst teriol 173:5359–5362 16:593–609 Collette BB, Carpenter KE, Polidoro BA, Juan-Jorda MJ, Boustany A, Babcock RC, Bull GD, Harrison PL, Heyward AJ, Oliver JK, Wallace Die DJ, Elfes C, Fox W, Graves J, Harrison LR, McManus R, CC, Willis BL (1986) Synchronous spawnings of 105 scleractinian Minte-Vera CV, Nelson R, Restrepo V, Schratwieser J, Sun CL, coral species on the Great Barrier Reef. Mar Biol 90:379–394 Amorim A, Brick Peres M, Canales C, Cardenas G, Chang SK, Baird AH, Marshall PA (1998) Mass bleaching of corals on the Great Chiang WC, de Oliveira Leite N, Jr., Harwell H, Lessa R, Fredou Barrier Reef. Coral Reefs 17:376 FL, Oxenford HA, Serra R, Shao KT, Sumaila R, Wang SP, Baird AH, Marshall PA (2002) Mortality, growth and reproduction in Watson R, Yanez E (2011) High value and long life-Double scleractinian corals following bleaching on the Great Barrier jeopardy for tunas and billfishes. Science 333:291–292 Reef. Mar Ecol Prog Ser 237:133–141 Danovaro R, Bongiorni L, Corinaldesi C, Giovannelli D, Damiani E, Baird AH, Guest JR, Willis BL (2009) Systematic and biogeographical Astolfi P, Greci L, Pusceddu A (2008) Sunscreens cause coral patterns in the reproductive biology of scleractinian corals. Annual bleaching by promoting viral infections. Environ Health Perspect Review of Ecology, Evolution and Systematics 40:551–571 116:441–447 Baker AC, Glynn PW, Riegl B (2008) Climate change and coral reef Davis KA, Lentz SJ, Pineda J, Farrar JT, Starczak VR, Churchill JH bleaching: An ecological assessment of long-term impacts, recovery (2011) Observations of the thermal environment on Red Sea trends and future outlook. Estuar Coast Shelf Sci 80:435–471 platform reefs: a heat budget analysis. Coral Reefs 30:25–36 Bellwood DR, Wainwright PC (2002) The history and biogeography DesRosiers N (2011) Growth and maturation of Plectropomus spp. in of fishes on coral reefs. In: Sale PF (ed) Coral reef fishes: the Saudi Arabian Red Sea. MSc thesis, King Abdullah University Dynamics and diversity in a complex ecosystem. Academic of Science and Technology, p 69 Press, San Diego, pp 5–32 Diaz MC, Tzler K (2001) Sponges: An essential component of Bellwood DR, Hughes TP, Folke C, Nystrom M (2004) Confronting Caribbean coral reefs. Bull Mar Sci 69:535–546 the coral reef crisis. Nature 429:827–833 DiBattista JD, Berumen ML, Gaither MR, Rocha LA, Eble JA, Choat Bellwood DR, Hughes TP, Hoey AS (2006) Sleeping functional JH, Craig MT, Skillings DJ, Bowen BW (2013) After continents group drives coral reef recovery. Curr Biol 16:2434–2439 divide: comparative phylogeography of reef fishes from the Red Belmaker J, Shashar N, Ziv Y (2005) Effects of small-scale isolation Sea and Indian Ocean. J Biogeogr 40:1170–1181 and predation on fish diversity on experimental reefs. Mar Ecol Edwards FJ (1987) Climate and oceanography. In: Edwards AJ, Head Prog Ser 289:273–283 S (eds) Key environments: Red Sea. Pergamon Press, Oxford, Bergman O, Haber M, Mayzel B, Anderson MA, Shpigel M, Hill RT, pp 45–68 Ilan M (2011) Marine-based cultivation of Diacarnus sponges Fishelson L, Baranes A (1997) Ontogenesis and cytomorphology of the and the bacterial community composition of wild and maricul- nasal olfactory organs in the Oman Shark, Iago omanensis tured sponges and their larvae. Mar Biotechnol 13:1169–1182 (Triakidae), in the Gulf of Aqaba, Red Sea. Anat Rec 209:409–421 Bond ME, Babcock EA, Pikitch EK, Abercrombie DL, Lamb NF, Fisher PR, Newton SF, Tatwany HMA, Goldspink CR (2001) Chapman DD (2012) Reef sharks exhibit site-fidelity and higher Variation in the diet of Ospreys Pandion haliaetus, Farasan relative abundance in marine reserves on the Mesoamerican Islands, southern Red Sea-preliminary observations. Vogelwelt Barrier Reef. PloS ONE 7:e32983 122:205–218 Bouchet P, Lozouet P, Maestrati P, Heros V (2002) Assessing the Floeter SR, Behrens MD, Ferreira CEL, Paddack MJ, Horn MH magnitude of species richness in tropical marine environments: (2005) Geographical gradients of marine herbivorous fishes: exceptionally high numbers of molluscs at a New Caledonia site. Patterns and processes. Mar Biol 147:1435–1447 Biol J Linn Soc 75:421–436 Furby KA, Bouwmeester J, Berumen ML (2013) Susceptibility of Bouchon-Navaro Y, Harmelin-Vivien ML (1981) Quantitative distri- central Red Sea corals during a major bleaching event. Coral bution of herbivorous reef fishes in the Gulf of Aqaba (Red Sea). Reefs 22:505–513 Mar Biol 63:79–86 Galal N, Ormond R, Hassan O (2002) Effect of a network of no-take Bouwmeester J, Khalil MT, De La Torre P, Berumen ML (2011) reserves in increasing catch per unit effort and stocks of Synchronous spawning of Acropora in the Red Sea. Coral Reefs exploited reef fish at Nabq, South Sinai. Egypt. Mar Freshw Res 30:1011 53:199 Brokovich E, Ayalon I, Einbinder S, Segev N, Shaked Y, Genin A, Glassom D, Zakai D, Chadwick-Furman NE (2004) Coral recruit- Kark S, Kiflawi M (2010) Grazing pressure on coral reefs ment: a spatio-temporal analysis along the coastline of Eilat, decreases across a wide depth gradient in the Gulf of Aqaba, Red northern Red Sea. Mar Biol 144:641–651 Sea. Mar Ecol Prog Ser 399:69–80 Gleason DF, Wellington GM (1993) Ultraviolet radiation and coral Burkepile DE, Hay ME (2006) Herbivore vs. nutrient control of bleaching. Nature 365:836–838 marine primary producers: context-dependent effects. Ecology Grimsditch GD, Salm RV (2006) Coral reef resilience and resistence 87:3128–3139 to bleaching. IUCN, Gland, Switzerland, p 52

