MARINE ECOLOGY PROGRESS SERIES Vol. 32: 307-310, 1986 - Published September 18 Mar. Ecol. Prog. Ser.

NOTE

Does giganteus protect host from predation by Acanthaster planci: predator pressure as a mechanism of coevolution?

L. M. DeVantier, R. E. Reichelt, & R. H. Bradbury

Australian Institute of Marine Science. PMB 3. Townsville MC. Queensland 4810, Australia

ABSTRACT A novel mutuahstlc relatlon has been observed Apart from such symbiotic relations, other predator- between serpulld worms Sp~robranchusyiganteus and mas- selected adaptations may also have evolved, while slve coral colon~esof Porites lutea and P lobata In response to some corals may be pre-adapted to survive predation. Acanthaster plancl predatlon on reefs of the central Great Barner Reef (GBR), Austraha As fe~rilngaggregatlons of A For example, prey preference by Acanthaster planci is planci have caused mass mortality of hard corals In these selective both among coral species and Individual col- areas, predator-lnduced selectlnn may be operating as an onles in relation to palatability, defensive mechanisms ecological mechanism In the development of the lnteractlon and availability (Barnes et al. 1970, Brauer et al. 1970, from the S giganteus cornmensalism Into a facultative mutu- allsm For the worms masslve Pontes colonies provide both Collins 1974, 1975a,b, Huxley 1976, Ormond et al. protect~on and an Ideal locat~on for suspension ieedlng 1976). While massive colonies of Porites spp. are less (Strathmann et a1 1984 Sm~thpers comm ) For the coral preferred prey (Barnes et al. 1970, Brauer et al. 1970, colony the worms provlde a refuge for adjacent polyps from Pearson 1973), A. planci do occasionally prey upon predatlon by A planci Such protected polyps ~n~tlateregen erahon ensurlng the survlval of the colony As Por~teslutea these species (Goreau 1964, Chesher 1969, Barnes et and P lobata have long colon~alllves (>500 yr) (Potts et a1 al. 1970, Clynn 1974, Done 1985). Indeed, the large 1985) the long term survlval of lnd~vldual colonies 1s an populations of starfish recently aggregated on reefs of Important factor In thelr hfe hlstory strategies and m maln- the central Great Barrier Reef (GBR) have preyed upon taln~ngthe observed broad dlstrlbutlon of slze classes of these many of these colonies (Cameron & Endean 1985, Done specles (Potts et a1 1985 Cameron & Endean 1985) In rela tlon to continued predahon by A planci this mutualism may 1985). be an important buffer actlng to preserve reef community A survey of a large number of Porites colonies on structure and accelerate the recovery process these reefs and one reef in Pulau Seribu (Indonesia) has demonstrated that such predation is often non- INTRODUCTION lethal to the colony. It has also indicated that a relation frequently exists between the distribution of certain Coral reef communities are renowned for their serpulid worms (Spirobranchus giganteus: taxonomic ecological complexity. This complex~tyis exemplified status currently under revision, R. Smith pers. comm.) by a high level of coevolution among interacting on these colonies and the remaining Living tissue. species. For instance, specific interactions between While working on Pontes in the Red Sea, Coreau predator and prey over evolutionary time scales have (1964) noted that many tubicolous worms survived led to symbiotic relations (Cameron 1983) that can death of the host coral by Acanthaster planci preda- ameliorate the predation effects. An example of this tion. He did not record the survival of any adjacent kind of interaction has been described by Glynn (1980) coral polyps. However, in our study area, the only in relation to hard coral predation by the crown-of- living polyps on Pontes colonies following predation thorns starflsh Acanthaster planci. He has shown that con~monlyoccur in patches hosting populations of ser- crustaceans living within pocdloporid coral colonies, phds, or directly below the worms' extended bran- and feedlng on mucus from the coral, can chemically chial crowns (3 to 5 cm diameter). On a broader geo- sense approaching A. planci and will attack their tube graphical scale the phenomenon is likely to be patchy, feet and spines, thus protecting the host coral. in relation to intensity of predation, although it has

