Internal bioerosion of the coral Porites lobata Dana (Coelenterata, Scleractinia) was examined on 5 reefs across the central region of the Great Barrier Reef, Australia: Orpheus Island, Britomart Reef, Rib Reef, Davies Reef, and Myrmidon Reef. Ninety corals were sliced into discs (n, = 270 replicates) and X-rayed to identify internal bioeroders and quantify the degree of bioerosion caused by each bioeroding group. Total internal bioerosion decreased significantly with distance offshore, across the continental shelf. Abundances of bivalves and sponges decreased the most significantly. Sponges were significantly lower on the outer shelf. Abundance of sipunculans and polychaetes showed only a nomlnal decrease across the shelf. Boring bivalves dominated the inshore area while the relative abundance of sponges increased with &stance from shore. The relative abundance of worms (polychaetes and sipunculans) remained generally constant across the shelf. The explanatory hypothesis is offered that there may be a link between internal bioerosion and the already documented decreasing productivity and increased abundance of grazing fish with distance from shore. Three predictions are also made: (1) that observed changes in internal bioeroders may influence sediment composition across the continental shelf, with inshore sediments being more poorly sorted, and possessing more fines produced by internal bioeroders; (2) that offshore lagoonal sediments may have a greater contribution derived from grazing fish; and (3) that the observed higher levels of bioerosion inshore may increase susceptibility of corals there to dislodgement and fragmentation imposed by physical disturbance.

Experiments were performed on the Great Barrier Reef, at Orpheus lsland (18" 40'S, 145" 30' E) and Lizard Island (14" 41' S, 145" 28' E), Australia, to examine allelopathic effects of soft corals on the larval recruitment of scleractinian corals. Ceramic tiles were used as settlement plates and arranged around the soft corals Sinularia flexibilis (Quoy & Gaimard) and Sarcophyton glaucurn (Quoy & Gaimard), serving as treatments. One control consisted of stacks of settlement plates uninfluenced by any organism, while a second control had settlement stacks surrounding a scleractinian coral to control for depletion of larvae via feeding. Coral spat recruitment was approximately 7 times higher at Lizard Island (7032) than at Orpheus lsland (1038). The pattern of coral recruitment and relative abundances of coral recruits around the controls and soft corals, however, was similar at the 2 sites. The average density of coral spat was always significantly less around the soft corals than the controls, indicating that soft corals at each experimental site inhibited scleractinian coral recruitment. Coral recruitment levels were very similar for the 2 control treatments. Coral spat were asymmetrically distributed around the soft corals, and the distribution varied significantly with respect to the distance from the soft coral and the direction of the prevailing current at both Orpheus and Lizard Island. There was a significant negative correlation between coral spat density and current direction under both soft coral treatments and at both sites. No such response was observed in the controls. In a second experiment, scleractinian coral spat were exposed to settlement plates impregnated with the extract of S. flexibilis and control plates. Settlement only occurred on the untreated controls.

