Advertisement
Research Article

Predator Crown-of-Thorns Starfish (Acanthaster planci) Outbreak, Mass Mortality of Corals, and Cascading Effects on Reef Fish and Benthic Communities

  • Mohsen Kayal mail,

    mohsen.kayal@gmail.com

    Affiliations: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia, Laboratoire d'Excellence “CORAIL”, Institut de Recherche pour le Développement, U 227 COREUS 2, Nouméa, New Caledonia

    X
  • Julie Vercelloni,

    Affiliations: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia, School of Mathematical Sciences, Queensland University of Technology, Brisbane, Queensland, Australia, Australian Institute of Marine Science, Townsville, Queensland, Australia

    X
  • Thierry Lison de Loma,

    Affiliation: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia

    X
  • Pauline Bosserelle,

    Affiliation: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia

    X
  • Yannick Chancerelle,

    Affiliation: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia

    X
  • Sylvie Geoffroy,

    Affiliation: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia

    X
  • Céline Stievenart,

    Affiliation: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia

    X
  • François Michonneau,

    Affiliations: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia, Department of Biology, University of Florida, Gainesville, Florida, United States of America

    X
  • Lucie Penin,

    Affiliations: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia, Laboratoire d'Excellence “CORAIL”, Laboratoire d'Écologie Marine – ECOMAR, Université de La Réunion, Saint-Denis, La Réunion, France

    X
  • Serge Planes,

    Affiliation: Laboratoire d'Excellence “CORAIL”, USR 3278 CNRS-EPHE, Centre de Recherches Insulaires et Observatoire de l′Environnement, Moorea, French Polynesia

    X
  • Mehdi Adjeroud

    Affiliation: Laboratoire d'Excellence “CORAIL”, Institut de Recherche pour le Développement, U 227 COREUS 2, Nouméa, New Caledonia

    X
  • Published: October 08, 2012
  • DOI: 10.1371/journal.pone.0047363

Abstract

Outbreaks of the coral-killing seastar Acanthaster planci are intense disturbances that can decimate coral reefs. These events consist of the emergence of large swarms of the predatory seastar that feed on reef-building corals, often leading to widespread devastation of coral populations. While cyclic occurrences of such outbreaks are reported from many tropical reefs throughout the Indo-Pacific, their causes are hotly debated, and the spatio-temporal dynamics of the outbreaks and impacts to reef communities remain unclear. Based on observations of a recent event around the island of Moorea, French Polynesia, we show that Acanthaster outbreaks are methodic, slow-paced, and diffusive biological disturbances. Acanthaster outbreaks on insular reef systems like Moorea's appear to originate from restricted areas confined to the ocean-exposed base of reefs. Elevated Acanthaster densities then progressively spread to adjacent and shallower locations by migrations of seastars in aggregative waves that eventually affect the entire reef system. The directional migration across reefs appears to be a search for prey as reef portions affected by dense seastar aggregations are rapidly depleted of living corals and subsequently left behind. Coral decline on impacted reefs occurs by the sequential consumption of species in the order of Acanthaster feeding preferences. Acanthaster outbreaks thus result in predictable alteration of the coral community structure. The outbreak we report here is among the most intense and devastating ever reported. Using a hierarchical, multi-scale approach, we also show how sessile benthic communities and resident coral-feeding fish assemblages were subsequently affected by the decline of corals. By elucidating the processes involved in an Acanthaster outbreak, our study contributes to comprehending this widespread disturbance and should thus benefit targeted management actions for coral reef ecosystems.

Introduction

The crown-of-thorns seastar Acanthaster planci (Figure 1) is the major natural enemy of reef-building corals [1], [2]. This specialized coral-feeder is found on tropical reefs across the planet, except in the Atlantic Ocean. Populations of Acanthaster commonly display cyclic oscillations between extended periods of low-density with individuals scarcely distributed among large reef areas, and brief episodes of unsustainably high densities commonly termed ‘outbreaks’ [3]. These outbreaks are among the most destructive disturbances observed on tropical reefs [4], [5]. They result in mass mortalities of corals, sometimes annihilating populations, with typically second-order and long-term consequences on various communities [6][8]. Cascading effects of Acanthaster outbreaks usually spread to the entire reef ecosystem and commonly lead to increases in benthic algae, a loss of coral-feeding assemblages, an overall collapse of reef structural complexity, and a decline in biodiversity and productivity [6], [9][11] (see Figure 2 for illustrations). As a result, measures are often taken by local populations and management authorities to eradicate Acanthaster from reefs (e.g., [12]). However, such efforts often have limited success against the magnitude of outbreaks [6], [13]. Acanthaster outbreaks are increasingly documented [11], [14], yet these reports have mostly been restricted to short-term, sporadic observations (e.g., [5], [8], [15]). Until now, no study has quantitatively described the entire progression of an Acanthaster outbreak, including spatio-temporal dynamics of predator population and resultant impacts to the biological reef community. As a result, relatively little is known about the origins, development, or processes that influence the outcome of this disturbance [3], [7], [12], [16][26].

thumbnail

Figure 1. Photographs illustrating the outbreaking seastar Acanthaster and its feeding-scars as found on colonies preyed upon.

