The role of corallivory on coral reefs is becoming increasingly important to coral reef ecology given the number of other stressors coral reefs worldwide currently face [1]. Invertebrates are the majority of corallivores, outnumbering their fish counterparts nearly 3 to 1 [2], [3]. However, most invertebrate species have long been overlooked due to their small size and cryptic nature [4]. Corallivorous invertebrates may play an important role in coral health, inflicting minor or lethal damage on their coral hosts, which may subsequently have deleterious effects on coral growth and fitness [1]. They have also been implicated in transmitting or increasing vulnerability to coral disease [5], which indirectly contributes to coral loss or shifts in community composition. As scleractinian corals are the major reef builders, more attention is required to identify their predators and determine the roles they might play in maintaining or conserving coral reef ecosystems.

Two species of polyclad flatworms are known to prey on scleractinian corals [6], [7], yet very little is known about their impacts on coral reefs. As they are small and difficult to detect due to their excellent camouflage against the coral host, they may have been overlooked thus far in most studies of coral-associated animals. One such cryptic polyclad, the Acropora-eating flatworm (commonly known as the AEFW), was recently identified and classified as Amakusaplana acroporae Rawlinson et al., 2011 [7]. Known only from aquaria as a notorious pest of Acropora coral, this species has never been found in the wild. In fact, the taxonomic assignment was based on multiple specimens collected from two aquaria in the United States. Although most small animals that live and feed on corals have negligible, if any, ill effects on the coral host [8], infestations of A. acroporae on acroporids in captivity can result in rapid and complete colony death [9]. A. acroporae is a destructive predator of at least nine aquarium-reared Indo-Pacific acroporids (Acropora valida, A. pulchra, A. millepora, A. tortuosa, A. nana, A. tenuis, A. formosa, A. echinata and A. yongei), individuals lay multiple egg batches on an Acropora host and the hatchlings have a low dispersal capability [7]. These life history characteristics, combined with high prey specificity to Acropora, lend this species the potential to be a significant corallivore of Acropora corals.

Corallivory on Acropora corals is of particular interest to conservation management as Acropora is one of the most ecologically important coral genera to coral reefs worldwide. It is the largest extant coral genus, occurring in all tropical oceans as the dominant reef building coral [10]. Acroporids are a source of critical habitat and food for an immense diversity (∼150 species) of coral-associated animals [3], [11], [12]. They are extremely abundant and fast-growing branching corals yet are among the most susceptible corals to bleaching [13] and disease [14]. Furthermore, many corallivores actively select species of Acropora as their preferred prey [1], [2], such as the crown-of-thorns sea star, Acanthaster planci [15] and the gastropod Drupella conus [16]. Acroporids are also commercially important, being among the top three genera collected for the aquarium trade [17]. Thus acroporids are often the focus of conservation efforts, such as reef restoration [18], and an understanding of what biotic and abiotic interactions affect the growth, survival and distribution of acroporid corals is critical to their effective conservation.

Given Amakusaplana acroporae’s preference for Indo-Pacific Acropora species it is assumed that the worm is endemic to that region. Its cryptic coloration and relatively small size would make it difficult to detect in situ, hence its easy introduction into aquaria as Acropora epifauna. Locating A. acroporae in its natural environment would permit further study of its biology and ecological interactions, and this, in turn, could lead to the discovery of effective biological controls for this corallivore in captivity. This study aimed to determine whether an as-yet unidentified polyclad flatworm found on Acropora valida colonies from Lizard Island, Australia, was Amakusaplana acroporae.

This study identifies a polyclad flatworm found on Acropora valida colonies around Lizard Island as Amakusaplana acroporae and represents the first report of this animal in the wild. Evidence that A. acroporae is a corallivore in its natural habitat, as it is in aquaria, is supported by the presence of Symbiodinium and cnidarian spirocysts in the gut and parenchyma. In addition, the extracellular distribution of Symbiodinium implies that they were ingested and are not symbionts living within A. acroporae. Discovering A. acroporae in its natural environment and documenting a method of extracting the animals from their coral host alive will aid further research into the abundance, distribution and ecology of this corallivore.

Polyclad flatworms are morphologically quite homogeneous and over the past two centuries species descriptions and classifications have been based on a limited number of taxonomic characters [20], [22], [23]. These characters, used at all taxonomic levels, are described from the animal’s gross morphology and anatomy; for example the presence of a ventral sucker, the type and position of the pharynx, the presence of tentacles, details of the reproductive system and patterns of eyes. However, these last two sources of taxonomic characters, which are important for species level identification within the Prosthiostomidae, demonstrate plasticity during maturation as observed in this study and Kato [24] (in Prosthiostomum (L.) purum). Therefore, without access to Amakusaplana ohshimai material for comparative morphological and molecular analysis we cannot rule out the possibility that Amakusaplana acroporae is synonymous with A. ohshimai, given that the characters that distinguish the two species (eye arrangement and number, morphometrics of the male and female reproductive systems and the presence of an egg chamber) vary with body length and maturation. This highlights the need to include in future species descriptions changes in morphological characters during development, and within and between populations, where possible. Nevertheless, from our morphological and molecular diagnoses we are confident that the polyclads collected from Lizard Island are the same species as that described from aquaria [7].