123 Coral Reefs (2013) 32:737–748 747

Gugel J, Wagler M, Bruemmer F (2011) Porifera, one new species natural and aquarium environments. FEMS Microbiol Lett Suberea purpureaflava n. sp. (Demospongiae, Verongida, Ap- 276:106–113 lysinellidae) from northern Red Sea coral reefs, with short Kvitt H, Rosenfeld H, Zandbank K, Tchernov D (2011) Regulation of descriptions of Red Sea Verongida and known Suberea species. apoptotic pathways by Stylophora pistillata (Anthozoa, Pocillo- Zootaxa 2994:60–68 poridae) to survive thermal stress and bleaching. PLoS ONE Hanafy MH, Aamer MA, Habib M, Rouphael AB, Baird AH (2010) 6:e28665 Synchronous reproduction of corals in the Red Sea. Coral Reefs Lasker HR (2003) Zooxanthella densities within a Caribbean 29:119–124 octocoral during bleaching and non-bleaching years. Coral Reefs Hannak JS, Kompatscher S, Stachowitsch M, Herler J (2011) 22:23–26 Snorkelling and trampling in shallow-water fringing reefs: Risk Lowe WH, Allendorf FW (2010) What can genetics tell us about assessment and proposed management strategy. J Environ Man- population connectivity? Mol Ecol 19:3038–3051 age 92:2723–2733 Lutcavage M, Kraus S, Hoggard W (1997) Aerial survey of giant Harrison PL (2011) Sexual reproduction of scleractinian corals. In: bluefin tuna, Thunnus thynnus, in the great Bahama Bank, Straits Dubinsky Z, Stambler N (eds) Coral reefs: An ecosystem in of Florida, 1995. Fish Bull 95:300–310 transition. Springer, Netherlands, pp 59–85 Magnino G, Sara` A, Lancioni T, Gaino E (1999) Endobionts of the Harrison PL, Babcock RC, Bull GD, Oliver JK, Wallace CC, Willis BL coral reef sponge Theonella swinhoei (Porifera, Demospongiae). (1984) Mass spawning in tropical reef corals. Science 223:1186–1189 Invertebr Biol 118:213–220 Hay ME (1981) Spatial patterns of grazing intensity on a Caribbean Maier E, Tollrian R, Rinkevich B, Nu¨rnberger B (2005) Isolation by barrier reef: herbivory and algal distribution. Aquat Bot 11:97–109 distance in the scleractinian coral Seriatopora hystrix from the Hay ME (1984) Patterns of fish and urchin grazing on Caribbean coral Red Sea. Mar Biol 147:1109–1120 reefs: are previous results typical? Ecology 65:446–454 Marshall PA, Baird AH (2000) Bleaching of corals on the Great Head SM (1987) Introduction. In: Edwards AJ, Head SM (eds) Key Barrier Reef: differential susceptibilities among taxa. Coral environments: Red Sea. Pergamon Press, Oxford, UK, pp 1–21 Reefs 19:155–163 Hilborn R (2007) Moving to sustainability by learning from Marshall NA, Marshall PA, Abdulla A, Rouphael T (2010) The links successful fisheries. AMBIO 36:0044–7447 between resource dependency and attitude of commercial fishers Hoey AS, Bellwood DR (2008) Cross-shelf variation in the role of to coral reef conservation in the Red Sea. Ambio 39:305–313 parrotfishes on the Great Barrier Reef. Coral Reefs 27:37–47 McClanahan TR, Muthiga NA, Mangi S (2001) Coral and algal changes Hoey AS, Bellwood DR (2011) Suppression of herbivory by after the 1998 coral bleaching: interaction with reef management macroalgal density: a critical feedback on coral reefs? Ecol Lett and herbivores on Kenyan reefs. Coral Reefs 19:380–391 14:267–273 McCook LJ (1996) Effects of herbivores and water quality on Hughes TP (1994) Catastrophes, phase shifts, and large-scale Sargassum distribution on the central Great Barrier Reef: cross- degradation of a Caribbean coral reef. Science 265:1547–1551 shelf transplants. Mar Ecol Prog Ser 139:179–192 Hughes L (2003) Climate change and Australia: Trends, projections McCook LJ, Almany GR, Berumen ML, Day JC, Green AL, Jones and impacts. Austral Ecol 28:423–443 GP, Leis JM, Planes S, Russ GR, Sale PF, Thorrold SR (2009) Hughes TP, Rodrigues MJ, Bellwood DR, Ceccarelli D, Hoegh- Management under uncertainty: guide-lines for incorporating Guldberg O, McCook L, Moltschaniwskyj M, Pratchett MS, connectivity into the protection of coral reefs. Coral Reefs Steneck RS, Willis B (2007) Phase shifts, herbivory, and the 28:353–366 resilience of coral reefs to climate change. Curr Biol 17:360–365 McCook LJ, Ayling T, Cappo M, Choat JH, Evans RD, De Freitas Ilan M, Loya Y, Kolbasov GA, Brickner I (1999) Sponge-inhabiting DM, Heupel M, Hughes TP, Jones GP, Mapstone B (2010) barnacles on Red Sea coral reefs. Mar Biol 133:709–716 Adaptive management of the Great Barrier Reef: A globally IUCN/UNEP (1988) Coral reefs of the world: Volume 2: Indian significant demonstration of the benefits of networks of marine Ocean, Red Sea and Gulf. UNEP/IUCN, Gland, Switzerland and reserves. Proc Natl Acad Sci USA 107:18278–18285 Cambridge UK/UNEP, Nairobi, Kenya McMahon KW, Berumen ML, Thorrold SR (2012) Linking habitat Ivanov VA, Lipshitz A (2006) Description of a new diphyllidean mosaics and connectivity in a coral reef seascape. Proc Natl parasite of triakid sharks from the deep Red Sea. J Parasitol Acad Sci USA 109:15372–15376 94(841):846 Meesters EH, Bak RPM (1993) Effects of coral bleaching on tissue Jimenez C (2001) Bleaching and mortality of reef organisms during a regeneration potential and colony survival. Mar Ecol Prog Ser warming event in 1995 on the Caribbean coast of Costa Rica. 96:189–198 Rev Biol Trop 49:233–238 Mervis J (2009) The big gamble in the Saudi desert. Science Jin D, Kite-Powell HL, Hoagland P, Solow AR (2012) A bioeconomic 326:354–357 analysis of traditional fisheries in the Red Sea off the coast of the Mora C, Andrefouet S, Costello MJ, Kranenburg C, Rollo A, Veron J, Kingdom of Saudi Arabia. Mar Res Econ 27:137–148 Gaston KJ, Myers RA (2006) Coral reefs and the global network Jones GP, Russ GR, Sale PF, Steneck RS (2009) Theme section on of marine protected areas. Science 312:1750–1751 ‘‘Larval connectivity, resilience and the future of coral reefs’’. Moriarty DJW, Pollard PC, Hunt WG (1985) Temporal and spatial Coral Reefs 28:303–305 variation in bacterial production in the water column over a coral Karpestam B, Gustafsson J, Shashar N, Katzir G, Kru¨ger RHH (2007) reef. Mar Biol 85:285–292 Multifocal lenses in coral reef fishes. J Exp Biol 210:2923–2931 Mumby PJ (1999) Bleaching and hurricane disturbances to popula- Kelman D, Kashman Y, Rosenberg E, Ilan M, Ifrach I, Loya Y (2001) tions of coral recruits in Belize. Mar Ecol Prog Ser 190:27–35 Antimicrobial activity of the reef sponge Amphimedon viridis Mumby PJ (2006) The impact of exploiting grazers (Scaridae) on the from the Red Sea: evidence for selective toxicity. Aquat Microb dynamics of Caribbean coral reefs. Ecol Appl 16:747–769 Ecol 24:9–16 Myers R, Baum J, Shepherd T, Powers S, Peterson C (2007) Kochzius M (2007) Community structure of coral reef fishes in El Cascading effects of the loss of apex predatory sharks from a quadim bay (El quseir, Egyptian red sea coast). Zool Middle East coastal ocean. Science 315:1846–1850 42:89–98 Ormond RFG, Edwards AJ (1987) Red Sea fishes. In: Edwards AJ, Kooperman N, Ben-Dov E, Kramarsky-Winter E, Barak Z, Kushmaro Head SM (eds) Red Sea: Key environments. IUCN/Pergamon A (2007) Coral mucus-associated bacterial communities from Press, Oxford, pp 251–287