(A) An Acanthaster planci observed on a living tabular coral from the genus Acropora. (B) A partially-killed coral from the genus Acropora bearing feeding-scars left by successive predation events by Acanthaster: 1) live portion of the colony bearing the pigmented coral tissue, 2) freshly killed portion of the colony deprived of its pigmented living tissue (<1 day post-predation), 3) recently killed portion of the colony covered by early colonizing algae and cyanobacteria (~10 days post-predation), 4) dead portion of the colony killed long ago and covered by turf algae (>3 weeks post-predation). © Photos Mohsen Kayal.

doi:10.1371/journal.pone.0047363.g001
thumbnail

Figure 2. A portion of Moorean outer-reef (6 m-depth on the site Vaipahu) is shown through time.

(A) Corals dominate the healthy reef (coral cover >40%). (B) Algae have colonized dead coral skeletons following severe predation by the seastar Acanthaster (~10% coral cover). (C) Mostly dead and weakened coral skeletons were swept away by a cyclone occurring at the end of the seastar outbreak (Lison de Loma et al. unpublished data) and colonizing algae once again dominate the devastated reef (~5% coral cover). © Photos Mohsen Kayal.

doi:10.1371/journal.pone.0047363.g002

Here we describe an Acanthaster outbreak that occurred during the last decade in French Polynesia (South Pacific), a region where these disturbances occur with a periodicity of ~20 years and, along with bleaching events and cyclones, are the major drivers of community dynamics on coral reefs [27]. Our study has specifically focused on the island of Moorea, where we tracked the distribution of Acanthaster aggregations and quantified their impacts on corals, other sessile communities, and resident coral-feeding fishes. In contrast to prior studies that were mostly restricted to reporting a posteriori observations of the consequences of Acanthaster on reefs, we quantitatively describe the processes leading to the community changes resulting from Acanthaster outbreaks.

Methods

This study was approved and conducted as part of ongoing research of the Centre de Recherches Insulaires et Observatoire de l′Environnement (CRIOBE, USR 3278 CNRS-EPHE, LABEX “CORAIL”).

Spatio-temporal scope of the study

Starting in 2002, unusually elevated densities of Acanthaster were progressively reported from the different high volcanic islands of the Society Archipelago (Tahiti, Moorea, Huahine, Raiatea, Tahaa, Bora Bora, Maupiti) and then from the Australes (Rurutu) in French Polynesia. These islands are scattered over a broad geographical scale spreading 675 km north-south and 330 km east-west. Our study was conducted in Moorea (17°30′ S, 149°50′ W, see Figure 3), where long-term reef monitoring sites have been sampled for ~40 years, and where the first Acanthaster aggregation was observed in 2003. Two complementary sampling approaches were used to quantify the dynamics of this outbreak and its consequences on reef communities. The first sampling approach was a periodic survey of various key functional assemblages among benthic and fish communities. These surveys were conducted at a small scale covering few hundred square meters of reef at nine reference reef locations, consisting of three water depths (6, 12, 18 m) at each of three sites (Vaipahu, Tiahura, Haapiti). The second sampling approach consisted of a yearly survey of seastar aggregations conducted at a large scale all around Moorea and its ~100 km perimeter of reefs. The small-scale surveys were initiated between 2003 and 2005 depending on the different assemblages sampled (see Sampling section below), whereas the large-scale survey of seastars started in 2006, once Acanthaster aggregations were observed to spread to multiple sides of the island. Surveys were conducted until 2010, as long as remaining aggregations were observed. All the sampling was conducted using SCUBA on the outer reef slopes where, in Moorea as in other islands in French Polynesia, the highest coral biomass and the most diverse reef communities are concentrated [28]. This is also where Acanthaster aggregations were systematically first observed (refer to results of this study). Around Moorea, the outer reef habitat typically extends from the water surface at the crest of the barrier-reef where oceanic waves break, down to a depth of ~35 m where sand plains begin. A peak in diversity and coral coverage is typically observed at the 10–20 m depth range [28], [29]. These outer reef habitats are exposed to the open ocean, undergo relatively little direct human pressure, and experience maximum exposure to natural disturbances [27]. By the end of the Acanthaster outbreak, Moorean reefs underwent the additional impacts of the tropical cyclone Oli (Lison de Loma et al. unpublished data) whose immediate effects were partially captured by the present study.

thumbnail

Figure 3. Distributions and densities of Acanthaster feeding-scars as observed around Moorea through time.

The dimensions of the circles are proportional to the abundance of scars (n scars per 2 min-towing section), and a color code is used to distinguish different abundance classes (refer to the legend). A mean relation of 8.6±1.7 SE scars per individual seastar was estimated during the outbreak (see Table S1). The positions of the three reference sites where reef communities were surveyed are also displayed on the graphs: Haapiti (H), Tiahura (T), Vaipahu (V). Original satellite image from © Google Earth.