The presence of Symbiodinium and cnidarian spirocysts in the gut and parenchyma provides evidence that Amakusaplana acroporae is a corallivore in its natural habitat. No other prey items were observed in the gut of A. acroporae indicating that perhaps they are obligate corallivores (as has been demonstrated in the only other known scleractinian-eating polyclad Prosthiostomum (Prosthiostomum) montiporae [25]). As spirocyst morphology is fairly homogeneous within the Anthozoa [26] more direct evidence that A. acropora is feeding on A. valida would involve comparisons of molecular fingerprints of coral tissue in the gut contents with tissue from the coral host. Unlike some polyclad species that sequester nematocysts from their cnidarian prey in the lateral and posterior margins of their dorsal epidermis [27]–[29], there was no evidence of spirocysts being sequestered in A. acroporae in this study. While some corallivores have morphological adaptations that provide them with protections from coral nematocysts [30], how A. acroporae overcomes Acropora nematocysts is unknown.

As Amakusaplana acroporae is quite small, cryptic and possesses excellent camouflage against its acroporid coral host, this species is easy to overlook and thus far, their corallivory in the wild has probably been attributed to another species or even coral disease [30]. Moreover, until now this species has been unknown to marine ecologists, hampering any potential to learn about its role in coral health. Although Sweet et al. [31] reported that previous studies had found A. acroporae (or AEFW, as it would have been known at the time) in Indonesia and the Red Sea (citing [32]–[34]), Haapkylä et al. [33] actually refer to the acoel worm from the genus Waminoa which is a known coral-associate [35], and the other two studies do not mention flatworms. Although it is highly likely that the distribution of A. acroporae mirrors that of its Acropora species prey, and it could therefore be found in Indonesia and the Red Sea, visual surveying methods alone would probably not be sufficient to see A. acroporae in situ (Stella pers obs), although bite marks in the coral tissue and egg capsules on the bare coral skeleton might be visible on a heavily infested colony. As A. acroporae has been found in association with other Indo-Pacific Acropora species in aquaria (A. pulchra, A. millepora, A. tortuosa, A. nana, A. tenuis, A. formosa, A. echinata and A. yongei [7]), it is possible these species would be suitable natural hosts as well and might serve as a logical basis for learning more about these animals under natural conditions.

Gaining knowledge of the natural rates of occurrence and ecology of these polyclad worms will be vital to understanding its ecological role on coral reefs. No obvious tissue damage was evident on the Acropora valida colonies sampled in this study. That may, in part, be due to the small abundances (averaging less than three worms per colony) or the presence of natural predators within the coral colony. The estimates of abundance per colony in this study are somewhat underrepresented given that our sample size was small and the method was biased towards individuals greater than 1 mm². Amakusaplana acroporae hatchling size is 250–300 µm [7] and these juvenile stages would have escaped collection. In aquaria some wrasse species have been observed to eat dislodged adult worms in the water column [7], [9]. Embryonic and hatchling life history stages may be vulnerable to a different set of predators, such as gastropods and decapods, which are highly diverse on acroporid corals [12]. Coral crabs, belonging to the genus Tetralia, have high occurrence rates on tightly branching acroporids [11] and are known to provide the coral host with cleaning services [36]. It is possible that these crabs may eat the adult worms and egg capsules, thus controlling the worms’ numbers. It is also possible that A. acroporae only becomes a serious pest in disturbed coral systems and aquarium environments, as is the case with P. (P.) montiporae [9], [25]. Further observations of A. acroporae in the field are needed to determine rates of coral tissue consumption (and subsequent colony mortality), identify its natural predators and quantify spatio-temporal patterns in its abundance.

Scleractinian corals are the most functionally important corals to reef processes, thus it is essential to understand what factors affect their growth and survival. Corallivores represent a biotic stressor that can detrimentally affect coral growth and fitness. In order to effectively manage conservation efforts of Acropora on coral reefs and to successfully rear colonies in aquaria, it is critically important to understand what biotic interactions are important to coral growth and fitness. This discovery of Amakusaplana acroporae in the wild and at Lizard Island will facilitate easy access to populations of this coral symbiont, enabling investigation of A. acroporae ecology, biology and life history in its natural habitat.