MARINE ECOLOGY PROGRESS SERIES Vol. 124: 181-188,1995 , Published August 10 Mar Ecol Prog Ser
Allogeneic and xenogeneic interactions in reef-building corals may induce tissue growth without calcification
Uri Frank1,*,Itzchak Brickner2.4, Baruch Rinkevichl, Yossi Loya2, Rolf P. M. ~ak~, Yair Achituv4, Micha 11an2
'The National Institute of Oceanography, Israel Oceanographic and Limnological Research, PO Box 8030. Haifa 31080. Israel 2Department of Zoology, The George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel-Aviv 69978, Israel 3Netherlands Institute for Sea Research, PO Box 59, 1790 AB Den Burg, Texel, The Netherlands; and Institute of Systematics and Population Biology, University of Amsterdam, Mauritskade 61, PO Box 94766, 1090 GT Amsterdam, The Netherlands 4Faculty of Life Sciences, Bar Ilan University, Ramat Gan 52900, Israel
ABSTRACT Tissue growth w~thoutthe depos~t~onof calcium carbonate skeletons was recorded in 2 Red Sea hermatyp~ccnldarians dunng competitive interactions Tissue contacts between allogeneic colonles of the hydrocoral Millepola dichotorna resulted in unilateral overgrowth In 39% of the assays the overgrowing tissue d~dnot secrete a skeleton for up to 10 wk These tissues were loosely attached to the overgrown branch, and rapidly advanced by up to 25 mm ~v~th~nthe f~rst 2 wk Thereafter, tissue growth slowed down or stopped and calcium carbonate was depos~tedover the subordinate branch, starting at the original contact area In xenogenelc interactions between the scleract~nlancoral Cyphas- trea chalc~dicurnand the c~rnpedbarnacle Savignium dentatum, tlssues of the coral always overla~d the plates of the barnacle without depos~tingcalc~um carbonate as long as the barnacles were ahve (up to 5 yr) Calcium carbonate was depos~tedby the coral s tlssue on the barnacle's plates only following barnacle death In both cases, the non-calc~fy~ngovergrowing t~ssueslacked polyps but appeared nor- mal in histological sections and contalned typlcal cnldarian cells and endosymbiotic zooxanthellae Thls type of tissue growth without calc~ficationis a newly descnbed allogeneic/xenogeneic response ellcited by hermatypic cnidanans
KEYWORDS Coral . Cirnpedla . Growth Calclf~catlon. Competit~veInteractions Histocompat~bil~ty
INTRODUCTION Jokiel et al. 1978, Davies 1989) and incorporation of the radioactive isotope 14C (Rinkevich & Loya 1984), are Hermatypic cnidarians are the most important related to the deposition of skeletons rather than to the framework builders in modern tropical reef communi- growth of tissues. Tissue growth without calcification ties (Fagerstrom 1987). Tissue growth in this group of in hermatypic cnidarians or free tissue movements organisms is considered to be closely associated with above already deposited skeletons have only rarely calcification, the deposition of aragonitic calcium car- been documented, although Brown et al. (1994) bonate by the overlying calicoblastic, ectodermal layer observed periodical withdrawal of coral tissues into the (Barnes & Chalker 1990). Moreover, the most fre- calices as a response to air exposure during low tides. quently used methods to estimate coral growth rates, Hard substratum for settlement and growth is proba- such as alizarin red staining (Barnes 1970, Lamberts bly a limited resource for sessile reef invertebrates 1978), buoyant weight measurements (Bak 1973, such as hermatypic corals, which are known to com- pete for space during their growth (Jackson & Buss 1975, Sheppard 1985, Lang & Chornesky 1990). This 'E mail: [email protected]~on.ac.il competition is expressed as a complex array of antago-
O Inter-Research 1995 Resale of full article not pej-mtted 182 Mar Ecol Prog Ser