123 748 Coral Reefs (2013) 32:737–748

Pandolfi JM, Bradbury RH, Sala E, Hughes TP, Bjorndal KA, Cooke habitats in the Gulf of Aqaba (Red-Sea). Environ Biol Fish RG, McArdle D, McClenachan L, Newman MJH, Paredes G, 31:0378–1909 Warner RR, Jackson JBC (2003) Global trajectories of the long- Stehli FG, Wells JW (1971) Diversity and age patterns in hermatypic term decline of coral reef ecosystems. Science 301:955–958 corals. Syst Biol 20:115–126 PERSGA/GEF (2010) The Status of coral reefs in the Red Sea and Stephenson W, Searles RB (1960) Experimental studies on the Gulf of Aden: 2009 PERSGA Technical Series Number 16. ecology of intertidal environments at Heron Island. Aust J Mar PERSGA, Jeddah Freshw Res 11:241–267 Pikitch EK, Chapman DD, Babcock EA, Shivji MS (2005) Habitat Taylor MW, Radax R, Steger D, Wagner M (2007) Sponge-associated use and demographic population structure of elasmobranchs at a microorganisms: Evolution, ecology, and biotechnological Caribbean atoll (Glover’s Reef, Belize). Mar Ecol Prog Ser potential. Microbiol Mol Biol Rev 71:295–347 302:187–197 Turak E, Brodie J, DeVantier L (2007) Reef-building corals and coral Randall JE (1965) Grazing effects on sea greases by herbivorous reef communities of the Yemen Red Sea. Fauna of Arabia 23:1–40 fishes in the West Indies. Ecology 46:255–260 Veal CJ, Carmi M, Dishon G, Sharon Y, Michael K, Tchernov D, Randall JE, Levy MF (1976) A near-fatal shark attack by a mako in Hoegh-Guldberg O, Fine M (2010) Shallow-water wave lensing the northern Red Sea. Isr J Zool 25:61–70 in coral reefs: a physical and biological case study. J Exp Biol Rasher DB, Hoey AS, Hay ME (2013) Consumer diversity interacts 213:4304–4312 with prey defences to drive ecosystem function. Ecology. doi: Vine PJ (1974) Effects of algal grazing and aggressive behavior of the 10.1890/12-0389.1 fishes Pomacentrus lividus and Acanthurus sohal on coral-reef Riegl B, Piller WE (2003) Possible refugia for reefs in times of ecology. Mar Biol 24:131–136 environmental stress. Int J Earth Sci 92:520–531 Walker TI (1998) Can shark resources be harvested sustainably? A Robbins WD, Hisano M, Connolly SR, Choat JH (2006) Ongoing question revisited with a review of shark fisheries. Mar Freshw collapse of coral-reef shark populations. Curr Biol 16:2314–2319 Res 49:1448–6059 Rohwer F, Youle M (2010) Coral reefs in the microbial seas. Plaid Wilkinson CR, Fay P (1979) Nitrogen fixation in coral reef sponges Press with symbiotic cyanobacteria. Nature 279:527–529 Rohwer F, Seguritan V, Azam F, Knowlton N (2002) Diversity and Winters G, Loya Y, Beer S (2006) In situ measured seasonal distribution of coral-associated bacteria. Mar Ecol-Prog Ser variations in F-v/F-m of two common Red Sea corals. Coral 243:1–10 Reefs 25:593–598 Rosenberg E, Ben-Haim Y (2002) Microbial diseases of corals and Winters G, Beer S, Ben Zvi B, Brickner I, Loya Y (2009) Spatial and global warming. Environ Microbiol 4:318–326 temporal photoacclimation of Stylophora pistillata: zooxanthella Rosenberg E, Koren O, Reshef L, Efrony R, Zilber-Rosenberg I size, pigmentation, location and clade. Mar Ecol Prog Ser (2007) The role of microorganisms in coral health, disease and 384:107–119 evolution. Nature Reviews Microbiology 5:355–362 Wismer S, Hoey AS, Bellwood DR (2009) Cross-shelf benthic Sale PF, Cowen RK, Danilowicz BS, Jones GP, Kritzer JP, Lindeman community structure on the Great Barrier Reef: relationships KC, Planes S, Polunin NVC, Russ GR, Sadovy YJ, Steneck RS between macroalgal cover and herbivore biomass. Mar Ecol (2005) Critical science gaps impede use of no-take fishery Prog Ser 376:45–54 reserves. Trend Ecol Evol 20:74–80 Wo¨rheide G (2006) Low variation in partial cytochrome oxidase Samy M, Sanchez Lizaso JL, Forcada A (2011) Status of marine subunit I (COI) mitochondrial sequences in the coralline protected areas in Egypt. Anim Biodivers Conserv 34:165–177 demosponge Astrosclera willeyana across the Indo-Pacific. Sheppard C, Price A, Roberts C (1992) Marine ecology of the Arabian Mar Biol 148:907–912 region. Academic Press, London, p 359 Wo¨rheide G, Epp LS, Macis L (2008) Deep genetic divergences Shinn EA, Smith GW, Prospero JM, Betzer P, Hayes ML, Garrison V, among Indo-Pacific populations of the coral reef sponge Leucetta Barber RT (2000) African dust and the demise of Caribbean chagosensis (Leucettidae): founder effects, vicariance, or both? coral reefs. Geophys Res Lett 27:3029–3032 BMC Evol Biol 8:24 Shlesinger Y, Goulet T, Loya Y (1998) Reproductive patterns of Zeevi-Ben-Yosef D, Benayahu Y (2008) Synergistic effects of UVR scleractinian corals in the northern Red Sea. Mar Biol 132:691–701 and temperature on the survival of azooxanthellate and zooxan- Shnit-Orland M, Kushmaro A (2009) Coral mucus-associated bacte- thellate early developmental stages of soft corals. Bull Mar Sci ria: a possible first line of defense. FEMS Microbiol Ecol 83:401–414 67:371–380 Zvuloni A, Artzy-Randrup Y, Stone L, Kramarsky-Winter E, Barkan Shpigel M, Fishelson L (1991) Experimental removal of piscivorous R, Loya Y (2009) Spatio-temporal transmission patterns of groupers of the genus Cephalopholis (Serranidae) from coral Black-Band disease in a coral community. PloS ONE 4:e4993

123 Coral Reefs (2013) 32:505–513 DOI 10.1007/s00338-012-0998-5

REPORT

Susceptibility of central Red Sea corals during a major bleaching event

K. A. Furby • J. Bouwmeester • M. L. Berumen

Received: 18 April 2012 / Accepted: 11 December 2012 / Published online: 4 January 2013 Ó Springer-Verlag Berlin Heidelberg 2013