doi:10.1371/journal.pone.0047363.g003

Sampling strategy

Nine reef locations were surveyed on Moorean outer reefs in order to quantify community dynamics during the Acanthaster outbreak (see above). The composition of the sessile communities (i.e., relative coverage of coral populations and of other benthic components) was sampled in n = 10 random quadrats by recording the type of substrate beneath 81 points defined by a grid of 10 cm-mesh within the 1 m2 areas. As major reef corallivores, populations of the outbreaking seastar Acanthaster and resident coral-feeding butterflyfishes were surveyed in n = 3 replicate belt-transects (50×4 m) placed randomly along constant depth contours. These fishes belong to the genus Chaetodon which encompasses several specialized coral-feeding species whose populations are tightly associated with corals [1], [11]. The corallivorous representatives commonly encountered on Moorean outer reefs are C. ornatissimus, C. reticulatus, C. trifascialis, C. lunulatus, C. auriga, C. lunula, C. pelewensis, C. ulietensis, C. unimaculatus, C. quadrimaculatus [10], [30], [31]. To track the distributions and densities of Acanthaster aggregations at an island-scale around Moorea, we developed the SCUBA-tow technique, an adaptation of the manta-tow [8], [32] conducted using SCUBA [33]. The observer was positioned at 8 m-depth in the water column, ~5 m above the reef substrate at a constant depth of ~13 m, and towed by a boat all around the island at a speed of 4 knots (7.4 km h−1). In clear oceanic waters surrounding Moorea, this position of the observer allowed for a survey of the reef substrate in an ~30 m-wide band in the 10–30 m depth range. Due to the cryptic character of Acanthaster [6], the density of seastars was estimated here by counting the number of characteristic feeding-scars found on colonies that were recently preyed upon (Figure 1). These white scars are denuded portions of the coral skeleton recently deprived of their pigmented living tissues, and can be used for tracking recent predation events [21], [34]. In the oligotrophic oceanic waters of French Polynesia, Acanthaster feeding-scars remain clearly visible for ~3 weeks before being covered by colonising algae and other sessile organisms (see Figure 1). Counts were conducted in sections of 2 min of towing each covering a portion of ~7,500 m2 of reef, and GPS coordinates were simultaneously recorded. The correspondence between the number of scars and the density of seastars was established by subsequent counts performed in transects on a restricted number of sites throughout the process of the outbreak (see Table S1).

Statistical analysis

Variability in the coverage of benthic communities and in the density of butterflyfish assemblages was tested using three-way nested ANOVAs in which Time was nested within Depth, and Depth nested within Site. When significant differences were detected by ANOVA, Fisher's Least Significant Differences (LSD) post-hoc test was used to compare data among groups. Linear Mixed Models (LMMs) were used to test for correlations between the dynamics of the different reef communities: Acanthaster, corals, other sessile communities, and coral-feeding butterflyfishes. LMMs have the advantage of taking into account correlated observations, and were also used to examine how populations of different coral genera were affected during the decline of coral communities facing Acanthaster. This was performed by drawing a linear regression between the coverage of each coral genus (dependant variable Genus cover) and the coverage of the overall coral community (explicative covariable Coral cover) as quantified at our nine reference reef locations on four different years during the process of the outbreak (2005, 2008, 2009, 2010). For these regressions, autocorrelations were tested for the fixed effects of the covariable Coral cover, the random effects of the grouping factor Reef location (as the result of the interaction Site × Depth), and their interaction Coral cover × Reef location, and were taken into account in the calculation of parameters where significant [35], [36]. Before ANOVAs and LMMs, data were tested for normality and homoscedasticity, and were transformed when needed. Acanthaster and butterflyfish densities were log(x+1) transformed, and arcsin(√x) transformation was applied to percent-cover data. All statistics were computed in R version 2.12.0 (R Development Core Team 2008) complemented by the NLME package [36].

Results

Acanthaster outbreak

In Moorea, the first aggregation of Acanthaster was observed at 18 m-depth at the outer reef site Tiahura in October 2003 (Figure 4). The swarms of seastars affected the Vaipahu site also situated on the north shore of the island in May 2004. The Haapiti site on the west coast was affected in March 2006. SCUBA-tows performed in late 2006 showed that particularly high densities of Acanthaster feeding-scars (>150 scars per 2 min-towing section) were mostly concentrated on the north-eastern corner of Moorea, with elevated densities also found near Tiahura and on the central west side of the island (Figure 3). These scars were used as indirect evidences of recent Acanthaster predation on corals (see Figure 1), and an average ratio of 8.6±1.7 SE feeding-scars per seastar was calculated over the process of the outbreak (Table S1). Through consecutive years, these intense predation events spread over new reefs that were not yet affected, eventually affecting the entire coastline of Moorea. By 2009, most Acanthaster predation was concentrated near the southern tip of the island, with few feeding-scars observed on the formerly affected north shore (Figure 3).

thumbnail

Figure 4. Dynamics of various communities surveyed at the reference reef locations.

These nine locations consist of three sites (Haapiti: H, Tiahura: T, Vaipahu: V) × three water depths (6, 12, 18 m). Y-axes on the left indicate cover values (mean ± SE) of the sessile communities: reef-building corals and other benthic components. Y-axes on the right indicate densities (mean ± SE) of coral-predators: populations of the outbreaking seastar Acanthaster and butterflyfish assemblages. Arrows on the x-axes indicate the occurrence of the tropical cyclone Oli. Refer to Figure 5 for correlations between the dynamics of different communities. See Figure 2 for an illustration of the changes observed on reefs.

doi:10.1371/journal.pone.0047363.g004

Expanding waves of Acanthaster swarms developed similarly at all sites surveyed, consistently starting at the deepest locations of the outer reef and progressing upward with the migration of seastars along the reef-slope (Figure 4). This pattern varied among stations in terms of observed peak densities (min. 2.3±0.3 SE ind.200 m−2 or 11,500 ind.km−2 at Haapiti-6 m; max. 30.3±6.1 SE ind.200 m−2 or 151,650 ind.km−2 at Vaipahu-6 m) and residence times of predators (min. 30 months at Haapiti-6 m; max. 72 months at Tiahura-18 m), which, combined with the sequential time of arrival of seastar swarms at the different reef locations, resulted in complex spatio-temporal variability in the rate of predation on corals. This generated asynchronicity in the decline of corals among sites and depths (three-way nested ANOVA, factor Time(Depth(Site)), p<0.01; see Figure 4). The upward migration of aggregated seastars on the reef slope was observed in one (Vaipahu) or several (Tiahura and Haapiti) slow waves. By April 2010, densities of Acanthaster had fallen to zero on all surveyed reef locations, and no additional individuals were observed during subsequent surveys.

