<I>Coralliophila Abbreviata:</I> a Significant Corallivore!

BULLETINOFMARINESCIENCE.32(2): 595-599, 1982 CORALREEFPAPER

CORALLIOPHILA ABBREVIATA: A SIGNIFICANT CORALLIVORE!

Susan H. Brawley and Walter H. Adey

ABSTRACT Predation by Coralliophila abbreviata on Acropora palmata was observed in the field and in a coral reef microcosm. The mollusks removed as much as 16 cm2 of coral tissue/dayl animal. The damage was irreversible, and algal colonization quickly occurred over the skeleton in areas stripped of living tissue. We describe aspects of Coralliophila behavior during feeding and following displacement experiments. The extent of Coralliophila predation on corals appears to be limited on some reefs by fish predation, but these mollusks are significant corallivores when predation pressure is low. Our observations suggest that feeding by Coralliophila can also attract other coraUivores (butterflyfish) to a colony, accelerating its death.

Scleractinian coral predators include some species of fishes, crustaceans, mollusks, echinoderms and polychaetes (Robertson, 1970). The importance of several of these predators has been quantitatively assessed (Glynn et al., 1972; Glynn et al., 1979) and found to have major impact on reef growth. The importance of the molluscan family Coralliophilidae, especially of Coralliophila spp., as significant coral predators, remains controversial. Ward (1965) found coral tissue in the digestive tract and concluded that these mollusks were partly responsible for damage to colonies of Montastrea annularis, the coral with which they were primarily associated in Barbados. Data on feeding rates were not presented, however, and there is no evidence of recently removed tissue in the documentary photograph in Ward's paper. Other reports (Ott and Lewis, 1973) cite the association of Coralliophila with the living margin of scleractinian corals (Montastrea annularis, Acropora cervicornis, A. palmata, Siderastrea siderea, Diploria clivosa, D. stri- gosa, Agaricia agaricites, Mycetophyllia lamarckiana) but little, if any, damage to the coral was noted in the field. Since feeding activity appeared to be minimal and the mollusks seemed sedentary, Ott and Lewis (1973) stated that the mollusks could not be primarily responsible for the large dead areas of coral found near Coralliophila. These mollusks lack a radula; the coral tissue is stripped through enzymatic breakdown and proboscal pumping (Ward, 1965). The coral skeleton is undamaged. To our knowledge, a detailed analysis of the diet of these mollusks has not been made, but previous work (Ward, 1965) indicates an exclusive con·· sumption of coral tissue. We report here on predation by C. abbreviata on A. palmata from Haitian reefs and from a coral reef microcosm.

METHODS

Field observations were made on Acropora palmata colonies growing on the fore reef (6-10 m) of a barrier reef located at 200N, now on the north coast of Haiti. These studies were made using SCUBA during the June, 1980, cruise of the RN MARSYSRESOLUTE. The field observations were compared with coral predation in a 12-kl coral reef microcosm at the Smithsonian Institution (Washington, D.C.). Mollusks and corals in the microcosm tank were collected from Eleuthera (the Bahamas). The coral reef microcosm contains over 150 species of plants and animals, including approximately 15 species of fish (mostly herbivores, including 2 spp. of po- macentrids and 3 spp. of scarids), 30 algal species, 12 species of scleractinian corals and a wide variety of other invertebrates (sea urchins, crabs, sponges, mollusks, anemones, polychaetes and smaller invertebrates). No known predators of Coralliophila are included in the microcosm community. Natural lighting is simulated with high intensity metal halide lights, and waves are generated

595 5% BULLETIN OF MARINE SCIENCE. VOL. 32, NO.2, 1982

Figure 1. Ventral view of one of the C. abbreviata removed from a colony of A. palmata in the microcosm tank. This animal was marked and used in displacement experiments. x 1.4 (photograph by C. Clark). Figure 2. Coralliophila abbreviata (small arrows) feeding on a small colony of Acropora palmala in the coral reef microcosm at the Smithsonian Institution. Areas within the dotted line are in various stages of algal colonization following mollusk predation in the previous week. The large arrow indicates an area in which a mollusk has just eaten the coral tissue. x 1.0 (photograph by C. Clark). Figure 3. Same colony as in Fig. 2. Dotted line encloses area of skeleton where predation by C. abbreviata occurred. xO.4 (photograph by C. Clark).

with dump buckets so that some surge is experienced by reef organisms. During these experiments, the light cycle was 15:9 L:D, temperature ranged from 26-29°C, and salinity varied from 36-37"%0. See Adey (In Press) for a detailed description of a similar microcosm. Feeding rates were assessed by daily observations of the surface area of tissue removed over the previous 24 h. Directional movements of snails over the coral were also noted. Two snails were marked with finger nail polish and displaced varying distances (0.3-1.3 m) from their feeding sites to areas of old reef substrate. The movements and feeding behavior of these animals were noted over succeeding days. These animals were 2.0 and 2.3 cm in length (Fig. 1).