Abstract A major coral bleaching event occurred in the were depth of the reef and distance of the reef from shore. central Red Sea near Thuwal, Saudi Arabia, in the summer Shallow reefs and inshore reefs had a higher prevalence of of 2010, when the region experienced up to 10–11 degree bleaching. This bleaching event shows that Red Sea reefs heating weeks. We documented the susceptibility of vari- are subject to the same increasing pressures that reefs face ous coral taxa to bleaching at eight reefs during the peak of worldwide. This study provides a quantitative, genus-level this thermal stress. Oculinids and agaricids were most assessment of the vulnerability of various coral groups susceptible to bleaching, with up to 100 and 80 % of col- from within the Red Sea to bleaching and estimates sub- onies of these families, respectively, bleaching at some sequent mortality. As such, it can provide valuable insights reefs. In contrast, some families, such as mussids, pocil- into the future for reef communities in the Red Sea. loporids, and pectinids showed low levels of bleaching (\20 % on average). We resurveyed the reefs 7 months Keywords Coral bleaching Mortality Red Sea later to estimate subsequent mortality. Mortality was highly Saudi Arabia Sea surface temperature Susceptibility variable among taxa, with some taxa showing evidence of full recovery and some (e.g., acroporids) apparently suf- fering nearly complete mortality. The unequal mortality Introduction among families resulted in signiﬁcant change in community composition following the bleaching. Signiﬁcant fac- Anomalous temperatures of even 1 °C higher than average tors in the likelihood of coral bleaching during this event summer sea surface temperature (SST) can trigger coral bleaching (Glynn 1991; Goreau and Hayes 1994; Hoegh- Guldberg 1999). Alarmingly, the frequency of bleaching Communicated by Biology Editor Dr. Hugh Sweatman events and mortalities has increased dramatically since Electronic supplementary material The online version of this 1979 (Glynn 1993; Hoegh-Guldberg 1999). However, coral article (doi:10.1007/s00338-012-0998-5) contains supplementary species vary in their susceptibility to bleaching and sub- material, which is available to authorized users. sequent rates of mortality (e.g., Marshall and Baird 2000; McClanahan et al. 2004), and understanding the patterns of K. A. Furby (&) Marine Biology Research Division, Scripps Institution of bleaching susceptibility is critical for accurate predictions Oceanography, 9500 Gilman Dr, La Jolla, CA 92093-0202, USA of future reef communities (Marshall and Baird 2000). e-mail: [email protected] Differences in the bleaching susceptibility among coral species have been attributed to phylogeny, colony size, J. Bouwmeester M. L. Berumen Red Sea Research Center, King Abdullah University of Science depth, or habitat type (Glynn 1990; Marshall and Baird and Technology, Thuwal, Jeddah 23955-6900, Kingdom of 2000; Baird and Marshall 2002; McClanahan 2004). In Saudi Arabia previous studies, corals were most susceptible if they were in shallow water (Brown and Suharsono 1990; Hoeksema M. L. Berumen Biology Department, Woods Hole Oceanographic Institution, 1991; Glynn 1996) or in a genus of fast-growing corals 266 Woods Hole Rd, Woods Hole, MA 02543-1050, USA (e.g., Acropora and Pocillopora) (Marshall and Baird 123 506 Coral Reefs (2013) 32:505–513

2000). Alternatively, corals from stagnant inner lagoons town approximately 90 km north of Jeddah on the Saudi and shallow reef areas may be less susceptible to thermal Arabian Red Sea coast, were on Bleaching Alert Level 2 stress compared with conspecifics that live in cooler hab- during this time period, the highest risk level watch for itats offshore or at greater depths (Hoeksema 1991; Oliver NOAA’s DHW charts. and Palumbi 2009). We captured potential differences in bleaching suscep- Some authors have suggested that a history of bleaching tibility among taxa, depth, and location. We aimed to test and thermal stress may drive coral populations to develop two hypotheses: (1) the combined and independent effects resistance to future bleaching through selection for ther- of depth and distance from shore on corals changed mally tolerant individuals (Brown 1997; reviewed in van bleaching prevalence and (2) coral taxa differed in their Oppen et al. 2009; Eakin et al. 2010). Previous surveys of susceptibility to bleaching, leading to subsequent changes mass bleaching in other locations have found up to 83 % of in the compositions of the coral communities. corals were affected (Harriott 1985; Glynn 1990; Marshall and Baird 2000; DeVantier and Pilcher 2000; Glynn et al. 2001), and mortality in various families reached 88 % Materials and methods (Baird and Marshall 2002). Because bleaching causes different rates of mortality in different coral species and its Sites effects on subsequent reproductive success also vary, bleaching events can change coral community composition Eight reefs near Thuwal, Kingdom of Saudi Arabia, in the and diversity (Glynn 1990; Hughes and Connell 1999; Red Sea, were surveyed in September 2010 (Fig. 1). Three Baird and Marshall 2002; McClanahan et al. 2007). offshore reefs (Abu Madafi, Al-Mashpah, Shi’b Nazar) were The Red Sea is a deep, evaporative basin with high located on the edge of the continental shelf, *15–20 km salinity ([38 ppt) and high water temperatures. It receives from shore. Three midshelf reefs (Palace Reef, Qita al-Kirsh, little freshwater input (30 mm year-1) and is isolated from Umm al-Kiethl) were located *8–9 km from shore. Two larger bodies of water (Berman et al. 2003). Sea surface inshore reefs (Tahla and Fsar), *2–4 km from shore, were temperature along the Saudi Arabian coast of the Red Sea also surveyed. All surveys were conducted on the exposed typically shows seasonal variations of *10 °C and sum- side of the reefs (facing into the prevailing wind and wave mer temperatures often exceed 32 °C (e.g., Davis et al. direction). Six of the surveyed reefs (all except for Al- 2011). SST anomalies in the Red Sea increased abruptly in Mashpah and Umm al-Kiethl) were revisited in April 2011, 1994, followed by a 0.7 °C increase in the last decade after the bleaching event, to record any changes in community (Raitsos et al. 2011). The Red Sea is also a biodiversity hot composition. spot (e.g., Hughes et al. 2002; Roberts et al. 2002). How- ever, much of the published work from the Red Sea comes from a very limited area (primarily the Gulf of Aqaba/Eilat, Benthic cover for example, Spaet et al. 2012). Comparatively little information is available from the central and southern Red Benthic cover was recorded along 10 m transects at three Sea, although many climate-change-oriented studies con- depths: 5, 10, and 15 m, using the line-intercept method. sider this to be a key region for understanding future Three replicate transects were completed at each depth, for environmental conditions on reefs worldwide (e.g., Merg- a total of nine transects at each reef. Transects were hap- ner 1984; Cantin et al. 2010). hazardly placed at each site. We quantified benthic cover In this study, we quantified the effects of a bleaching using 43 categories and recorded each colony or individual event in 2010 in the northern hemisphere summer along as bleached or non-bleached, using 10 m line-intercept the Saudi Arabian Red Sea coast. According to NOAA transects following Berumen et al. (2005). Such was the Coral Reef Watch (2000), the degree heating weeks intensity of bleaching that almost all corals that we (DHW) for the Thuwal reefs for the weeks before Sep- recorded as ‘‘bleached’’ were fully bleached; only a tember 27, 2010, were 9.0–10.0 near the coast and lessened handful of colonies showed evidence of ‘‘patchy bleach- to 8.0 on reefs farther offshore. Degree heating weeks are a ing’’ or ‘‘partial bleaching’’ (sensu Spalding 2009). The metric for describing the accumulated stress placed on low number of colonies in these latter categories led us to corals by high temperature over long durations. One DHW combine them into one category, ‘‘bleached’’, for our equals 1 week of SSTs that exceed the maximum expected analyses. Coral groups were identified to genus (following summer temperature by C1 °C (Toscano et al. 1999). A Hoeksema (1989) for Fungiidae; and Veron (2000) for all sharp increase occurred in the Thuwal region in late other families). As detailed below, in select analyses and August 2010, peaked in late September (11.0 DHW), and figures, we pooled data at the family level due to the decreased in October. Thus, the reefs near Thuwal, a small absence of several species and genera on some transects. 123 Coral Reefs (2013) 32:505–513 507

Fig. 1 Map of reefs surveyed in central Red Sea, near Thuwal, Saudi Arabia, during a bleaching event in 2010. Surveyed reefs are labeled and indicated with an ‘‘9’’