Impacts on corals and other reef communities

Acanthaster predation resulted in a sharp collapse of coral populations and communities (Fisher's p<0.05 between consecutive samplings). This decline of corals progressively affected the different reef locations, as the seastars migrated through the reef system (Figures 4 and 5). The coral coverage decreased gradually from values mostly above 40% in 2005 to values often below 5% in 2010, sometimes <1% with the combined effects of the cyclone (at 12 m depth on the sites Tiahura and Vaipahu). Mass mortality of corals was accompanied by a decline in the diversity of coral assemblages, and was correlated with an increase in turf algae and dead-coral rubble and sand substrates (refer to the slopes of regressions in Figure 5, p(a)<0.05). In contrast, no significant trend was observed in the cover of macro-algae, coralline algae, soft corals, or sponges (p(a)>0.05, see Figure 4 and 5).

thumbnail

Figure 5. Correlations between the dynamics of different communities surveyed at the reference reef locations.

Each point on the graphs (n = 36) represents the value (mean ± SE) of one sampling year (2005, 2008, 2009, 2010) on one site (Vaipahu, Tiahura, Haapiti) at one depth (6, 12, 18 m); refer to Figure 4. The equations and significance values (p) of the regression lines are given on the graphs (with a, the slope, and b, the intercept). The dynamics of the dominant coral genera are plotted against the dynamics of entire coral communities to quantify the sequential extirpation of populations by Acanthaster. An asterisk (*) indicates regressions for which significant autocorrelations were detected and taken into account in the calculation of parameters. For Acropora, data are split in two groups to distinguish samplings conducted during the decline phase (Acropora-cover >0.5%, solid dots, dashed regression line) and past the virtual extirpation of these populations from reefs (Acropora-cover <0.5%, empty circles, no regression line), the continuous line being the regression over the whole data set. For Montipora, data are split in two groups to distinguish samplings conducted at Haapiti-6 m where this taxon is predominating (empty circles, short-dashed regression line; see Figure 4) from the other stations (solid dots, long-dashed regression line), the continuous line being the regression over the whole data set. For coral genera whose populations were extirpated from the reef significantly earlier than the total coral community (i.e., p(b)<0.05), the x-intercept X0 (± SE) is also displayed.

doi:10.1371/journal.pone.0047363.g005

Coral genera were not equally impacted by Acanthaster predation, which resulted in a sequential extirpation of taxa from local communities (Figures 4 and 5). Branching and table-shaped species belonging to the genus Acropora were affected first and most heavily. Their populations declined abruptly and were extirpated from the reefs by the time coral cover fell below values of 13.7% (refer to the x-intercepts of regressions in Figure 5, arcsin(√X0) = 0.38±0.03 SE). The collapse of Acropora populations was followed by those of sub-branching Pocillopora (eliminated at 1.1% coral cover, arcsin(√X0) = 0.11±0.02 SE). Populations of encrusting Montipora, massive Porites, and other hard-coral assemblages also declined, showing a synchronized collapse with the entire coral communities (refer to the intercepts of regressions in Figure 5, p(b)>0.05). The calcifying hydrozoan Millepora was rarely preyed upon by the seastars, and its populations did not vary with the decline in live coral. The selective predation of Acanthaster on corals, combined with the sequential time of arrival of seastars at the different reef locations and the variability in the observed peak densities, generated high spatio-temporal heterogeneity in the structure of benthic communities. Within the seven years of the Acanthaster outbreak on Moorea (2003–2010), a gradual shift of the outer-reef system was observed, from a coral-dominated one, to an ecosystem where space was mainly occupied by turf algae, coralline algae, rubble, and sand (Figure 4). The community shift from corals to algae coincided with a shift in the composition of coral assemblages, from a state where branching genera Acropora and Pocillopora dominated in 2005, to one almost exclusively occupied by massive Porites in 2010.

Assemblages of coral-feeding butterflyfishes showed a tight correlation with corals in terms of size and diversity, resulting in a synchronous collapse of these populations with the mortality of corals (Figures 4 and 5). In February 2010, cyclone Oli generated ~8 m waves that strongly affected the reef landscape by breaking and removing many live and dead coral skeletons at our study locations (see Figure 2). However, the potential effects of this second disturbance were partly diminished by the prior and ongoing occurrence of the Acanthaster outbreak which already had shown profound impacts on reef communities. The cyclone did not modify the pattern of decline in coral cover, increase in turf algae, and collapse of coral-feeding butterflyfishes as initiated since the beginning of the Acanthaster outbreak (Figure 4). After the passage of Acanthaster and cyclone, the once flourishing and polymorphic coral communities were mostly restricted to surviving fragments of massive Porites scattered among opportunistic algae (refer to Figure 2 for a synthetic illustration of the changes observed on the reef landscape).

