Demographic plasticity in tropical reef fishes

We use age-based analyses to demonstrate consistent differences in growth, mortality, and longevity of coral reef fishes from similar habitats (exposed reef crests) 20 km apart. On outer-shelf reef crests of the northern Great Barrier Reef (GBR), size in four taxa of reef fishes (Chlorurus sordidus, Scarus frenatus, and S. niger and the acanthurid Acanthurus lineatus) was systematically and significantly smaller when compared with the same taxa on adjacent mid-shelf reef crests. Differences in size could be attributed to differences in growth between habitats (shelf positions). On outer reef crests the species examined had consistently lower size at age profiles and also reduced life spans compared with populations from mid-shelf reefs. To confirm this relationship, two of the most abundant species (C. sordidus and S. frenatus) were selected for more detailed spatial analysis of demographic patterns. Sampling adults of both taxa from reef crests on three mid- and three outer-shelf reefs revealed that most of the variation in growth was explained by shelf position, although C. sordidus also displayed differences in growth among mid-shelf reefs. We conclude that differences in body sizes across the continental shelf of the northern GBR are primarily determined by these trends in growth. Strong spatial patterns also existed in the mean ages of populations and longevity estimates for C. sordidus and S. frenatus between shelf positions. Both species on outer-shelf reefs displayed less variable cohort sizes, significantly reduced mean ages, and foreshortened longevity compared with populations on mid-shelf reefs. Furthermore, differences in these parameters were rare among replicate reefs within mid- and outer-continental-shelf positions. Age-based catch curves suggested that rates of S. frenatus natural mortality on the outer shelf were nearly twice as high as on the mid shelf. Visual surveys indicated that total scarid densities on outer-shelf reef crests are on average fourfold higher than for equivalent mid-shelf habitats. This fact, coupled with significantly reduced growth rates, reduced mean ages, and increased mortality rates, suggests that density-dependent processes may be responsible for observed differences among localities.     

Microstructure of settlement-marks in the otoliths of tropical reef fishes

The morphology and ultrastructure of the otolith settlement-mark was examined in 44 tropical reef-fish species spanning nine families. A classification scheme based on similar otolith characteristics is presented. Three major categories are identified based on changes in increment width and optical qualities of the settlement-mark. Of the 44 species examined, 39 possessed “abrupt” settlement-marks (Type I) characterised by a rapid decrease in increment width (up to 50% reduction) over settlement. Type I settlement-marks were found in all nine families examined. The 39 species spanned the whole range of possible larval durations (Pomacentrus moluccensis, 15 d ± 0 SE; Naso hexacanthus, 91.2 d ± 2.97 SE). Four of the 44 species possessed “zonal” settlement-marks (Type II), featuring a band of increments that are wider than pre-settlement increments. Species in this category are the labrids Corisaygula, Thalassoma bifasciatum, T. lunare and an unidentified acanthurid (Acanthurus sp. 2). One species of acanthurid (N. brevirostris) possessed a “gradual” settlement-mark (Type III), manifest as a gradual decrease in increment width during the settlement period. A possible fourth type was identified from the literature. Gnatholepis thompsoni and Coryphopterus glaucofraenum possessed a settlement-mark with increment widths that increased post-settlement. Available data suggest a poor relationship between the structure of the settlement-mark and the magnitude of metamorphosis (previously reported as internal and external morphological change). Evidence suggests that the increment profile over early development and the increment transitions associated with the settlement event are taxon-specific and may enable late-larval stage fishes to be identified to species level. Received: 21 May 1997 / Accepted: 3 February 1999     

Association of color and feeding deterrence by tropical reef fishes

Summary. While many marine molluscs have been suggested to use aposematic coloration to avoid predation, few studies have tested the ability of marine predators to learn to associate colors with distasteful prey. In field experiments, we tested the ability of two populations of reef fishes to discriminate among red, yellow, and black artificial nudibranch models when one color was paired with a feeding deterrent. We offered fishes (1) the models without any feeding deterrents, (2) the models with a feeding deterrent coated onto one color, and (3) the models without deterrents again. If reef fishes learn to associate colors with noxious prey, we expected the color paired with the feeding deterrent to be eaten less frequently in the final assay than the initial assay. In both populations, fishes formed clear associations between color and feeding deterrence. However, when the experiment was repeated in one population, changing the color paired with the feeding deterrent, fishes did not form an association between color and feeding deterrence. In this case, prior learning may have affected subsequent trials. Our study indicates that common colors of nudibranchs are recognizable by fishes and can be associated with noxious prey. Received 24 September 1998; accepted 18 December 1998.     