Calculations transects were averaged for each reef. One-way ANOVA was used to test for differences in richness between years. On each transect, the intercept length (cm) of healthy cover All calculations were performed using R (R Development (lh) and bleached cover (lb) in each taxonomic category Core Team 2011) or Systat (Ver 11.2). was used to calculate the percent bleaching (Pb) using the formula: Results l P ¼ 100 b b l þ l b h General patterns of bleaching We assessed the spatial patterns of percentage of bleached corals using a two-way nested ANOVA with distance off- In 2010, surveys of eight reefs found that the inshore reefs shore (3 levels: offshore, midshelf, and inshore) and depth (3 experienced higher levels of bleaching (74 % of hard corals levels: 5, 10, and 15 m) as fixed factors, and site (reef) as a were bleached) than offshore reefs (14 % of hard corals) random factor nested in distance categories. The response (Table 1). The reefs with the most dramatic bleaching were variable was the percentage of bleached coral. A PerMA- inshore reefs at 5 m, some of which experienced [95 % NOVA was used to test for differences in community hard coral bleaching (Fig. 2; Electronic Supplementary structure at a family level between 2010 and 2011 using the Material, ESM Table S1). Depth had a significant impact on vegan package in R (Oksanen et al. 2011). Due to limited bleaching prevalence (ANOVA, df = 2, F = 4.573, replication, depth and distance factors were pooled. This p = 0.012). Tukey’s HSD was used to test for differences analysis used year as a fixed factor and community com- among factor levels. The prevalence of bleaching at 5 m position as the response variable. To visualize the impact of depth was significantly higher than coral bleaching at 10 and bleaching on coral community composition between years, 15 m; however, 10 and 15 m were not significantly different we compared the six reefs surveyed in both 2010 and 2011 from each other (Fig. 2). by calculating the Bray-Curtis dissimilarity among all reefs Distance from shore, as a categorical variable, also had a plotted using non-metric multidimensional scaling (NMDS) significant impact on bleaching prevalence (ANOVA, df = 2, using the vegan package in R (Oksanen et al. 2011). F = 6.269, p = 0.003). The bleaching prevalence was sig- Richness was calculated for each transect, defined as nificantly higher among corals on inshore reefs compared the number of genera recorded on each transect, and all with those on midshelf and offshore reefs. However, the

123 508 Coral Reefs (2013) 32:505–513

Table 1 Mean proportion (% ± SE) of cover of coral families that were bleached in 2010 on reefs in the Central Saudi Arabian Red Sea, by site (all depths combined), with sites arranged from inshore to offshore Inshore 1 Inshore 2 Midshelf 1 Midshelf 3 Midshelf 3 Offshore 1 Offshore 2 Offshore 3

Acroporidae 78.5 ± 9.7 61.3 ± 15.8 33.8 ± 13.3 43.0 ± 13.8 9.8 ± 5.6 10.7 ± 5.9 14.3 ± 14.3 9.9 ± 9.9 Agariciidae 88.0 ± 8.4 75.9 ± 10.2 43.1 ± 20.2 61.9 ± 14.8 18.3 ± 10.6 22.0 ± 12.1 75.0 ± 25.0 31.6 ± 16.1 Dendrophylliidae 50.0 ± 0.5 Faviidae 78.8 ± 0.1 66.5 ± 0.2 37.5 ± 0.1 47.3 ± 0.1 19.6 ± 0.1 12.1 ± 0.1 36.9 ± 0.1 23.3 ± 0.1 Fungiidae 100 ± 0.0 65.7 ± 0.2 33.3 ± 0.3 100.0 ± 0.0 22.2 ± 0.2 45.0 ± 0.3 0.0 ± 0.0 0.0 ± 0.0 Merulinidae 100 ± 0.0 0.0 ± 0.0 100.0 33.3 ± 0.3 50.0 ± 0.5 0.0 ± 0.0 0.0 Mussidae 33.3 ± 0.3 0.0 0.0 ± 0.0 100.0 0.0 ± 0.0 0.0 ± 0.0 Oculinidae 54.6 ± 0.2 67.5 ± 0.2 62.0 ± 0.2 53.1 ± 0.2 33.3 ± 0.3 100.0 50 ± 0.5 50.0 ± 0.5 Pectiniidae 78.6 ± 0.2 33.3 ± 0.3 100.0 100.0 0.0 ± 0.0 Pocilloporidae 75.0 ± 0.1 60.9 ± 0.2 8.4 ± 0.0 2.9 ± 0.0 2.2 ± 0.0 0.7 ± 0.0 1.5 ± 0.0 0.5 ± 0.0 Poritidae 85.5 ± 0.1 35.5 ± 0.1 22.3 ± 0.1 39.7 ± 0.1 32.3 ± 0.2 17.2 ± 0.1 12 ± 0.1 16.5 ± 0.1 Siderastreidae 85.0 ± 0.2 66.7 ± 0.3 50.0 ± 0.5 100 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 Average (reef) 82.4 51.4 32.3 58.9 33.7 28.6 17.2 13.2 If there is no ±, then N = 1 for that site (i.e., the family was only recorded on one transect at the site). If the cell is blank, then N = 0 for that site (i.e., the family was never recorded at that site) bleaching prevalence did not differ significantly between Poritidae (Fig. 3a). After the bleaching event, these corals midshelf and offshore reefs (Tukey’s HSD). Depth and dis- remained the most common except on inshore reefs where tance had no significant interaction (ANOVA, df = 5, communities were most affected by the bleaching event. F = 0.672, p = 0.646). The NMDS plot illustrates qualitative changes in the community composition of the six reefs surveyed after the Susceptibility among taxa coral bleaching event (stress = 1.75 %, Fig. 4). The most dramatic changes are apparent in the two inshore reefs On average, the three most abundant families were Pocillo- (Fig. 4). To further examine the impact of bleaching on the poridae, Acroporidae, and Faviidae (Fig. 3a). Coral families community composition, we plotted the percent cover of showed variation in their susceptibility to bleaching (Table 1; the six most abundant families at 5 m (where bleaching Fig. 3b). Bleaching affected 10–60 % of the most common was most prevalent) for these six reefs (Fig. 5). Some scleractinian corals (16 families) in the central Red Sea families of corals on inshore reefs (e.g., Acroporidae) were (Fig. 3b). The families Oculinidae, Agariciidae, and Fungii- extremely susceptible to bleaching, and these were virtu- dae were most sensitive to bleaching (an average of 45–60 %, ally absent in 2011 (Fig. 5e, f). up to 80–100 % at some sites) (Fig. 3b; Table 1). The level In 2010, inshore reefs had the highest levels of richness of bleaching experienced by the three most common families among all sites. Following the bleaching event, richness (Pocilloporidae, Acroporidae, and Faviidae) averaged across differed significantly across depths and reefs (ANOVA, all sites, and depths was 19, 33, and 40 % (respectively). The df = 1, F = 8.693, p = 0.0146). Overall richness declined mussids, pocilloporids, and pectinids bleached less (\20 % most at shallow depths and the inshore reefs. Richness on average). Some families appeared to be highly resistant to decreased on five of the six that were resurveyed reefs in bleaching, such as the Astrocoeniidae and Euphyllidae which 2011. The average richness (per 10 m transect) was 15 were rarely observed to bleach during our surveys, although genera at the inshore reefs in 2010, decreasing to 10 in they were not common families. 2011 (Fig. 6).