Discussion

Periodic outbreaks of the coral predator seastar Acanthaster planci constitute major disturbances to reef ecosystems in several regions throughout the Indo-Pacific [4], [5], [27]. Yet lack of observations of the development of these disturbances has restrained our knowledge of the ecological processes surrounding these events. During a particularly intense episode of Acanthaster outbreak around Moorea, French Polynesia, elevated Acanthaster densities spread from restricted source areas at the base of the northern outer reef and over several years propagated to the entire insular reef system. This propagation was based on a consecutive migration of Acanthaster aggregations toward unaffected adjacent and shallower reef locations. These waves of predatory seastars strongly impacted coral communities by decimating populations as encountered across reefs. The end of the outbreak coincided with the decimation of corals on the last affected reefs at the south of the island. This Acanthaster outbreak has been the most intense disturbance recorded on Moorean reefs since the establishment of scientific observations on this island about 40 years ago [27].

As observed in Moorea, Acanthaster outbreaks typically start at deeper locations at the base of reefs, where elevated cover in dead-coral rubble and coralline algae possibly favor settlement of the seastar larvae, provide shelter and food for the young juveniles that feed on coralline algae, and promote Acanthaster recruitment into adult coral-eating populations [6], [37][39]. The observed aggregative behavior of Acanthaster during outbreaks is thought to promote reproductive success [3], [6], [40], while resulting in the mass mortality of corals. Acanthaster has been shown to move relatively little in the presence of adequate food, however movements increase with higher densities of individuals and lower prey availability [6], [21]. During aggregations such as those recently observed in Moorea, the rapid local shortage in coral prey seems to engender an intensified foraging behaviour in Acanthaster. This behaviour is probably the major engine of the observed waves of migration, leading starving seastars to search for food in surrounding localities and spreading densities to unaffected reef locations. Such hunger-motivated directional movement of Acanthaster during outbreaks has already been suggested [38], [40], [41], and may explain the formation of high-density Acanthaster feeding fronts, as it is observed in other species (see [42], [43]). Interestingly, during the previous outbreak of this predator that was observed around Moorea in the early 1980′s, the first individuals of Acanthaster were also reported from the north shore of the island near the pass Taotoi [34], which is situated close to our site Tiahura where the first individuals were observed for the outbreak reported here. It remains unclear why this specific area would constitute a favourable settlement spot, nursery, or aggregating area for the development of Acanthaster outbreaks. The base of outer reefs on the north coast of Moorea accumulate relatively high concentrations of coral-rubble covered by coralline algae [29], which could favour recruitment of seastars [6], [39]. Yet further investigation is still needed to elucidate why specific reef locations constitute potential sources for Acanthaster infestations.

Acanthaster outbreaks typically induce considerable declines in corals, however the magnitude of decline is highly variable among outbreak events [5]. The coral communities around Moorea had shown relatively little fluctuations in size and structure since the turn of the millennium [27], and were drastically depleted by Acanthaster within a few months. Feeding preferences of Acanthaster consistently alter the structure of coral communities toward dominance by non-preferred species [1−3], [23], . Our observations show how this food selectivity results in a singular scheme of coral decline on affected reefs: Acanthaster hierarchically consumes preferred species and sequentially extirpates local populations. Density and residence time of predator Acanthaster, and local abundance and composition in prey corals, thus influence the magnitude of coral decline and the structure of surviving coral communities (see also [7]). In French Polynesia as on other reefs throughout the Indo-Pacific affected by periodic outbreaks, targeted attacks of Acanthaster on faster growing branching Acropora and Pocillopora populations result in episodic shifts of coral communities toward a temporary dominance by slower growing massive Porites [1], [2], [7], [15]. Thus, these natural disturbances constitute important historical drivers that shape the structure of coral communities in these regions.

Alterations of coral communities by Acanthaster are accompanied by subsequent changes in the demography of various reef species. As major competitors of corals, algae communities typically increase during outbreaks by colonizing the space released as corals die [16], [37], [38]. The recent increase in algae around Moorea has further been correlated to increases in herbivore assemblages, which in turn are preventing the development of macro-algal blooms as observed following coral mortality on other reefs [14], [44], [45]. Another re-emerging consequence of Acanthaster outbreaks is the collapse of resident corallivore assemblages that suffer from trophic limitations following the decimation of corals [9][11], [46]. Such shortage in food, rather than loss of refuges and habitat, was probably the main driver of the observed decline of coral-feeding butterlyfishes in Moorea. Indeed, Acanthaster predation does not alter the skeleton of corals and, over the short term, leaves the reef framework unaffected [1], [5], [11] (see Figures 1 and 2); yet these fishes showed a synchronous collapse with the decline of live coral during the outbreak, and were highly decimated on most reef areas before the physical alteration of their habitats by the cyclone. Similar Acanthaster-mediated loss of corallivores was also reported in decapod communities living within the branches of corals [26].

Following the relatively slow and diffusive devastation of corals by Acanthaster, the passage of the cyclone Oli resulted in an additional pulse disturbance that mostly affected the north shore of Moorea. This disturbance literally flattened the reef topography by breaking and removing the mostly-dead coral skeletons that remained after Acanthaster predation (Lison de Loma et al. unpublished data; see Figure 2 for illustrations). However, losses of reef structural complexity over the long-term have been attributed to Acanthaster outbreaks alone [9], [11]. While the cyclone showed limited immediate effects on surveyed benthic and fish species that were previously impacted by the seastar outbreak, the loss of physical structure on reefs likely had detrimental effects on other communities relying on corals as habitats and refuges [11], [26], [47], [48]. Furthermore, loss in reef structure will undoubtedly influence community regulation in the long-term [7], and may hinder the resilience of Moorean reefs. However, in contrast with a trend increasingly observed on reefs [4], [14], the strict regulation of algal communities to early-colonizing turf forms and the continuous flow of coral larvae among reefs in this region should lead to a progressive recovery of Moorean reefs, despite the tight recurrence of recent disturbances [27], [44], [45], [49].