nistic responses against conspecifics, as well as against MATERIALS AND METHODS unrelated organisms. During these interactions, vari- ous competitive processes, some of which are still All field work was carried out by SCUBA diving in poorly understood, occur on the morphological, cellu- the shallow-water (3 to 6 m) coral reef adjacent to the lar and biochemical levels (e.g. nematocyst discharge, H. Steinitz Marine Biology Laboratory at Eilat, Israel development of mesenterial filaments and sweeper (29" N, 35" E) on the Red Sea. The fringing reefs of the tentacles, allelopathic interactions, unidirectional trans- northern Gulf of Eilat are characterized by high coral location of photosynthates, unilateral and reciprocal coverage and species diversity (Loya 1972). overgrowth: Bak et al. 1982, Rinkevich & Loya 1983, Branches (ramets), about 10 cm long each, were Chornesky 1991, Salter-Cid & Bigger 1991, Chadwick- removed in the field from 9 large Millepora dichotoma Furman & Rinkevich 1994, Frank & Rinkevich 1994, colonies (identified as A to I) using wire cutters. The Rinkevich et al. 1994). Overgrowth during both allo- isolated ramets were secured to plastic clothespins that geneic and xenogeneic interactions is one of the most had been pre-glued to 5 X 20 cm Plexiglas plates. Ram- common and well-documented competitive mecha- ets from 6 colonies (A to F) were arranged in pairs, in 2 nisms. In calcareous cnidarians, such overgrowth has replicates each of all 15 possible pairwise combina- always been marked by simultaneous tissue formation tions. Ramets from 3 colonies (G to I) were paired in 8 and skeleton deposition as in the regular growth pat- replicates each of all 3 possible pair combinations. The tern (e.g. Potts 1976, Bak & Criens 1982, Rinkevich & tips of adjacent branches in each allogeneic pair were Loya 1983, Chornesky 1991, Chadwick-Furman & carefully put in gentle direct contact. This procedure Rinkevich 1994). minimizes injury to contacting tissues, as indicated by Tissues and skeletons of living hermatypic cnidari- in situ follow-up observations on allogeneic pairs, ans also serve as substrata for a variety of parasitic which displayed healthy tissue with extending polyps and commensal sessile invertebrates (Patton 1976, as in other branch tips. One isogeneic control assay Morton 1983). Several groups of invertebrates, such was set up for each of the 9 studied colonies. Interact- as serpulid tubeworms (Strathmann et al. 1984, ing branches were observed in situ every 10 to 14 d DeVantier et al. 1986, ten Hove 1989, Marsden et al. for up to 12 wk following the first tissue-to-tissue con- 1990, Frank & ten Hove 1992), cirriped crustaceans tact. A palr of branches was vitally stained with (Ross & Newman 1973), molluscs (Morton 1983) and alizarin red S dye in situ every 3 to 4 wk (10 to 15 mg sponges (Wulf & Buss 1979), are known to inhabit the 1-l, 24 h: Lamberts 1978). Plexiglas plates bearing the living surfaces of hydrocorals and madreporarians. pairs of interacting colonies were put into 30 1 trans- These epibionts exhibit a variety of mechanisms to parent plastic bags, into which 30 mg of dissolved avoid overgrowth by the coral's tissue and skeleton powdered dye was injected. For histological studies, (Hiro 1938, Gohar & Soliman 1963, Ross & Newrnan contacting pairs of branches were fixed in 2.5% glu- 1973, Morton & Scott 1980, Anderson 1992, Brickner taraldehyde in sea-water at room temperature for 3 to 1994). 