Community composition change Discussion There was a signiﬁcant difference in community composition at a family level between 2010 and 2011 (PerMA- This study provides an important, quantitative, genus-level NOVA, F = 3.0156, p = 0.013). Average coral cover in analysis of susceptibility and mortality associated with 2010 was similar across all reefs surveyed; however, there bleaching for Red Sea corals. Previous studies in the Red was a general decrease in coral cover when reefs were Sea have recorded bleaching due to freshwater runoff resurveyed in 2011. Dominant coral families in 2010 at all (Antonius 1988) and the 1997/8 ENSO bleaching event reefs were Acroporidae, Faviidae, Pocilloporidae, and (DeVantier et al. 2000). The central Red Sea experienced

123 Coral Reefs (2013) 32:505–513 509

100 100 100 90 Offshore 1 90 Midshelf 1 90 80 Inshore 1 80 80 70 70 70 60 60 60 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 5m 10m 15m 5m 10m 15mm 5m 10m 15m 100 100 100 90 Offshore 2 90 Midshelf 2 90 Inshore 2 80 80 80 70 70 70 60 60 60 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 5m 10m 15m 5m 10m 15mm Percentage of hard coral cover bleached 5m 10m 15m 100 100 90 Offshore 3 90 Midshelf 3 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 5m 10m 15m 5m 10m 15m Depth

Fig. 2 Percent of bleached scleractinian corals at each of three depths Graphs are arranged according to relative position on the continental (5, 10, 15 m) at 8 reefs in the Saudi Arabian Red Sea. Bar indicates shelf: offshore, midshelf, and inshore average percent bleaching of hard corals per 10 m transect (±SE).

0.9–1.3 °C higher SST than average (*14 DHW, NOAA found bleaching severely affecting several agaricids (in Coral Reef Watch 2000) during the 1998 global bleaching concurrence with the present study), although they observed event (Goreau et al. 2000). However, prior to this study, bleaching far less frequently in oculinids (in contrast to the little information about bleaching susceptibility and asso- present study). Also consistent between the two studies was ciated mortality among taxa has been available for the the pattern of increased bleaching in the shallow areas corals of the central Red Sea. (B5 m) compared to deeper transects. Anecdotal evidence suggests that the 2010 Thuwal Shallow areas of reefs were disproportionately more bleaching was restricted to the central Red Sea region, as affected by thermal stress than deep reef areas in our study, other Saudi Arabian Red Sea reefs that we surveyed during consistent with studies from other locations (e.g., Harriott this time did not bleach signiﬁcantly. We are aware of only 1985; Brown and Suharsono 1990; Hoeksema 1991). The one previous report of thermal bleaching in the central Red thermal stress affecting the shallow-water corals may have Sea: DeVantier et al. (2005), which provides rapid ecolog- been compounded by UV radiation (Glynn 1996; Torres- ical assessment categorical data (following DeVantier et al. Perez and Armstrong 2012), exacerbating the bleaching 1998) of the 1998 global bleaching event. Interestingly, they and/or mortality due to bleaching. Deep areas of reefs may

123 510 Coral Reefs (2013) 32:505–513

9 offshore is due to the proximity of deep and relatively cooler 8 a 7 water to these reefs, while inshore reefs on the continental 6 shelf may not have had much exchange with deeper water. 5 4 The link between upwelling events and coral health (e.g., 3 Riegl and Piller 2003) may be a factor in the success of the 2 Coral cover (%) 1 offshore corals. Offshore reefs, adjacent to much deeper 0 waters, are likely to receive cooler water input, which can 70 mediate the effect of high SSTs (sensu Davis et al. 2011). An 60 b additional potential compounding factor was the decreased 50 average surface wind speeds in the Thuwal region during the 40 summer of 2010 (JT Farrar, pers comm), resulting in reduced 30 20 circulation on the inshore shelf. Goreau and Hayes (1994) Bleaching (%) 10 found that bleaching ‘‘hot spots’’ were often coupled with 0 low wind and high temperature, as we found in the 2010 Red Sea event, and low cloudiness and low rainfall, which are

FaviidaePoritidae Mussidae both typical characteristics of this Red Sea location. OculinidaeFungiidae Pectiniidae Acroporidae Agariciidae Merulinidae Euphyllidae Pocilloporidae Siderastreidae CaryophylliidaeAstrocoeniidae Euphorbiaceae In the same manner as DeVantier et al. (2005), we Dendrophylliidae Coral families (in order of % cover) informally surveyed several dive operators on the Saudi Red Sea coast, but also found no reports of bleaching Fig. 3 Susceptibility of 16 coral families to bleaching in the Saudi during 2010 outside of the central region (near Thuwal). Arabian Red Sea as observed at three depths at each of eight reefs in The relative rarity of bleaching events in the Red Sea September 2010. a Mean percent cover (total bleached and non- bleached) on 10 m line-intercept transects. b Mean proportion of hinders a full understanding of the capacity of these corals cover within each family that was bleached to adapt to changing conditions, but this study clearly shows that these reefs are susceptible to anomalous tem- therefore act as important refugia for healthy corals (Riegl perature events. Investigating the susceptibility of partic- and Piller 2003). ular taxa is an important first step toward an understanding Corals on the offshore reefs appeared to suffer less from what kind of community composition future reefs may the thermal stress than the inshore reefs we surveyed. Our have (McClanahan et al. 2009). follow-up surveys in 2011 showed that the coral bleaching The bleaching event we observed had a significant impact event resulted in a loss of richness on inshore reefs, pre- the coral communities, causing declines in diversity and sumably due to mortality of corals. Multiple factors could total coral cover on many reefs. Specifically, some reefs saw affect the inshore-to-offshore gradient in bleaching preva- nearly a complete local extinction of certain taxa. Inshore lence. Decreased circulation and/or terrestrial input may reefs lost virtually all of the taxa that were dominant in our compound the effect of thermal stress on the corals (Jokiel initial surveys, and several of these taxa (e.g., acroporids and and Brown 2004). Although we lack formal oceanographic pocilloporids) are important as food and shelter for many data, we suspect that the low prevalence of bleaching reef fishes (Pratchett et al. 2008, 2009). As expected, not all

Fig. 4 Qualitative changes in Community changes following 2010 bleaching coral communities at six reefs Offshore2 2010 (named and color coded) in the central Saudi Arabian Red Sea Offshore1 2011 following a bleaching event in 50 Midshelf2 2010 2010. Plot illustrates a non- metric multidimensional scaling Offshore2 2011 using Bray-Curtis dissimilarities Midshelf2 2011

(NMDS stress = 1.75 %) 0 Midshelf1 2010

NMDS2 Midshelf1 2011 Inshore2 2011 Offshore1 2010 Inshore1 2011 -50

Inshore1 2010 Inshore2 2010

-60 -40 -20 0 20 40 60 NMDS1

123 Coral Reefs (2013) 32:505–513 511

a Offshore 1 2010 b Offshore 2 2010 30 30 2011 2011 25 25

20 20

15 15

10 10

5 5

0 0

c Midshelf 1 2010 d Midshelf 2 2010 30 30 2011 2011 25 25

20 20

15 15

10 10

5 5

0 0 Coral cover (Average % per 10m) Coral cover (Average e Inshore 1 2010 f Inshore 2 2010 30 30 2011 2011 25 25

20 20

15 15

10 10

5 5

0 0 Faviidae Faviidae Poritidae Poritidae Fungiidae Fungiidae Agariciidae Agariciidae Acroporidae Acroporidae Pocilliporidae Pocilliporidae