Conclusion

Our observations of an Acanthaster outbreak around Moorea coincide with records of a prior outbreak on this island [34] and with limited observations on other islands throughout French Polynesia also affected by the recent wave of infestations (unpublished data). Far from being unorganized and random events, outbreaks of the coral predator Acanthaster planci appear as ordered, relatively slow and diffusive biological disturbances. On the reefs surrounding the high volcanic islands of French Polynesia, these outbreaks were observed to originate from localized source areas situated at the base of outer-reef slopes, and to progressively spread to the entire reef systems by aggregative migrations of seastars. This pattern of propagation of Acanthaster from deeper parts of reefs toward unaffected locations also coincides with reports of previous outbreaks from other regions [38], [40], [41]. These seemingly hunger-driven assaults on corals decimate populations in a predictable sequence determined by feeding preferences consistently observed for Acanthaster [7], [15], [23], and their effects typically cascade down to many reef communities whose fates are directly or indirectly related to corals [9][11], [26], [40]. These findings improve our understanding of reef dynamics and have critical implications for management of coral ecosystems where Acanthaster is observed. We advocate the importance of monitoring the ocean-orientated bases of reefs, particularly those where coral-rubble and coralline algae are abundant. These measures could help detect Acanthaster outbreaks at the earliest stages and, when appropriate, improve the efficiency of control efforts.

Supporting Information

Table S1.

Ratios of the density of feeding-scars to the number of predator seastar Acanthaster as observed in transect-counts on Moorean reefs during the outbreak.

doi:10.1371/journal.pone.0047363.s001

(PDF)

Acknowledgments

We thank volunteers from CRIOBE and Planète Urgence for assistance in the field, and Madeleine Thwaites, Michael Gil, Pr Hunter Lenihan, and two anonymous reviewers for comments that improved the manuscript.

Author Contributions

Conceived and designed the experiments: MK TLDL MA. Performed the experiments: MK TLDL PB YC SG CS FM LP MA. Analyzed the data: MK JV. Contributed reagents/materials/analysis tools: MK JV TLDL SP MA. Wrote the paper: MK JV TLDL MA.