4 h. Non-calcifying overgrowing tissues, which were Here we describe the phenomenon of tissue growth loosely attached to the overgrown branch, were then without calcium carbonate deposition in 2 Indo-Pacific removed using fine forceps, rinsed in double distilled hermatypic cnidarians, Millepora dichotoma (Hydro- water (DDW), dehydrated with increasing concentra- zoa) and Cyphastrea chalcidicum (Anthozoa), which tions of ethanol and embedded in glycol methacrylate develops following allogeneic and xenogeneic interac- plastic. Sections (1 to 2 pm) were prepared using glass tions, respectively. Allogeneic contacts between M. knives and stained with hematoxylin and eosin (Ban- dichotoma colonies are usually followed by rapid over- croft et al. 1990). Control sections were prepared from growth of one colony over the other (Frank & Rinke- naive M. dichotoma branches after decalcifying the vich 1994). In some cases, however, we observed allo- skeletons in a formic acid/sodium citrate solution geneic overgrowths which were not associated with (Rinkevich & Loya 1979). calcification. Xenogeneic interactions between the Ten colonies of the scleractinian coral Cyphastrea scleractinian coral C. chalcidicum and the barnacle chalcidicum infested by the pyrgomatide barnacle Savignium dentatum occur in about 84% of all coral Savignium dentatum were marked in situ with plastic colonies near Eilat, in the Red Sea (Brickner 1994). The tags along a 10 m transect and used for 2 experiments. barnacle bases are embedded within the coral skele- In the first, 5 coral colonies were stained with alizarin ton and their plates are covered by the coral's tissue up red S dye as described above. Immediately thereafter, to the rims of the barnacle aperture (Ross & Newman coral tissues were removed using a jet of tap water. 1973). This coral tissue covering the barnacle's plates Newly calcified zones were visualized under a stereo- does not deposit calcium carbonate as does normal microscope as pink-reddish areas (Lamberts 1978). tissue (Brickner 1994). Plates of 5 barnacles were taken from 2 of these Frank et al.: Coral growth without calcification 183
colonies in order to record coral calcification on the a few days following tissue-to-tissue contacts (Table 1). plates. Tissue residues were removed by soaking the Most (83.3 %) outcomes of replicate pairs were consis- plates in concentrated household bleach (sodium tent in terms of type of response and the directlonality hypochlorite) for several hours. Subsequently they of overgrowth. In 7 pair combinations (17 assays, were rinsed in DDW, mounted on aluminum stubs 38.9% of all combinations), the overgrowing tissues of using conductive glue, coated with gold palladium and 6 of the 9 studied colonies (all except colonies B, G and observed with a JEOL JSM 840 scanning electron H; Table 1) started to develop without any sign of a microscope (SEM). In the second experiment, 30 bar- simultaneous calcification process. In the field they nacles of various sizes from the other 5 coral colonies were recognized as loosely attached, delicate sheets of were killed without harming their shells, by inserting a tissue over the subordinate (= overgrown) branches sharp metal needle into their apertures. After 14 d the (Fig. l), which could be peeled off easily with a finger- alizarin red staining protocol was carried out on the 5 nail. These non-calcifying sheets of tissue advanced on coral colonies, and the barnacles were observed under the other colony's branch by up to 25 mm from the con- the stereomicroscope. SEM preparations were made tact area within the first 15 d (combination BD: Table from several plates of the previously killed barnacles. 