Fig. 5 Percent cover of six most common coral families at 5 m depth length) that was bleached (white bars) compared to unbleached (black for six reefs surveyed in 2010 and 2011. Data for 2010 are color bars). Data for 2011 (gray bars) indicate percent of benthic cover in coded to indicate the percent of benthic cover (i.e., line-intercept 2011 (no bleaching was observed in 2011) taxa were equally affected by the bleaching event. However, bleaching at this site. During the 2010 surveys, we were as in previous studies, reefs severely affected by the coral unable to install permanent transects due to logistical and bleaching event experienced a decline in richness (e.g., permitting complications. We did resurvey the same areas Glynn 1990). Previous work suggests that faster-growing of the reefs in 2011, aided by GPS coordinates of our initial coral genera (e.g., Acropora, Pocillopora, Stylophora,etc.) surveys. It is likely that the decrease in richness reﬂects an will be most extremely affected by bleaching, whereas unfortunate positioning of our transects rather than a result slower-growing genera (e.g., Favia, Porites, Fungia, etc.) of bleaching related mortality. The genera absent in some will be less severely affected (sensu Harriott 1985;Hoek- of the 2011 transects include, for example, Acropora, sema 1991;Glynn1993; Marshall and Baird 2000). These Favia, and Echinopora. These genera were recorded in patterns hold in our study. The difference in communities 2010 and showed among the lowest levels of bleaching at from 2010 to 2011 was largely driven by declines in this reef (Table 1). Complete summary data for each acroporids and pocilloporids on the shallow and inshore family, depth, and site are presented in the ESM Table S1. reefs. It is interesting to note that Pocillopora was resistant It can be expected that the observed changes in the coral to bleaching on all but the inshore reefs, where it was community composition of these central Red Sea reefs heavily bleached at all survey depths. will have ﬂow-on effects in the non-coral community, We do note an anomalous decrease in richness at particularly in the abundances of organisms that are Offshore Reef 1 (Shi’b Nazar), although there was little dependent on corals for food or shelter (e.g., Berumen and

123 512 Coral Reefs (2013) 32:505–513

20 References 2010 18 a 2011 16 Antonius A (1988) Distribution and dynamics of coral diseases in the th 14 Eastern Red Sea. Proc 6 Int Coral Reef Symp 2:293–298 Baird AH, Marshall PA (2002) Mortality, growth and reproduction in 12 scleractinian corals following bleaching on the Great Barrier 10 Reef. Mar Ecol Prog Ser 237:133–141

Richness 8 Berman T, Paldor N, Brenner S (2003) Annual SST cycle in the 6 Eastern Mediterranean, Red Sea and Gulf of Elat. Geophys Res 4 Lett 30:1261 Berumen ML, Pratchett MS (2006) Recovery without resilience: 2 Persistent disturbance and long-term shifts in the structure of fish 0 and coral communities at Tiahura Reef, Moorea. Coral Reefs Inshore 1 Inshore 2 Midshelf 1 Midshelf 2 Offshore 1 Offshore 2 25:647–653 Reefs Berumen ML, Pratchett MS, McCormick MI (2005) Within-reef 16 differences in diet and body condition of coral-feeding butter- b 2010 flyfishes (Chaetodontidae). Mar Ecol Prog Ser 287:217–227 14 2011 Brown BE (1997) Coral bleaching: causes and consequences. Coral Reefs 16:129–138 12 Brown BE, Suharsono (1990) Damage and recovery of coral reefs affected by El Nino related to seawater warming in the Thousand 10 Islands, Indonesia. Coral Reefs 8:163–170 8 Cantin NE, Cohen AL, Karnauskas KB, Tarrant AM, McCorkle DC (2010) Ocean warming slows coral growth in the central Red Richness 6 Sea. Science 329:322–325 Davis KA, Lentz SJ, Pineda J, Farrar JT, Starczak VR, Churchill JH 4 (2011) Observations of the thermal environment on Red Sea platform reefs: a heat budget analysis. Coral Reefs 30:25–36 2 Devantier L, Pilcher N (2000) The status of coral reefs in Saudi Arabia. In: Wilkinson CR (ed) Status of coral reefs of the world 0 5m 10m 15m (2000). Australian Institute of Marine Science, Townsville Depth Devantier LM, De’ath G, Done TJ, Turak E (1998) Ecological assessment of a complex natural system: a study from the Great Fig. 6 Average genus richness of hard corals at six reefs in the Barrier Reef. Ecol Appl 8:480–496 central Saudi Arabian Red Sea surveyed in 2010 and 2011. a Richness Devantier LM, Turak E, Al-Shaikh K, De’ath G (2000) Coral averaged by reef (5, 10, and 15 m depths combined), presented in communities of the central-northern Saudi Arabian Red Sea. order of increasing distance from shore (from left to right). b Richness Fauna of Arabia 18:23–66 averaged among all sites by depth. Error bars are ±SE Devantier LM, Turak E, Al-Shaikh K (2005) Coral bleaching in the central-northern Saudi Arabian Red Sea August-September 1998. In: Abuzinada A (ed) Proceedings of the International Workshop on the Extent and Impact of Coral Reef Bleaching in the Arabian Pratchett 2006; Pratchett et al. 2011; Stella et al. 2011). Region. Kingdom of Saudi Arabia Publications, National Com- Organisms that depend on corals in a facultative manner mission for Wildlife Conservation and Development may have a delayed but noticeable response to changes in Eakin CM, Morgan JA, Heron SF, Smith TB, Liu G, Alvarez- Filip L, Baca B, Bartels E, Bastidas C, Bouchon C, Brandt M, Bruckner the coral communities (McClanahan et al. 2009; Pratchett AW, Bunkley-Williams L, Cameron A, Causey B, Chiappone M, et al. 2009). These effects remain poorly understood in the Christensen TRL, Crabbe MJC, Day O, de la Guardia E, Dıáz- Red Sea ecosystem and warrant further investigation. Pulido G, DiResta D, Gil-Agudelo D, Gilliam DS, Ginsburg RN, This study shows that Red Sea reefs are not immune to Gore S, Guzmań HM, Hendee JC, Hernańdez-Delgado EA, Husain E, Jeffrey CFG, Jones RJ, Jordań-Dahlgren E, Kaufman increasing global pressures and the rising SSTs may be LS, Kline DI, Kramer PA, Lang JC, Lirman D, Mallela J, approaching the thermal tolerance limits of the corals. Manfrino C, Marećhal JP, Marks K, Mihaly J, Miller WJ, Estimates of the susceptibility of various taxa in the Red Mueller EM, Orozco Toro CA, Oxenford HA, Ponce-Taylor D, Sea to bleaching allows insight into the potential for further Quinn N, Ritchie KB, Rodrı´guez S, Rodrı´guez Ramı´rez A, Romano S, Samhouri JF, Sańchez JA, Schmahl GP, Shank BV, community changes in the tropical reefs of the region. Skirving WJ, Steiner SCC, Villamizar E, Walsh SM, Walter C, Weil E, Williams EH, Woody Roberson K, Yusuf Y (2010) Acknowledgments This work was supported by the Red Sea Caribbean corals in crisis: Record thermal stress, bleaching, and Research Center at the King Abdullah University of Science and mortality in 2005. PLoS ONE 5(11):e13969. doi:10.1371/jour Technology, with logistical support from the Coastal and Marine nal.pone.0013969 Resources Core Lab. The authors thank A Baird, M Pratchett, and B Glynn PW (1990) Coral mortality and disturbances to coral reefs in Riegl for helpful discussions of the study; A Baird and J Veron for the tropical Eastern Pacific. In: Glynn PW (ed) Global ecological assistance with taxonomy; G Williams, N Price, M Johnson, and consequences of the 1982–83 El Nino-Southern Oscillation S Sandin for guidance in statistical analysis; and two anonymous event on corals and coral reefs of the eastern Pacific. Elsevier reviewers for comments on the manuscript. Oceanogr Ser 52, Amsterdam, pp 55–126