O Inter-Research/Printed ~n F. R. Germany 308 Mar. Ecol. Prog. Ser. 32: 307-310, 1986

been observed elsewhere on the Great Barrier Reef (R. Thirdly, more than 2 yr after predahon, the areas of Smith pers. comm.). living polyps beneath the worms' branchlal crowns As a result of such avoidance of Spirobranchus showed evidence of regrowth. During this period, giganteus, a refuge is provided for the adjacent polyps these remnant polyps had grown above the surround- and a ring of protected tissue persists after predation. ing algal covered skeleton, and asexual budding was These rehct polyps could inltiate colony regeneration occurring in the outermost polyps. Monitoring of these and ensure survival of the colony. colonies will allow continued assessment of recovery. Porites polyps were not protected in all cases. Field observations indicate that these worms gener- METHODS AND RESULTS ally retreat into their tubes on initial contact with foreign organisms, reappearing after a short interval to To exarmne the significance of these observations, continue feeding. The length of time between retrac- field censuses were carried out at Potter Reef (17"42'S, tion and reappearance seems to vary among individual 146'33'E) in the central section of the GBR in April worms in relation to the type of stimulus, its intensity and June 1985. This reef, along with many others in the and frequency. Preliminary field experiments indicate region, supported large populations of Acanthaster that for certain worms, contact by Acanthaster planci planci throughout 1983-84. At the coral population induces retraction followed by almost immediate level, significant positive associations between the co- reappearance, with the operculum and branchial occurrence of the remaining living Porites lutea and P. crowns pushing against the tube feet and arms of the lobata colonies and Spirobranchus giganteus estab- starfish. This caused the predator to move quickly lished prior to the predation event were found away. However, other worms on the same Porites col- (Table 1). ony did not attempt to expose their branchlal crowns More specifically, the distributions of Spirobranchus during the period of observation, thereby allowing the giganteus and the living tissue on those colonies show starfish to remain in position over their tubes. strong positive association (Table 2). Whether these different behavioural responses are related to genotypic, or perhaps phenotypic, variations Table 1. Visual census data from Potter Reef, GBR (April within the Spirobranchus giganteus species complex, 1985). Contingency analysis (Kendall & Stuart 1979) of par- hally versus totally consumed Porites colonies shows a strong with only certain types capable of repelling Acanthas- positive relation between the presence of Spirobranchus ter (R. Smith pers. comm.), is currently unknown. giganteus and live colonies Further areas of research include ethology, chemistry and genetic studies of the coral-worm-starfish interac- State of Number of colonies with Total hon. predation or without (W/WO)worms of colony W WO DISCUSSION Partial 15 5 20 Total 0 11 11 It seems likely that those worms which attempt to Fisher Exact Test P < 0.001 expose their branchial crowns and opercular hooks upon coverage by Acanthaster planci sufficiently irri-

Table 2. Contingency analysis of randomly located, paired, 25 cm X 25 cm quadrats on partially consumed Porites colonies. For 5 of the 7 colonies there was a significant positive association between the presence of Spirobranchus giganteus and living tissue. a: S. giganteus and dead coral; b: no S. giganteus and dead coral; c: S. giganteus and live coral; d: no S. giganteus and live coral; P = Fisher Exact Probability (Kendall & Stuart 1979; or G-Test (Sokal & Rohlf 1969) for Totals

Colony Frequencies Total P Colony dimensions number a b C d number of Circumference Height Live cover quadrats (m) (m) ("10) DeVantier et al.: Does Spirobranchus giganteus protect host Porites? 309