1).This was 12 times faster than the most rapid calcify- Histological sections from C, chalcidicum tissues over- ing pair (2 mm; combination CF: Table 1). After 2 wk, growing the plates of live barnacles, and from regular the average extent of growth of non-calcifying tissues coral tissue, were prepared following decalcification as (13.4 * 8.2 mm) was significantly higher than that of described. normal calcifying tissues (1.1 rt 0.31 mm, p < 0.001; t- test). After 6 wk, non-calcifying tissues still covered a significantly greater portion of the subordinate branch RESULTS than did calcifying tissues (15.4 + 7.5 and 1.6 * 7.86 mm, respectively; p < 0.001).After 10 wk, all over- Allogeneic interactions in Millepora dichotorna growing tissues which had started as non-calcifying tissues had already deposited calcium carbonate, and Tissue fusion was evident in all Millepora dichotoma the rapid advancement of these tissues slowed down or isograft assays within 1 to 3 wk after initial tissue-to- stopped. Still, these tissues occupied a significantly tissue contact. Allogeneic combinations did not reveal greater part of the overgrown branch as compared to any tissue fusion (cf. Frank & Rinkevich 1994). Primary those that had started as calcifying tissues (16.4 + 7.86 incompatible allogeneic responses (Frank & Rinkevich and 2.5 + 0.97 mm, respectively; p < 0.001). Polyps 1994) in all 18 M. dichotoma colony-combinations were not observed in non-calcifying tissues. Histologi- were recorded as unilateral overgrowths, starting only cal sections revealed, however, that the overgrowing
Table 1 Primary allogeneic responses between Millepora dichotoma conspecifics as recorded after 2, 6 and 10 wk following Initial tissue-to-tissue contact. Numbers of replicates are given; numbers in parentheses (only for colonies A to F) indicate the average overgrowth distance in mm. < and > indicate overgrowth directionality; Ca: calcification; NCa: no calcification
Combination After 2 wk After 6 wk After 1.0 %vk
AB 2 A> B, NCa; (5) 2 A>B, NCa; (7) 2 A>B, Ca; (7) AC 2 A>C, Ca; ( rapid advancement of the 'superior' colony's tissue slowed down signifi- cantly or even stopped, and calcium carbonate was deposited, starting at the tip area of the original tissue-to-tis- sue contact. This result was then con- firmed by in situ observations (Fig. 1) and by alizarin red staining (pink-red- dish color on the overgrown tips of the subordinate branch). Xenogeneic interactions between Cyphastrea chalcidicum and Savig- nium dentatum The plates of all living Savignium dentatum inhabiting Cyphastrea chal- cidicum colonies were found to be cov- ered by homogeneous zooxanthellate coral tissue without calcification when observed under the stereomicroscope. Polyps were absent on the plates of all living barnacles (Figs. 3 to 5). The coral tissue advanced up to the barnacle's aperture, and was tightly attached to the barnacle's plates. This type of over- CI growth differs, therefore, from the loosely attached non-calcifying Mille- en- pora dichotoma tissue envelopes that arise during allogeneic encounters. All studied colonies of C. chalcidicum stained well and evenly with alizarin red, as expected for a hemispheric colony. However, the plates of all living Figs. 1 & 2. Millepora dichotoma.Fig. 1. Allogeneic overgrowth, 4 wk following initial tissue-to-tissue contact. Tissue growth without calcification has been fol- barnacles remained white and did not lowed by calcificati.on, startinq from the onqinal contact area. Arrowheads stain at all, indicatinq that no calcium indicate the original contact aria (l),edge of ca~cif~in~overgrowing tissue (2) carbonate was by the overly- and non-calcifying tissue (3).Scale bar = 10 mm. Fig. 2. Cross section (2 pm) in.g coral tissue. SEM preparations from through overgrowing tissue without calcification (h~matoxylinand eosin). ec: ectoderm; en: endoderm; m: mesoglea; n: nucleus of endodermal cell; nem: plates of living S. nemdtocyst; z: zooxanthella cell. Scale bar = 10 pm only a few, limited spots of calcified material of coral origin (data not shown), which might explain the tight tissues contained endosymbiotic zooxanthellae, nema- attachment of the coral