References

  1. 1. Rotjan RD, Lewis SM (2008) Impact of coral predators on tropical reefs. Mar Ecol-Prog Ser 367: 73–91 doi:10.3354/meps07531. doi: 10.3354/meps07531
  2. 2. Stella JS, Pratchett MS, Hutchings PA, Jones GP (2011) Coral-associated invertebrates: diversity, ecology importance and vulnerability to disturbance. Oceanogr Mar Biol Annu Rev 49: 43–104. doi: 10.1201/b11009-3
  3. 3. Uthicke S, Schaffelke B, Byrne M (2009) A boom–bust phylum? Ecological and evolutionary consequences of density variations in echinoderms. Ecol Monogr 79: 3–24 doi:10.1890/07-2136.1. doi: 10.1890/07-2136.1
  4. 4. Bruno JF, Selig ER (2007) Regional decline of coral cover in the Indo-Pacific: timing, extent, and subregional comparisons. PLoS ONE 2: e711 doi:10.1371/journal.pone.0000711. doi: 10.1371/journal.pone.0000711
  5. 5. Osborne K, Dolman AM, Burgess SC, Johns KA (2011) Disturbance and the dynamics of coral cover on the Great Barrier Reef (1995–2009). PLoS ONE 6: e17516 doi:10.1371/journal.pone.0017516. doi: 10.1371/journal.pone.0017516
  6. 6. Moran PJ (1990) Acanthaster planci (L.): biographical data. Coral Reefs 9: 95–96 doi:10.1007/BF00258218. doi: 10.1007/bf00258218
  7. 7. Kayal M, Lenihan HS, Pau C, Penin L, Adjeroud M (2011) Associational refuges among corals mediate impacts of a crown-of-thorns starfish Acanthaster planci outbreak. Coral Reefs 30: 827–837 doi:10.1007/s00338-011-0763-1. doi: 10.1007/s00338-011-0763-1
  8. 8. Sweatman H, Delean S, Syms C (2011) Assessing loss of coral cover on Australia's Great Barrier Reef over two decades, with implications for longer-term trends. Coral Reefs 30: 521–531 doi:10.1007/s00338-010-0715-1. doi: 10.1007/s00338-010-0715-1
  9. 9. Sano M (2000) Stability of reef fish assemblages: responses to coral recovery after catastrophic predation by Acanthaster planci. Mar Ecol-Prog Ser 198: 121–130 doi:10.3354/meps198121. doi: 10.3354/meps198121
  10. 10. 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 doi:10.1007/s00338-006-0145-2. doi: 10.1007/s00338-006-0145-2
  11. 11. Wilson SK, Graham NAJ, Pratchett MS, Jones GP, Polunin NVC (2006) Multiple disturbances and the global degradation of coral reefs: are reef fishes at risk or resilient? Glob Change Biol 12: 2220–2234 doi:10.1111/j.1365-2486.2006.01252.x. doi: 10.1111/j.1365-2486.2006.01252.x
  12. 12. Houk P, Raubani J (2010) Acanthaster planci outbreaks in Vanuatu coincide with ocean productivity, furthering trends throughout the Pacific Ocean. J Oceanogr 66: 435–438 doi:10.1007/s10872-010-0038-4. doi: 10.1007/s10872-010-0038-4
  13. 13. Johnson DB, Moran PJ, Driml S (1990) Evaluation of a crown-of-thorns starfish (Acanthaster planci) control program at Grub Reef (central Great Barrier Reef). Coral Reefs 9: 167–171 doi:10.1007/BF00258231. doi: 10.1007/bf00258231
  14. 14. Bellwood DR, Hughes TP, Folke C, Nyström M (2004) Confronting the coral reef crisis. Nature 429: 827–833 doi:10.1038/nature02691. doi: 10.1038/nature02691
  15. 15. Pratchett MS (2010) Changes in coral assemblages during an outbreak of Acanthaster planci at Lizard Island, northern Great Barrier Reef (1995–1999). Coral Reefs 29: 717–725 doi:10.1007/s00338-010-0602-9. doi: 10.1007/s00338-010-0602-9
  16. 16. Dulvy NK, Freckleton RP, Polunin NVC (2004) Coral reef cascades and the indirect effects of predator removal by exploitation. Ecol Lett 7: 410–416 doi:10.1111/j.1461-0248.2004.00593.x. doi: 10.1111/j.1461-0248.2004.00593.x
  17. 17. Pratchett MS (2005) Dynamics of an outbreak population of Acanthaster planci at Lizard Island, northern Great Barrier Reef (1995–1999). Coral Reefs 24: 453–462 doi:10.1007/s00338-005-0006-4. doi: 10.1007/s00338-005-0006-4
  18. 18. Sweatman H (2008) No-take reserves protect coral reefs from predatory starfish. Curr Biol 18: R598–R599 doi:10.1016/j.cub.2008.05.033. doi: 10.1016/j.cub.2008.05.033
  19. 19. Fabricius KE, Okaji K, De'ath G (2010) Three lines of evidence to link outbreaks of the crown-of-thorns seastar Acanthaster planci to the release of larval food limitation. Coral Reefs 29: 593–605 doi:10.1007/s00338-010-0628-z. doi: 10.1007/s00338-010-0628-z
  20. 20. Timmers MA, Bird CE, Skillings DJ, Smouse PE, Toonen RJ (2012) There's no place like home: crown-of-thorns outbreaks in the Central Pacific are regionally derived and independent events. PLoS ONE 7: e31159 doi:10.1371/journal.pone.0031159. doi: 10.1371/journal.pone.0031159
  21. 21. Keesing JK, Lucas JS (1992) Field measurement of feeding and movement rates of the crown-of-thorns starfish Acanthaster planci (L.). J Exp Mar Biol Ecol 156: 89–104 doi:10.1016/0022-0981(92)90018-6. doi: 10.1016/0022-0981(92)90018-6
  22. 22. De'ath G, Moran PJ (1998) Factors affecting the behaviour of crown-of-thorns starfish (Acanthaster planci L.) on the Great Barrier Reef: 1: Patterns of activity. J Exp Mar Biol Ecol 220: 83–106 doi:10.1016/S0022-0981(97)00085-3. doi: 10.1016/s0022-0981(97)00085-3
  23. 23. De'ath G, Moran PJ (1998) Factors affecting the behaviour of crown-of-thorns starfish (Acanthaster planci L.) on the Great Barrier Reef: 2: Feeding preferences. J Exp Mar Biol Ecol 220: 107–126 doi:10.1016/S0022-0981(97)00100-7. doi: 10.1016/s0022-0981(97)00100-7
  24. 24. Pratchett MS (2001) Influence of coral symbionts on feeding preferences of crown-of-thorns starfish Acanthaster planci in the western Pacific. Mar Ecol Prog Ser 214: 111–119 doi:10.3354/meps214111. doi: 10.3354/meps214111