tate the predator to induce the starfish to move away. CONCLUSIONS That irritation of the starfish's tube feet or everted stomach discourages feeding is well documented (Bar- Boucher et al. (1982) have considered the wider nes et al. 1970, Glynn 1980). However, the level of implications of the evolution of mutualism: 'In cases response by the starfish may be mediated by its nutri- starting as symbiotic commensalisms, the evolution of tional and physiological state, the amount of food mutualism may proceed by the commensal's providing available, and the population densities of predator and some benefit that would be selected for if it increased worm. the host's chances of survival . . .a change in ecologi- The long term consequence is the continued growth cal circumstances or the presence of an ecological of those polyps protected by Spirobranchus giganteus opportunity could transform the relationship into a and the survival of the colony. The long term survival mutualism if the commensal happens to render the of the same colonies is an important strategy in the life host better able to survive or take advantage of the histories of these corals. Indeed, one colony has been situation'. dated at over 600 yr (Potts et al. 1985). Such large, old Whether polyp protection is a specialized response colonies, comprising perhaps dons of polyps, are to Acanthaster planci predation or other selective potentially highly fecund, producing large numbers of pressures over coevolutionary time, or merely an inad- gametes annually (Kojis & Quinn 1981, Haniott 1983). vertent by-product of the Acanthaster-Spirobranchus Further, under certain conditions such old corals may interaction, remains to be seen. That the repellent produce new daughter colonies by asexual fragmenta- effect may be limited to certain types within the tion (Highsmith 1980), thereby perpetuating the parent Spirobranchus species complex suggests that preda- genotype (Potts et al. 1985). Cameron & Endean (1985) tion-induced selection may be currently taking place. have proposed that 'complex coral reef communities If large-scale outbreaks of A. planci are a relatively are based on (such) long lived corals which persist as recent phenomenon, then it is conceivable that we may individual colonies at specific sites for many years'. be witnessing the initial stages in the coevolutionary Certainly, with regard to predation by Acanthaster development of this mutualism. plancj, the survival of these colonies by regrowth is important in the recovery process, allowing eventual Acknowledgements We are grateful to Terry Done, Peter restochng of decimated areas (Glynn 1976) with Moran, Peter Isdale and Richard Smith for stimulating discus- sons, the Comn~onwealthCommunity Employment Program planulae, or by fragmentation. Partial predation by A. funded Crown-of-Thorns starfish study for logistlc support, planci may enhance the latter process by isolating and 2 reviewers for construct~vecriticism. This is contribution patches of living polyps beneath S. giganteus and in No. 328 from the Australian Institute of Manne Science. crevices etc. During regeneration these polyps may form separate colonies which become increasingly subject to dislodgement from the parent colony during LITERATURE CITED storm events. Barnes, D. J., Brauer, R. W., Jordan, M. R. (1970). Locomotory The continued upward growth of protected polyps response of Acanthaster planci to various species of coral. and Spirobranchus giganteus would ensure the worms Nature (Lond.) 228: 342-344 remain in advantageous positions for feeding, above Boucher, D. H., James, M,, Keeler, K. H. (1982). The ecology the boundary layers of lower strata benthos (Strath- of mutuahsm. Ann. Rev. Ecol. Syst. 13: 315-347 Brauer, R. W., Jordan, M. R., Barnes, D. J. (1970). Triggering mann et al. 1984). In contrast, when host Porites col- of the stomach eversion reflex of Acanthaster planc~by onies die, settlement by fouling species and growth of coral extracts. Nature (Lond.) 228: 344-346 the surrounding organisms may tend to block the Bryan, P. G. (1973). Growth rate, toxicity, and distribution of essential water currents or lead to con~petitiveinterac- the encrusting sponge Terpios sp. (Hadromerida: Sub- eritidae) in Guam, Mariana Islands. Micronesica 9: tions for space. It seems reasonable that worms living 237-242 in growing Porites colonies may be selectively Cameron, A. M. (1983). Toxicity in coral reef organisms. favoured. Indeed, recruiting S. gjganteus larvae Toncon. Suppl. 3: 105-109 actively select living Pontes colonies as settlement Cameron, A. M., Endean, R. (1985). Do long-lived species sites (Smith 1985). Over evolutionary time, mechan- structure coral reef ecosystems? Proc. 5th Int. Coral Reef Symp. 6: 211-215 isms which assist the suMval of the adjacent polyps Chesher, R. H. (1969). Destruction of Pacific corals by the sea- may have developed. Bryan (1973) noted that spirorbid star Acanthaster planci. Science 165: 28C283 worms residing on a Porites lutea colony in Guam were Cobs, A. R. S. (1974). Biochemical investigations of two not encrusted by the sponge Terpios sp., possibly due responses involved in the feeding behaviour of Acanthas- terplanci (L.). 1. Assay methods and preliminary results. J. to their mechanical movements. Considering the exp. mar. Biol. Ecol. 15: 173-184 above, it appears possible that polyp refuges may sirni- Cohns, A. R. S. (1975a). Biochemical investigahons of two larly evolve in this and other overgrowth interactions. responses involved in the feeding behaviour of Acanthas- 310 Mar. Ecol. Prog. Ser. 32: 307-310, 1986