The Octocorallia are a highly diverse group, rich in secondary compounds which serve several ecological functions, including predator defense, interspecific competition for space, antifouling, and reproduction. The Alcyonacea are characterized by high frequencies of ichthyotoxicity and feeding deterrence which are not generally correlated with each other. This may be due to the multiple functions these compounds perform, m a s l n g correlations which would otherwise be apparent. Correlations between ichthyotoxicity and physical defense exist, but only at higher levels of taxonomic resolution. One resultant selective advantage of varlance in ichthyotoxlcity and feeding deterrence may be Diffuse Batesian Mimicry, whereby the toxic or feeding-deterrent characteristics of a majority of species help to protect a taxon with a small number of general morphological types and small variance in coloration from generalist predators. Surface-brooded planulae of Indo-Pacific coral species, similar in appearance but not possessing toxic or feeding deterrent attributes, may be similarly protected. The chemical characters of octocorals may serve the same functions which morphological characters in scleractinian corals serve, particularly in predator defense and competition for space. Most octocorals lack a rigid skeleton possessed by scleractinians, which can protect polyps from predation. Octocorals also lack toxic stinging nenlatocysts common to scleractinians, but possess allelopathic capabilities. In some octocorals, 'non-toxic' nematocysts on sweeper tentacles may still be effective in competition for space; in others, aggressive interactions can be avoided through re-orientation or movement, resulting in directed growth. Octocoral orders are distributed inequitably between the tropical Atlantic and Indo-Pacific. The Gorgonacea dominate the former; the Alcyonacea. the latter. Octocoral species diversity is higher in the western Indo-Pacific, as is the abundance and diversity of secondary rnetabolites and their functions. This may be a result of differential levels of extinction between the 2 regions dunng the early Miocene and the late Pliocene/early Pleistocene, with high extinction rates having been experienced in the Caribbean. These extinctions may also have affected radiation in the secondary metabolites produced by those species and the functions which those rnetabolites serve. Secondary metabolites appear to have contributed to the evolutionary success of the Indo-Pacific Alcyonacea. In accordance with the Time-Stability hypothesis, we believe that this high abundance and diversity evolved in the Indo-Pacific under conditions of reduced climatic change and lower extinction levels, when compared to the Caribbean.

We have tested 136 specimens of soft corals (Coelenterata. Alcyonacea) for toxicity by exposing Gambusia affinis (Vertebrata, Pisces) to aqueous extracts of coral macerate and assessing mortality. Sixty-eight corals were examined in detail utilizing behavioral responses of fish as well as mortality rates to establish relative toxicities among and within coral genera. The responses exhibited by the fish ranged from rapid mortality through slow mortality and varying levels of narcotisation to negligible effects. With respect to individual soft corals, the fish exhibited varying patterns of response through time, and were sorted into 9 distinct groups via multivariate computer analyses according to similarity of these behavioral patterns. Only 2 of the genera tested, Lemnalia and Sarcophyton, were restricted to the most toxic set of groups. A number of genera such as Lobophytum, Sinularia, Nephthea, and Cespitularia had representatives which spanned the entire spectrum of responses from highly toxic to non-toxic. The wide range of toxicity determined for the specimens examined and the fact that approximately 50 % of all specimens were toxic suggests that chemical defences against predation are common in the Alcyonacea of the central region of the Great Barrier Reef. Toxicity, however, occurs significantly less frequently in this group than in those examined from the northern Great Barrier Reef in an earlier study.

Experiments were performed in Discovery Bay (Jamaica, W I.) to determine the relationship between behavior of threespot damselfish Eupornacentrus planifrons Cuvier, and distribution of the common shallow-water echinoid Diaderna antillarum Philippi. E planifrons actively excludes Diadema from its algal lawn, causing a distinct alteration in local distribution of Diaderna on the reef. Physical attacks on the urchin were usually sustained for 2 5 min. The damselfish's agonistic behavior occurred only dunng the day. Diadema densities were higher withln algal lawns at night than during the day but were always significantly lower than the mean density outside. The size of the damselfish's algal lawn was positively correlated with damselfish density, implying that damselfish which occur in aggregations are able to defend larger territories together than they would when occurring alone. The impact of interactions between fishes, urchins and sessile epibenthos are predictable from experimental data derived from a concurrent study. E. planifrons territories are probably sites of locally high mortality in coral spat, particularly in Agaricia and Porites spp. This would result from competition for space with algae and encrusting epifauna. The algal lawns, however, may also represent refuges for other corals such as Favia fragum which are well-adapted to survive competition with other sessile epifauna as well as with algae. Thus, the algal lawns could represent dlscrete patches In the environment within which mortality of dominant species is high and that of rarer species is relatively low, causing an increase in the latter's representation in the benthic community. The net effect of this would be a n overall increase in coral diversity within the community. This interpretation is consistent with hypotheses concerning the role of patch-formation in determining con~munity structure.