  25. 25. Bergsma GS (2012) Mutualists alter coral susceptibility and response to biotic disturbance through cascading trait-mediated indirect interactions. Coral Reefs 31: 461–469 doi:10.1007/s00338-011-0861-0. doi: 10.1007/s00338-011-0861-0
  26. 26. Leray M, Béraud M, Anker A, Chancerelle Y, Mills SC (2012) Acanthaster planci outbreak: decline in coral health, coral size structure modification and consequences for obligate decapod assemblages. PLoS ONE 7: e35456 doi:10.1371/journal.pone.0035456. doi: 10.1371/journal.pone.0035456
  27. 27. Adjeroud M, Michonneau F, Edmunds PJ, Chancerelle Y, Lison de Loma T, et al. (2009) Recurrent disturbances, recovery trajectories, and resilience of coral assemblages on a South Central Pacific reef. Coral Reefs 28: 775–780 doi:10.1007/s00338-009-0515-7. doi: 10.1007/s00338-009-0515-7
  28. 28. Adjeroud M (1997) Factors influencing spatial patterns on coral reefs around Moorea, French Polynesia. Mar Ecol-Prog Ser 159: 105–119 doi:10.3354/meps159105. doi: 10.3354/meps159105
  29. 29. Battistini R, Bourrouilh F, Chevalier JP, Coudray J, Denizot M, et al. (1975) Eléments de terminologie récifale indopacifique. Thétys 7: 1–111.
  30. 30. Harmelin-Vivien ML, Bouchon-Navaro Y (1983) Feeding diets and significance of coral feeding among Chaetodontid fishes in Moorea (French Polynesia). Coral Reefs 2: 119–127. doi: 10.1007/bf02395282
  31. 31. Penin L, Michonneau F, Baird AH, Connolly SR, Pratchett MS, et al. (2010) Early post-settlement mortality and the structure of coral assemblages. Mar Ecol Prog Ser 408: 55–64 doi:10.3354/meps08554. doi: 10.3354/meps08554
  32. 32. Moran PJ, De'ath G (1992) Suitability of the manta tow technique for estimating relative and absolute abundances of crown-of-thorns starfish (Acanthaster planci L.) and corals. Aust J Mar Fresh Res 43: 357–378 doi:10.1071/MF9920357. doi: 10.1071/mf9920357
  33. 33. Chancerelle Y, Maurin R, Poujade S, Vieux C (2008) Photo tow: a new method for estimating coral reef status and changes at large spatial scale. A coral bleaching event case study. J Soc Océanistes 126–127: 68–73. doi: 10.4000/jso.3312
  34. 34. Faure G (1989) Degradation of coral reefs at Moorea island (French Polynesia) by Acanthaster planci. J Coastal Res 5: 295–305.
  35. 35. Fox J (2002) Linear mixed models: appendix to an R and S-PLUS companion to applied regression. Available on R home page http://cran.r-project.org.
  36. 36. Pinheiro J, Bates D, DebRoy S, Sarkar D (2008) the R Development Core Team (2008) nlme: Linear and Nonlinear Mixed Effects Models. R package version 3: 1–89.
  37. 37. Endean R, Stablum W (1973) A study of some aspects of the crown-of-thorns starfish (Acanthaster planci) infestations of reefs of Australia's Great Barrier Reef. Atoll Res Bull 167: 1–76. doi: 10.5479/si.00775630.167.1
  38. 38. Laxton LH (1974) Aspects of the ecology of the coral-eating starfish Acanthaster planci. Biol J Linnaean Soc 6: 19–45 doi:10.1111/j.1095-8312.1974.tb00712.x. doi: 10.1111/j.1095-8312.1974.tb00712.x
  39. 39. Johnson CR, Sutton DC, Olson RR, Giddins R (1991) Settlement of crown-of-thorns starfish: role of bacteria on surfaces of coralline algae and a hypothesis for deepwater recruitment. Mar Ecol-Prog Ser 71: 143–162 doi:10.3354/meps071143. doi: 10.3354/meps071143
  40. 40. Ormond RFG, Campbell AC (1974) Formation and breakdown of Acanthaster planci aggregations in the Red Sea. Proc 2nd Int Coral Reef Symp 1: 577–590.
  41. 41. Suzuki G, Kai S, Yamashita H (2012) Mass stranding of crown-of-thorns starfish. Coral Reefs, in press. doi:10.1007/s00338-012-0906-z.
  42. 42. Silliman BR, van de Koppel J, Bertness MD, Stanton LE, Mendelssohn IA (2005) Drought, snails, and large-scale die-off of southern U.S. salt marshes. Science 310: 1803–1806 doi:10.1126/science.1118229. doi: 10.1126/science.1118229
  43. 43. Lauzon-Guay J-S, Scheibling RE, Barbeau MA (2008) Formation and propagation of feeding fronts in benthic marine invertebrates: a modeling approach. Ecology 89: 3150–3162 doi:10.1890/07-0643.1. doi: 10.1890/07-0643.1
  44. 44. Adam TC, Schmitt RJ, Holbrook SJ, Brooks AJ, Edmunds PJ, et al. (2011) Herbivory, connectivity, and ecosystem resilience: response of a coral reef to a large-scale perturbation. PLoS ONE 6: e23717 doi:10.1371/journal.pone.0023717. doi: 10.1371/journal.pone.0023717
  45. 45. Mumby PJ, Steneck RS (2008) Coral reef management and conservation in light of rapidly evolving ecological paradigms. Trends Ecol Evol 23: 555–563 doi:10.1016/j.tree.2008.06.011. doi: 10.1016/j.tree.2008.06.011
  46. 46. Bouchon-Navaro Y, Bouchon C, Harmelin-Vivien ML (1985) Impact of coral degradation on a Chaetodontid fish assemblage (Moorea, French Polynesia). Proc 5th Int Coral Reef Symp 5: 427–432. doi: 10.1007/bf02395282
  47. 47. Graham NAJ, McClanahan TR, MacNeil MA, Wilson SK, Polunin NVC, et al. (2008) Climate warming, marine protected areas and the ocean-scale integrity of coral reef ecosystems. PLoS ONE 3: e3039 doi:10.1371/journal.pone.0003039. doi: 10.1371/journal.pone.0003039
  48. 48. Halford AR, Caley MJ (2009) Towards an understanding of resilience in isolated coral reefs. Glob Change Biol 15: 3031–3045 doi:10.1111/j.1365-2486.2009.01972.x. doi: 10.1111/j.1365-2486.2009.01972.x
  49. 49. Magalon H, Adjeroud M, Veuille M (2005) Patterns of genetic variation do not correlate with geographical distance in the reef-building coral Pocillopora meandrina in the South Pacific. Mol Ecol 14: 1861–1868 doi:10.1111/j.1365-294X.2004.02430.x. doi: 10.1111/j.1365-294x.2005.02430.x