ter planci (L.). 2 Isolation and characterization of KendaU, M,, Stuart, A (1979). The advanced theory of statis- biochemical stimuli J. exp. mar Biol. Ecol. 17: 69-86 tics 2, Inference and relationsh~p. Charles Gnffln and Co., Collins, A. R. S. (197513). Biochemical investigations of two London responses involved in the feeding behaviour of Acanthas- Kojis, B. L.. Quinn. N. J. (1981). Reproductive strategies in ter planci (L.).3. Food preferences. J. exp. mar. Biol. Ecol. four species of Porites (Scleractinia). Prov. IVth Int. Coral 17: 87-94 Reef Symp. 2: 145-151 Done, T. J. (1985). Effects of two Acanthaster outbreaks on Ormond, R. F. G., Hanscomb, N. J., Beach, D. H. (1976). Food coral community structure - the meaning of devastation. selection and learning in Acanthaster planci. Marine Proc. Vth Int. Coral Reef Symp. 5: 315-320 Behav. Physiol. 4: 93-105 Glynn, P. W. (1974). Acanthaster: Effect on coral reef growth Pearson. R. G. (1973). Recovery of coral reefs on the Great in Panama. Science 180. 504-506 Barrier Reef following attack by Acanthaster. Symposium Glynn, P. W. (1976). Some physical and biological determin- on the Biology and Ecology of the Crown of Thorns Star- ants of coral community structure in the eastern Pacific. fish, Acanthaster planci (L.). Paclfic Science Association Ecol. Monogr. 46: 431456 Second Inter-Congress. University of Guam. Micronesia 9 Glynn, P. W. (1980). Defence by symbiotic crustacea of host (2): 223 corals elicited by chemical cues from the predator. Potts, D. C., Done, T. J., Isdale, P. J., Fisk, D. A. (1985). Oecologia (Berl.) 47: 287-290 Dominance of a coral community by the genus Pontes Goreau, T. F. (1964). On the predahon of coral by the spiny (Scleractinia). Mar. Ecol. Prog. Ser. 23: 79-84 starfish Acanthaster planci (L.) in the southern Red Sea. Smith, R. (1985). Development and settling of Spirobranchus Sea Fish. Res. Stn. Haifa Bull. 35: 23-26 giganteus (Polychaeta; ). Proc. 1st Int. Harriott, V. J. (1983). Reproductive ecology of four scleracti- Conference, Sydney: 461-483 nian specles at Lzard Island, Great Barrier Reef. Coral Sokal, R. R., Rohlf, F. J. (1969). Biometry. W. H. Freeman and Reefs 2: 9-18 Co., San Francisco Highsmith, R. C. (1980). Passive colonization and asexual Strathmann, R. R., Cameron, R. A., Strathmann, M. F. (1984). colony multiplication in the massive coral Porites lutea Spirobranchus giganteus (Pallus) breaks a rule for suspen- Milne Edwards and Hame. J, exp. mar. Biol. Ecol. 47. sion-feeders. J. exp. mar. Biol. Ecol. 79: 24L249 55-67 Huxley, C. J. (1976). Response of Acanthaster planci (L.) to Accepted for pnnting on June 11, 1986 partial stimuli. J. exp. mar. Biol. Ecol. 22: 199-206