tissue to the plates. Even lim- tocyts and other cell types, arranged in ectodermal and ited coral calcification over the barnacle plates can be endodermal cell layers as in regular M. dichotoma tis- easily seen because the radial strands and projections sue (Fig 2). The skeletons of the subordinate branches which are characteristic of the plates for S. dentatum below the non-calcifying allogenelc sheets of tissue did are not visible if covered by coral skeleton (Figs. 3 to 6). not stain with alizarin red, indicating that no new Plates from dead ~ndlviduals,on the other hand, were calcium carbonate was deposited there. In contrast, always covered by calcium carbonate of coral origin other branch termini, including those which were (Figs. 4 & 6) and stained well with alizarin red. Their overgrown by tissue and skeletons (37 assays; Table l), apertures were almost completely sealed by coral stained weakly pink-reddish. The prlmary allogeneic skeleton already 3 wk after their death (Fig. 4). Histo- responses in the non-calcifying overgrowth assays logical sections of C. chalcidicum did not reveal any were usually completed within 2 wk. Thereafter, the significant difference between calcifying and non- urrl O& = ieq ale3s .aleld pa!jtqesap jo uoqesol :d :hltnes sulse6 :6 !uriapopua :ua !wlapolsa :sa !anssg s,apeuieq :q '(u!soa pue u~~dxoleuraq)anss11 lelos 6u!n~oilj~anoaql pue apeuleq pa!jp[eJap q6noiql (urrl Z) uo!lsas ss013 .cunleluap .S pue cunJ!pp[eq>.3 .md 00~= ieq alms ,ur6!.1o [elos $0 aleuoq~un!3le3 411~paiaAos MOU 'aleid aql uo uoqsaroid salexput peaqmoliv ,6utldwes alojaq p p[ pal[!y sent q~tq~apeuieq e ruoij ale~de jo uoqeledald ~3s'urnlejuap 'S 'g '61~.mm T = ieq aless .ale~ds,aIseuieq uo uotparoid :d .(~a!~iaddn)apeuleq aA!I e molj aleld e jo uogeiedald ~3s'wiijejuap 'S 'S '6!g 'wur E =leg aless .dhlodle~os'd !alseuIeqal\!l'? !uola[ays [elos r(q paleas a~nllade:uayel seM qde~6oloqds!ql aJojaq p 12pa[[ry seM q31q~alseuleq 'a.urn]eluap 'S Aq pazruolos cunJ,rpp/eyJ .n 10 uolalays pasodxa .p ,61j ~LLILUg = leq aless 'dr([od[p103 :d !alseuleq :g ,urnjejuap ,S Aq paz~uo[o~runJ!ptJ[eqJ ,3 jo ~a!n~anoC '61.~ .ru~?jeluap urnrubl~e~ pue urnJrpr; ~seqdb ol E 's6:j a- 'm. - - 186 Mar Ecol Prog Ser calcifying tissues, except of the lack of polyps in the nutrient supply, as has been suggested by Ivker (1972) latter (Fig. 7). Cross sections of coral tissue overgrow- for allogeneic interactions between Hydractinia echi- ing living S. dentatum plates clearly demonstrated the nata conspecifics. Calcium carbonate is probably typical structure of the inter-polyp madreporarian tis- deposited by the overgrowing partner only on clean sue, consisting of 2 body walls separated by a gastric areas of the subordinate colony's skeleton. Rapid over- cavity (Fig. 7). Typical cnidarian cells were arranged in growth without calcification was displayed in at least 1 ectodermal and endodermal cell layers in each body combination by only 6 colonies out of the 9 studied wall. Symbiotic zooxanthellae were observed within (Table 1). The occurrence of non-calcifying tissue in endodermal cells, which had a normal appearance, only some of the combinations, and its high degree of like those of the control sections. reproducibility within repeatable assays of the same combinations (except for G1 and HI; Table l), suggest that a specific genotype-combination is required for DISCUSSION this effector mechanism to be expressed. Similar speci- ficity has also been recorded for other effector mecha- We have documented here the occurrence of growth nisms in the scleractinian corals Stylophora pistillata without calcium carbonate deposition in 2 hermatypic and Acropora hemprichi (Chadwick-Furman & Rinke- cnidarians, Millepora dichotoma and Cyphastrea chal- vich 