RESULTS During several dives on the Haitian fore reef near the base of the A. palmata zone, we observed predation on A. palmata colonies by C. abbreviata. Large branches commonly had as many as 30 animals clustered at the margin of living polyps and dead corallites. Since algal colonization occurs within days when a bare carbonate surface becomes available, the 4-cm long strip of bare skeleton common around the branches represented predation over no more than the previous 2 to 3 days. Successive stages of algal colonization, which appeared to be earlier feeding areas, were observed below this band for at least 0.3 m. C. ab- BRAWLEY AND ADEY: CORAL REEF PREDATOR 597 breviata was also observed feeding on other scleractinians, including M. annularis, but the greatest damage appeared to be to colonies of A. palmata. We made seven dives on the Haitian fore reef. The damage we observed in the shallower areas of the fore reef was not apparent at greater depth (13-30 m), but an extensive search was not undertaken, and quantitative studies might reveal consider- able predation at these depths as well as within the shallow zones. In our microcosm, two species of Coralliaphila, C. caribaea and C. abbreviata, have fed on A. palmata; much greater damage occurred following predation by C. abbreviata. Most of the predation occurred at night. One colony (Figs. 2 and 3) of A. palmata lost 195 cm2 of surface tissue from feeding by three animals over 4 days, i.e., an average feeding rate of about 16 cm2 tissue/animal/day. Predation was too rapid to allow regrowth of coral tissue into the grazed areas of the skeleton; algal colonization occurred in these areas within a few days. This colony was subsequently killed by predation from the mollusks and butterflyfish (C. striatus, C. ocellatus). Primary tissue loss was due to gastropod predation, and the butterflyfish fed during most of the observation period on residual tissue behind the feeding tracks of the gastropods; however, the butterflyfish began feeding steadily on the colony after about 60% of its tissue had been removed by the mollusks. C. abbreviata displaced from A. palmata colonies reappeared on the same or a different colony of A. palmata depending upon the location to which they wen: displaced. Typically, no movement occurred for at least 24 h following displace·· ment. This was true for four displacement trials of one marked animal, but only for the first two of five displacements of a second animal. Both of these animals moved over more than a meter of irregular surface on succeeding nights to reach a colony of A. palmata 1-3 days later. As they returned to A. palmata colonies, these animals passed many other species including M. annularis and M. caver·· nasa. Animals were moved within 3 cm of aM. annularis colony on two occa- sions. The second animal was more voracious than the first. It fed at an average rate of 8 cm2 tissue/day while the first animal fed about every third night during 2 weeks of observation, removing about 7 cm2 of tissue during such a feeding period . . Displacements were made to areas of the microcosm which eliminated the possibility that the animals were returning to a colony via a mucous-pheromonal trail secreted within the previous 2 weeks. Chemotactic response to the coral also seemed unlikely because straight-line paths of return to a colony were made both upstream and downstream of water flow in the microcosm. The shortest visible path of return was ordinarily used by the animal; however, the ability to return to a coral by an indirect route, when faced with highly interrupted substrate along a straight path, was remarkable. A cage of nylon screen (I-em mesh) placed around one colony of A. palmata had no effect upon return of the displaced Caralliophila to this colony.

