Feeding Preferences of Acanthaster planci (Echinodermata: Asteroidea) under Controlled Conditions of Food Availability

Feeding preferences of the crown-of-thorns sea star, Acanthaster planci (L.), were studied in a series of laboratory-based feeding trials wherein sea stars were provided with equal availability of six different coral species. The order in which corals were consumed was then used to ascertain feeding preferences. Crown-of-thorns sea stars exhibited strong and consistent feeding preferences across replicate feeding trials. The most readily eaten coral species was Acropora hyacinthus, followed by A. gemmifera, A. nasuta, A. formosa, Stylophora pistillata, Montipora undata, and Pocillopora damicornis. Crown-of-thorns sea stars also consumed Goniopora lobata, Fungia fungites, Goniastrea retiformes, and Pavona cactus but only after all Acropora and Montipora (Family Acroporidae) as well as Pocillopora and Stylophora (Family Pocilloporidae) were eaten. The leastpreferred corals were Favites abidita, Porites lobata, Symphyllia recta, Echinopora horrida, and Porites cylindrica. Of these, P. cylindrica was never eaten in any of the feeding trials in which it was offered. Observed feeding preferences substantiate findings from previous studies, where corals from the families Acroporidae and Pocilloporidae were preferred over all other corals. Further research is required to assess the underlying basis of feeding preferences of A. planci, but this study confirms that these starfish readily distinguish between different corals and have innate preferences for certain species. Still, most corals were eventually consumed, showing that when food is limited (during population outbreaks) A. planci is likely to consume virtually all different coral species, causing extreme devastation to coral reef ecosystems. The coral-feeding crown-of-thorns sea star, Acanthaster planci (L.), is renowned for its capacity to cause large-scale devastation on tropical coral reefs. Mostly, A. planci occurs at very low densities (typically <1 starfish ha ) and has limited effect on coral reef ecosystems (Dana et al. 1972, Glynn 1973). However, at very high densities, during outbreaks, A. planci can kill up to 80% of corals across large reef areas (Chesher 1969, Pearson and Endean 1969). Even at moderate densities A. planci has the potential to greatly modify coral community structure by selectively feeding on different coral species. In the eastern Pacific, for example, Glynn (1976) found that A. planci reduces coral diversity and increases dominance of Pocillopora damicornis by selectively feeding on rare coral species, such as Pavona varians (see also Branham et al. 1971, Glynn 1987, Chess et al. 1997). In contrast, in the western and central Pacific and in the Indian Ocean, A. planci tends to prefer corals of the genus Acropora, which are highly abundant and often dominate coral communities in those regions (Potts 1981, Moran 1986, Birkeland and Lucas 1990). By feeding selectively on highly abundant corals, A. planci may facilitate growth and recruitment of other less-common coral species and effectively increase coral diversity (Porter 1972). Pacific Science (2007), vol. 61, no. 1:113–120 : 2007 by University of Hawai‘i Press All rights reserved 1 This research was funded by the Great Barrier Reef Marine Park Authority. Manuscript accepted 13 April 2006. 2 ARC Centre of Excellence for Coral Reef Studies, James Cook University, Townsville, Q4811, Australia (phone: 61 7 4781 5747; fax: 61 7 4725 1570; e-mail: [email protected]). Acanthaster planci sea stars are clearly very selective in their choice of prey (Glynn 1974, 1987, Ormond et al. 1976, De’ath and Moran 1998, Pratchett 2001), and their feeding preferences are fundamental in determining their effects on coral communities. Despite this, feeding preferences of A. planci, as well as the factors that influence these preferences, remain poorly understood. The choice of diets for all animals is widely assumed to follow principles of optimal foraging theory, whereby foraging activities adopted by animals are those that maximize the net rate of energy intake (Ormond et al. 1976, Hughes 1980). However, there are many other factors, such as the distribution and abundance of prey types as well as interspecific competition, that can constrain the range of prey types an animal may consume (e.g., Berumen et al. 2005). Consequently, it can be very difficult to establish which factors influence feeding preferences, even for the most highly selective feeders (Hughes 1980). For A. planci, many factors have been proposed to influence feeding preferences, including the nutritional content of corals, coral growth form, coral defenses (e.g., mesenterial filaments, nematocysts, and secondary metabolites), host defense by crustacean symbionts, the distribution and abundance of corals, and prior conditioning and learned behavior of the sea stars (reviewed by Moran 1986, Birkeland and Lucas 1990). In the field, A. planci tends to exhibit a well-defined hierarchy of feeding preferences, with Acropora being the most preferred genus of corals and Porites the least preferred (Keesing 1990, 1992, De’ath and Moran 1998). These strong and consistent feeding preferences are partly attributable to variation in the energy and protein content among different corals (Keesing 1990). Keesing (1990) also showed that A. planci feeds much more efficiently on Acropora spp., compared with Porites spp. However, observed feeding preferences of A. planci are not consistent with differences in nutritional content or handling times for other coral genera (Keesing 1990). As a consequence, various constraints have been put forward to account for departures from optimal foraging theory and explain why A. planci does not consistently select the most nutritious prey corals. Potts (1981) suggested that coral prey that are readily consumed by A. planci are species that are least avoided, rather than those that are most preferred. Defensive mechanisms of scleractinian corals, including nematocysts, mesenterial filaments, secondary metabolites, and the antagonistic behavior of crustacean guards, may all deter starfish from feeding on certain corals (Potts 1981). In the bestdocumented example, Glynn (1974, 1976, 1987) and Pratchett (2001) showed that crustacean guards within some coral species attack A. planci and deter them from feeding on their host corals. Still, both Pocillopora and Stylophora, which contain highly antagonistic crustacean guards, are among the most highly preferred coral prey of A. planci on the Great Barrier Reef (e.g., Keesing 1990, De’ath and Moran 1998), suggesting that these crustacean guards have only limited influence on the overall feeding preferences of the sea stars. The specific role and relative importance of different factors in determining feeding preferences of A. planci have not been tested. Also, much of the current information about feeding preferences of A. planci is qualitative, rather than quantitative (reviewed by Potts 1981, Moran 1986). Further, most field-based studies infer feeding preferences from measures of dietary electivity, which are highly confounded by differences in the size, abundance, and accessibility of different corals (De’ath and Moran 1998). The purpose of this study was to examine feeding preferences of A. planci under laboratory conditions, providing the sea stars with equal availability of alternate prey corals in aquariums. Such studies are essential to test whether A. planci has definitive feeding preferences independent of variation in the size or abundance of different coral species, as suggested by Moran (1986). Previous studies that have explored feeding preferences of A. planci in the laboratory (Brauer et al. 1970, Ormond et al. 1976, Keesing 1990, Sonoda and Paul 1993) have been very limited in their extent or scope. Brauer et al. (1970), for example, examined behavioral responses of A. planci (stomach eversion versus withdrawal) 114 PACIFIC SCIENCE . January 2007