123 Coral Reefs (2013) 32:505–513 513

Glynn PW (1991) Coral reef bleaching in the 1980s and possible Heating Weeks Product, 26 Jul. to 28 Oct. 2010. Silver Spring, connections with global warming. TREE 6:175–179 Maryland, USA: NOAA Coral Reef Watch. Data set accessed Glynn PW (1993) Coral reef bleaching: ecological perspectives. Coral 2011-07-30 at http://coralreefwatch.noaa.gov/satellite/hdf/index.html Reefs 12:1–17 Oksanen J, Blanchet FG, Kindt R, Legendre P, O’Hara RB, Simpson Glynn PW (1996) Coral reef bleaching: facts, hypotheses and GL, Solymos P, Stevens MHH, Wagner H (2011) Vegan: implications. Global Change Biol 2:495–509 community ecology package. R package version 1.17-11. http:// Glynn PW, Mate-T JL, Baker AC, Calderon MO (2001) Coral CRAN.R-project.org/package=vegan bleaching and mortality in Panama and Ecuador during the Oliver TA, Palumbi SR (2009) Distributions of stress-resistant coral 1997–1998 El Nino Oscillation Event: Spatial/temporal patterns symbionts match environmental patterns at local but not regional and comparisons with the 1982–1983 event. Bull Mar Sci 69: scales. Mar Ecol Prog Ser 378:93–109 79–109 Pratchett MS, Berumen ML, Marnane MJ, Eagle JV, Pratchett DJ Goreau TJ, Hayes RL (1994) Coral bleaching and ocean ‘‘hot spots’’. (2008) Habitat associations of juvenile versus adult butterfly Ambio 23:176–180 fishes. Coral Reefs 27:541–551 Goreau TJ, Hayes RL, McClanahan T (2000) Conservation of coral Pratchett MS, Wilson SK, Graham NJ, Munday PL, Jones GP, reefs after the 1998 global bleaching event. Conserv Biol 14:1–18 Polunin NC (2009) Coral bleaching and consequences for motile Harriott VJ (1985) Mortality rates of scleractinian corals before and reef organisms: Past, present, and uncertain future effects. In: during a mass bleaching event. Mar Ecol Prog Ser 21:81–88 van Oppen MJH, Lough JM (eds) Coral bleaching: patterns, Hoegh-Guldberg O (1999) Climate change, coral bleaching and the processes, causes and consequences. Springer, Heidelberg, future of the world’s coral reefs. Mar Freshw Res 50:839–866 pp 139–158 Hoeksema BW (1989) Taxonomy, phylogeny and biogeography of Pratchett MS, Trapon M, Berumen ML, Chong-Seng K (2011) Recent mushroom corals (Scleractinia: Fungiidae). Zool Verh Leiden disturbances augment community shifts in coral assemblages in 254:1–295 Moorea, French Polynesia. Coral Reefs 30:183–193 Hoeksema BW (1991) Control of bleaching in mushroom coral Raitsos DE, Hoteit I, Prihartato PK, Chronis T, Triantafyllou G, populations (Scleractina:Fungiidea) in the Java Sea: stress Abualnaja Y (2011) Abrupt warming of the Red Sea. Geophys tolerance and interference by life history strategy. Mar Ecol Res Lett 38 doi:10.1029/2011GL047984 Prog Ser 74:225–237 Riegl B, Piller WE (2003) Possible refugia for reefs in times of Hughes TP, Connell JH (1999) Multiple stressors on coral reefs: A environmental stress. Int J Earth Sci (Geol Rundsch) 92:520–531 long term perspective. Limnol Oceanogr 44:932–940 Roberts CM, McClean CJ, Veron JEN, Hawkins JP, Allen GR, Hughes TP, Bellwood DR, Connolly SR (2002) Biodiversity hotspots, McAllister DE, Mittermeier CG, Schueler FW, Spalding M, Wells centres of endemicity, and the conservation of coral reefs. Ecol F, Vynne C, Werner TB (2002) Marine biodiversity hotspots and Lett 5:775–784 conservation priorities for tropical reefs. Science 295:1280–1284 Jokiel PL, Brown EK (2004) Global warming, regional trends and Spaet JLY, Thorrold SR, Berumen ML (2012) A review of inshore environmental conditions influence coral bleaching in elasmobranch research in the Red Sea. J Fish Biol 80:952–965 Hawaii. Global Change Biol 10:1627–1641 Spalding M (2009) Detecting and monitoring coral bleaching events. Marshall PA, Baird AH (2000) Bleaching of corals on the Great In: van Oppen MJH, Lough JM (eds) Coral bleaching: patterns, Barrier Reef: differential susceptibilities among taxa. Coral processes, causes and consequences. Springer, Heidelberg, Reefs 19:155–163 pp 69–82 McClanahan TR (2004) The relationship between bleaching and Stella JS, Munday PI, Jones GP (2011) Effects of coral bleaching on mortality of common corals. Mar Biol 144:1239–1245 the obligate coral-dwelling crab Trapezia cymodoce. Coral Reefs McClanahan TR, Baird AH, Marshall PA, Toscano MA (2004) 30:719–737 Comparing bleaching and mortality responses of hard corals R Development Core Team (2011) R: a language and environment for between southern Kenya and the Great Barrier Reef, Australia. statistical computing. R Foundation for Statistical Computing, Mar Pollut Bull 48:327–335 Vienna, Austria. URL: http://www.R-project.org/ McClanahan TR, Ateweberhan M, Graham NAJ, Wilson SK, Ruiz Torres-Perez JL, Armstrong RA (2012) Effects of UV radiation on the Sebastiań C, Guillaume MMM, Bruggemann JH (2007) Western growth, photosynthetic and photoprotective components, and Indian Ocean coral communities: bleaching responses and reproduction of the Caribbean shallow-water coral Porites susceptibility to extinction. Mar Ecol Prog Ser 337:1–13 furcata. Coral Reefs. doi:10.1007/s00338-012-0927-7 McClanahan TR, Weil E, Cortes J, Baird AH, Ateweberhan M (2009) Toscano MA, Strong AE, Guch IC (1999) New analyses for ocean Consequences of bleaching for sessile reef organisms. In: van HotSpots and coral reef bleaching. Reef Encounter 26:31 Oppen MJH, Lough JM (eds) Coral bleaching: patterns, van Oppen MJH, Baker AC, Coffroth MA, Willis BL (2009) processes, causes and consequences. Springer-Verlag Berlin Bleaching resistance and the role of algal endosymbiosis. In: Heidelberg, Heidelberg, pp 121–138 van Oppen MJH, Lough JM (eds) Coral bleaching: Patterns, Mergner H (1984) The ecological research on coral reefs of the Red processes, causes and consequences. Springer, Heidelberg, Sea. Deep Sea Res Part A 31:855–884 pp 83–102 NOAA Coral Reef Watch (2000, updated twice-weekly) NOAA Coral Veron JEN (2000) Corals of the World, vol 1–3. Australian Institute Reef Watch Operational 50-km Satellite Coral Bleaching Degree of Marine Science, Townsville, Australia

123 176

APPENDIX 2: LIST OF CONFERENCE PRESENTATIONS

Bouwmeester J, Khalil MT, Berumen ML (2013) Coral recovery and recruitment after a bleaching event in the central Red Sea, 87th Australian Coral Reef Society Conference, Sydney, Australia (oral presentation)

Bouwmeester J, Baird AH, Berumen ML (2012) Synchronous spawning of scleractinian corals in the Red Sea, 12th International Coral Reef Symposium, Cairns, Australia (oral presentation)

Bouwmeester J, Baird AH, Berumen ML (2012) Spawning patterns of scleractinian corals in the Red Sea, Coral Reefs of the Gulf Conference, Abu Dhabi, U.A.E. (oral presentation)

Bouwmeester J, Baird AH, Berumen ML (2012) Coral reproduction in the Red Sea, Coral Reef Conservation in the Red Sea Symposium, Thuwal, Saudi Arabia (oral presentation)