1994 and Rinkevich et al. 1994, respectively). cidicum, engaged in allogeneic and xenogeneic inter- Overgrowth of Savignium dentatum plates by actions, respectively. These species represent 2 dis- Cyphastrea chalcidicum tissue without skeleton depo- tinct classes of the Cnidaria: M. dichotoma belongs to sition and polyp development is probably the result of the Hydrozoa, C. chalcidicum to the Anthozoa. The a different mechanism. The non-calcifying coral tissue phylogenetic distance between these 2 species within above the barnacle's plates, which probably provides the Cnidaria suggests that growth without calcification protection against predation (Patton 1976), was found is not a unique phenomenon, and may be expressed by to be a long-term situation, lasting as long as the bar- other reef-building corals as well. nacle lives (up to 5 yr: Brickner 1994).Death of the bar- Unilateral overgrowth by tissue without calcification nacle was immediately followed by calcium carbonate or polyp development in allogeneic encounters of deposition by the overgrowing coral tissue and eventu- Millepora dichotoma was a temporary phase during ally by appearance of coral polyps above the plates. On the establishment of primary allogeneic responses. It living barnacles, calcium carbonate deposition by the was always followed by development of regular calci- overgrowing coral tissue may prevent the barnacle's fying tissue and polyps, usually within the 4 to 10 wk growth by cementing the sutures between plates and following initial tissue-to-tissue contact. The non- bases (Ross & Newman 1973), and eventually may calcifying tissue advanced rapidly over the subordi- reduce food and oxygen supply by sealing the aper- nate branch and was loosely attached to it until calcifi- ture. S. dentatum presumably possesses some physio- cation began. Overgrowth by non-calcifying tissue logical-biochemical mechanisms to prevent skeleton extended for significantly greater distances over the deposition by the overgrowing C. chalcidicurn tissue. subordinate branches, as compared to the regular We suggest that tissue growth without calcification, calcium-carbonate-depositing tissue (Table 1). The as observed in the 2 hermatypic species studied here, absence of polyps is probably attributable to the reflects 2 similar end products of 2 different biological extremely rapid growth. This rapid tissue advance- processes with unlike ecological backgrounds. Growth ment, without calcification or polyp development, may without calcification in Millepora dichotoma was prob- therefore be regarded as a newly described allogeneic ably triggered in the superior partner as a result of effector mechanism that evolved for rapid occupation tissue contact with a specific allogeneic colony. In of substratum. The efficiency of this type of growth is Cyphastrea chalcidicum, on the other hand, growth well demonstrated by the significantly greater areas without skeleton deposition was induced by the over- occupied on the subordinate branches by non-calci- grown barnacle Savigniurn dentatum. The non-calcify- fying tissues as compared to normal calcifying tissues. ing tissues were histologically indistinguishable from Similar responses involving rapid growth without regular tissues overlying skeleton in both studied spe- calcification are also found in other hermatypic corals cies. One may therefore assume that the calcification- in cases of rapid regeneration after tissue lesions, and inhibiting agent directly prevents the precipitation of produce features referred to as 'lip of tissue' (Bak et al. calcium carbonate at the biochemical level and not by 1977, Bak & Steward-Van Es 1980). The overgrowing obstructing the development of a normal calicoblastic tissue slowly eliminates the tissue