DISCUSSION Our brief observations in Haiti do not allow us to comment on the extent of Coralliophila damage to corals across the reef or to compare it quantitatively to that caused there by other corallivores. The rate of tissue removal that we observed in the field and in the microcosm does indicate that Caralliophila is a potentially serious corallivore. Why are these mollusks important corallivores on some reefs but apparently not on others? Coralliophila is subject to predator control by fish, such as puf- ferfish (Glynn et aI., 1979), but predator control of mollusk population density 598 BULLETIN OF MARINE SCIENCE, VOL. 32, NO.2, 1982 does not appear to determine the extent to which Coralliophila is an important corallivore. Recall that Ott and Lewis (1973) reported little damage to corals from Coralliophila predation. Yet they found a mean density of 13 individuals/m2 of living coral on the Barbados reefs they studied, a much higher density than in the reef microcosm. Another possibility is that the average feeding rate of C. abbreviata varies throughout its range. Kitting (1975) suggested that mollusks, such as Coralliophila, might feed at low rates to prevent "tracking" by predators. Certainly, the large areas of freshly grazed Haitian coral were apparent from meters away and drew the diver's attention to the presence of Coralliophila. Reefs with higher fish populations than this Haitian reef might select a population of Coralliophila that fed at a low rate and grew slowly, thereby avoiding "tracking" and intensive fish predation. The tendency of these mollusks to feed at the margin of living tissue would be a protective strategy, because basal areas of each colony often have dead zones caused by a wide variety of factors, and a low feeding rate would allow algal colonization to disguise the presence of the mollusks. Arguing against this idea of population selection by fish predators is the feeding situation we observed in our microcosm. Mollusks and corals in our microcosm were collected from an area in the Bahamas where extensive feeding was not obvious and where general fish populations were high. The mollusks were com- pletely free from predation in our microcosm and became significant corallivores there. This argues for some level of recognition by the mollusks themselves of the density of fish predators, producing a corresponding adjustment in their feeding levels on coral tissue. Different predator levels do influence feeding behavior in the urchin Eucidaris thouarsii (Glynn et aI., 1979). In Panama, there is high predation by fish on these urchins and they are algal grazers, while in the Gala- pagos these urchins are subject to relatively little fish predation and are significant corallivores. Physical factors, such as high wave energy, could regulate the amount of predation by Coralliophila on shallow water corals, such as. A. palmata, but our observations in the field and microcosm provide no data to support this possibility. It seems as likely to us that high wave energy would facilitate feeding on coral by reducing fish predation on Coralliophila. Under the typical, moderate energy conditions of the Acropora palmata zone, it seems unlikely that wave action would reduce feeding by dislodging the mollusks or making exposed areas of a colony less accessible. Coralliophila predation was apparently responsible for the demise of many of the corals that had survived the August, 1980, hurricane at the Discovery Bay reef on the north shore of Jamaica (J. Lang, pers. comm.). We have noted that butterftyfish in the microcosm appear to eat more coral tissue when Coralliophila predation is present, and such facultative targeting of corals to an array of corallivores may be of significance in a situation such as that at Discovery Bay. Intensive coral predation appears to occur on most reefs, but the particular corallivores involved clearly vary significantly depending upon the predator com- position of the particular reef community. It would be interesting to study feeding rates on coral by C. abbreviata exposed to different predator densities.

ACKNOWLEDGMENTS

We thank J. Rosewater (Smithsonian Institution) for identification of these mollusks as C. ahhre- viata, Research support for the coral reef microcosm was provided by the Smithsonian Institution and the National Science Foundation (O.S.S. #78-06909). J. Johnson and T. Goertemiller provided technical assistance for this work. BRAWLEYANDADEY:CORALREEFPREDATOR 599

LITERATURE CITED

Adey, W. H. In Press. Reefs and microcosms: Evening the odds for field researchers. Bull. Mar. Sci. Glynn, P. W., R. H. Stewart, and J. E. McCosker. 1972. Pacific coral reefs of Panama: Structure, distribution and predators. Sonderdruck aus der Geologischen Rundschau 61: 483-519. ---, F. A. Oramas, C. A. Montaner, and J. B. Achurra. 1979. Speculation on potential effects of molluscan corallivore introductions across the Isthmus of Panama. (abst.). Assoc. 1st. Mar. Lab. Carib. Santo Domingo, p. 15. ---, G. M. Wellington, and C. Birkeland. 1979. Coral reef growth in the Galapagos: Limitation by sea urchins. Science (Wash., D.C.) 203: 47-49. Kitting, C. L. 1975. The impact of molluscs feeding on some West Indian gorgonians. Bull. Am. Mal. Union 41: 73. Ott, B., and J. B. Lewis. 1973. The importance of the gastropod Coralliophila abbreviata (Lamarck) and the polychaete Hermodice caruncu/ata (pallas) as coral reef predators. Can. J. Zoot. 50: 1651-1656. Robertson, R. 1970. Review of the predators and parasites of stony corals, with special reference to symbiotic prosobranch gastropods. Pac. Sci. 24: 43-54. Ward, J. 1965. The digestive tract and its relation to feeding habits in the stenoglossan pro sob ranch Coralliophila abbreviata (Lamarck). Can. J. Zoo I. 43: 447-464.

DATE ACCEPTED: March 25, 1981.

ADDRESS: Marine Systems Laboratory, W-3/O, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. PRESENTADDRESS:(S.H.B.) The Mary Ingraham Bunting Insti- tute, Radcliffe College, /0 Garden Street, Cambridge, Massachusetts 02/38.