of the subordinate layer. The rapid calcification of C. chalcidicum tissue branch by the expression of unknown effector mecha- immediately following the barnacle's death supports nisms, or indirectly by drastically reducing light and this hypothesis. Rapid overgrowth without skeleton Frank et al.: Coral growth rv~thoutcalcification deposition is an allogeneic effector mechanism in M. DeVantier LM, Reichelt RE, Bradbury RH (1986) Does Spiro- dichotoma, and the inhibition of calcium carbonate branchus giganteus protect host Porites from predation by Acanthaster planci: predator pressure as a mechanism of deposition in C. chalcidicurn by S. dentaturn is a xeno- coevolution? Mar Ecol Prog Ser 32:307-310 geneic effector mechanism, induced by the barnacle. Fagerstrom JA (1987) The evolution of reef communities. As such, neither phenomenon has been described pre- John Wiley & Sons, New York viously. They should be taken into account in future Frank U, Rinkevich B (1994) Nontransitlve patterns of his- coral growth and competition studies. torecognltion phenomena in the Red Sea hydrocoral M~lleporadichotoma. Mar Biol 118:723-729 Frank U, ten Hove HA (1992) In vitro exposure of Spiro- Acknowledgements. We thank the staff of the H. Steinitz branchus giganteus and S tetraceros (Polychaeta, Serpuli- Marine Biology Laboratory in Eilat for diving facilities and dae) to various turbidities; branchial morphologies an hospitality. Yaacov Langsam (Department of Electron expression of filtering strategy? Oebalia 18:45-52 Microscopy. Bar Ilan University) is kindly acknowledged for Gohar HAF. Soliman GN (1963) On three mytilid species bor- making the SEM preparations. Funds were provided by the ing in living corals. Pub1 Mar Biol Stn Ghardaqa, Red Sea GIF, the German-Israeli Foundation for Scientific Research 12:65-98 and Development (B.R.). Hiro F (1938) Studies on the animals inhabiting reef corals 11. Cirripeds of the genera Creusia and Pyrgoma. Palao Trop Biol Stn Stud 1:391-416 LITERATURE CITED lvker FB (1972) A hierarchy of histo-~ncompatibility in Hydractrnia echinata. Biol Bull 143:162-174 Anderson DT (1992) Structure, functlon and phylogeny of Jackson JBC, Buss L (1975) Allelopathy and spatial competi- coral-inhabiting barnacles (Cirripedia, Balanoidea). Zoo1 J tion among coral reef invertebrates Proc Natl Acad Sci Linn Soc 106:277-339 USA 72:5160-5163 Bak RPM (1973) Coral weight increment in-situ. A new Jokiel PL, Maragos JE Franzisket L (1978) Coral growth: method to determine coral growth. Mar Biol 20:45-49 buoyant weight technique. In: Stoddard DR. Johannes RE Bak RPM, Brouns JJWM, Heys FML (1977) Regeneration and (eds) Coral reefs: research methods. UNESCO, Paris, p aspects of spatial competition in the scleractinian corals 529-54 1 Agaricia agaricites and Montastrea annularis. In: Taylor Lamberts AE (1978) Coral growth: alizarin method. In: Stod- DL (ed) Proc 3rd Int Symp Coral Reefs, Vol 1. School of dard DR. Johannes RE (eds) Coral reefs: research meth- Marine and Atmospheric Sciences. University of Miami, ods. UNESCO. Paris, p 523-527 Mlami, p 143-148 Lang JC, Chornesky EA (1990) Competition between sclerac- Bak RPM, Cnens SR (1982) Experimental fusion in Atlantic tin~an reef corals - a review of the mechanism and Acropora (Scleractinia). Mar Biol Lett 3.67-72 effects. In Dubinsky Z (ed) Ecosystems of the world; coral Bak RPM, Steward Van Es Y (1980) Regeneration of superfi- reefs. Elsevier, Amsterdam, p 209-252 cial damage in the scleractinian corals Agancia agaricites, Loya Y (1972) Community structure and specles diversity of f purpurea and Porites astreoides. Bull mar Sci 30: hermatypic corals at Eilat. Red Sea. Mar Biol 13:lOO-123 883-887 Marsden JR. Conlin BE, Hunte W (1990) Habitat selection in Bak RPM, Termaat RM, Dekker R (1982) Complexity of coral the tropical polychaete Splrobranchus giganteus 11. Larval interaction: influence of time, location of interaction and preference for corals. Mar Biol 104:93-99 epifauna. Mar Biol 69:215-222 Morton B (1983) Coral-associated bivalves of the Indo Pacific. Bancroft ID, Stevens A, Turner DR (1990) Theory and practice of In: Russell Hunter WD (ed) The Mollusca, Vol6. Academic histological techniques. Churchill Livingstone, Edinburgh Press, Orlando, p 139-224 Barnes DJ (1970) Coral skeletons: an explanation for their Morton B. Scott PJB (1980) Morphological and functional spe- growth and structure. Science 170.1305-1308 ciallsations of the shell, musculature and pallial glands in Barnes DJ, Chalker BE (1990) Calcification and photosynthe- the Lithophaginae (Mollusca: Bivalvia). J 2001, Lond 192: sis in reef-building corals and algae. In: Dubinsky Z (ed) 179-203 Ecosystems of the world; coral reefs Elsevier, Amsterdam, Patton WK (1976) Animal associates of living reef corals. In: p 109-131 Jones OA, Endean R (eds) Biology and geology of coral Brickner I (1994) Ecology, systematics and phylogenetics of reefs, Vol3. Academic Press, New York. p 1-36 coral inhabiting barnacles in the Gulf of Eilat. PhD thesis, Potts DC (1976) Growth interactions among morphological Bar Ilan University, Ramat Gan (in Hebrew with English variants of the coral Acropora palifera. In: Mackie GO (ed) summary) Coelenterate ecology and behavior. Plenum Press, New Brown BE, Le Tissier MDA, Dunne RP (1994) Tissue retraction York, p 7 9-88 in the scleractinian coral Coeloseris mayeri, its effect upon Rinkevich B, Frank U, Bak RPM, Miiller WEG (1994) Allo- coral pigmentation, and preliminary implications for heat immune responses between Acropora hemprichi con- balance Mar Ecol Prog Ser 105:209-218 specifics. nontransitive patterns of overgrowth and Chadwick Furman NE, Rinkevich B (1994) A complex allo- delayed cytotoxicity. Mar Biol 118:?31-737 recognition system in a reef bulldlng coral: delayed Rinkevich B, Loya Y (1979) The reproduction of the Red Sea responses, reversals and nontransltive hierarchies. Coral coral Stylophora pistillata. I. Gonads and planulae. Mar Reefs 13:57-63 Ecol Prog Ser 1:133-144 Chornesky EA (1991) The ties that blnd: inter-clonal coopera- knkevich B. Loya Y (1983) Intraspeclflc competitive networks tion may help a fragile coral dominate shallow high- in the Red Sea coral Stylophora pistillata. Coral Reefs 1: energy reefs. Mar Biol 109:41-51 161-172 Davies PS (1989) Short-term growth measurements of corals Rinkevich B, Loya Y (1984) Coral illumination through an using an accurate buoyant weighing technique. Mar Ecol optic glass-fiber: incorporation of I4C photosynthates. Mar Prog Ser 101:389-395 Biol 80:f-15 Mar Ecol Prog Ser 124: 181-188, 1995 Ross A, Newman W (1973) Revision of the coral inhablt~ng Oranchus giganteus (Pallas) breaks a rule for suspension- barnacles (Cirripedia: Balanidae). Trans San Diego Soc feeden. J exo mar Biol Ecol79:245-249 Nat Hist 17:137-174 ten Hove HA (1989) Serpulinae (PolychaetaJ from the Salter-Cld L. Bigger CH (1991) Allo~mrnunltyin the gorgon- Caribbean 1V Pseudovermilja rnadrdcicola su.n.. a svm- ian coral Swiftia excerta. Biol Bull 181:127-l34 biont of corals. Studles in honour of Dr. P~eterWagenaar Sheppard CRC (1985) Unoccupied substrate in the central Hummelinck. Found Scient Res Surinam Neth Antilles Great Barrier Reef: role of coral interactions. Mar Ecol 123:135-144 Proy Ser 25:259-268 Wulf JL. BussLW (1979) Do sponges help hold coral reefs Strathmann RR, Cameron RA, Strathrnann MF (1984) Spiro- together? Nature 28 1:474-475 This article was submitted to the editor Manuscript first received: October 31, 1994 Revised version accepted: February 24, 1995