Short-term temporal patterns of early recruitment of coral reef fishes in the tropical eastern Pacific

Short-term temporal patterns of recruitment have been described in a variety of coral reef fishes and have often been related with lunar and tidal cycles. While the relative importance of lunar and tidal factors in determining recruitment patterns has been difficult to assess, most studies have been done in the Caribbean and Indo-Pacific, where tidal amplitudes are small. We studied the short-term temporal dynamics of fish recruitment at Gorgona Island (tropical eastern Pacific), where there is a large tidal amplitude (~4.4 m). Every other day during three consecutive months in 1998, we directly measured the magnitude of reef fish recruitment to standardized coral units (SCUs) isolated from natural reefs. A total of 40 species from 21 families settled on the SCUs. Of 11 species with sufficient numbers for meaningful statistical analyses, two (Lutjanus guttatus and Pomacanthus zonipectus) had lunar recruitment with peaks near the new moon; three combined species of antennariids showed semilunar recruitment with peaks near moon quarters; and eight other species showed sporadic and aperiodic recruitment pulses. The contribution of lunar (moonlight intensity) and tidal factors (tidal amplitude and net tidal flow) to recruitment dynamics varied among species, although it was generally low (<18%) even among species with periodic patterns, except perhaps in L. guttatus. In this species, recruitment magnitude correlated negatively with moonlight intensity, accounting for 34.5% of the variance. Post-settlement predation by roving predators may be one cause of this relationship. In the remaining species, particularly those with sporadic and aperiodic recruitment pulses, stochastically varying weather and oceanographic events may be more important in determining temporal variation in recruitment.     

Reproductive isolation in temperate reef fishes

Opportunity and recognition isolation can lead directly to reproductive isolation, the former via divergence in the location and timing of breeding, and the latter via differential mate preferences. We describe the potential significance of these factors in the maintenance of reproductive isolation in a clade of triplefin fishes that occur sympatrically around coastal New Zealand. Specifically, we investigate the roles of spawning time and nesting habitat in promoting opportunity isolation, and of interspecific variation in male body length and breeding colouration in promoting recognition isolation. The triplefin species investigated are reproductively active over several months and show high overlap in breeding times, thus rejecting temporal isolation as a mechanism. Differences in nesting habitats resulted in a reduced probability of encounter between some species, especially between sister-species pairs. Interspecific colour differences generally decreased during the reproductive period, and males of sister-species pairs showed no interspecific colour differences in the ultraviolet light spectrum, thus mate selection based on male colour patterns is unlikely to lead to premating isolation. Finally, males of closely related triplefin species differed in body length, a secondary sexual trait often involved in assortative mating. Thus, spatial differences in nesting habitats reduce the chances of encountering allospecific mates, which may facilitate opportunity isolation and differences in male length, possibly related to species-specific female selection on male body size, may lead to recognition isolation. The combination of limited spatial overlap in nesting habitat and differences in male body size may facilitate species assortative mating in sympatry or parapatry.     

Video-monitored predation by Caribbean reef fishes on an array of mangrove and reef sponges

Although predation by fishes is thought to structure benthic invertebrate communities on coral reefs, evidence to support this claim has been difficult to obtain. We deployed an array of eight sponge species on Conch Reef (16 m depth) off Key Largo, Florida, USA, and used a remote video-camera to record fish activity near the array continuously during five daylight periods (6 h for 1 d, at least 11.5 h for 4 d) and one night period (11 h). Of the eight sponge species, four were from adjacent reefs (Agelas wiedenmayeri, Geodia neptuni, Aplysina fistularis, and Pseudaxinella lunaecharta), and four were from a nearby mangrove habitat (Chondrosia collectrix, Geodia gibberosa, Halichondria sp., andTedania ignis). Each species of reef sponge was chosen to match the corresponding mangrove species in form and color (black, brown, yellow, and red, respectively). Predation events only occurred during daylight hours. Tallies of the number of times fishes bit sponges revealed intense feeding by the expected species of sponge-eating fishes, such as the angelfishHolacanthus bermudensis, H. tricolor, andPomacanthus arcuatus, the cowfishLactophrys quadricornis, and the filefishCantherhines pullus, but surprisingly also by the parrotfishSparisoma aurofrenatum andS. chrysopterum. Of 35 301 bites recorded, 50.8% were taken by angelfish, 34.8% by parrotfish, and 13.7% by trunkfish and filefish. Mangrove sponges were preferred by all reef fishes; 96% of bites were taken from mangrove species, with angelfish preferringChondrosia collectrix and parrotfish preferringGeodia gibberosa. Fishes often bit the same sponge repetitively, and frequently consumed entire samples within 30 min of their deployment. Sponge color did not influence fish feeding. Two of the four mangrove sponge-species deployed on the array were also found living in cryptic habitats on adjacent reefs and were rapidly consumed by fishes when exposed. Our results demonstrate the importance of fish predation in controlling the distribution of sponges on Caribbean reefs.     

Interactions during feeding among certain coral reef fishes in Elat

From July 1983 to March 1984 coral reef fishes in Elat (Red Sea) were fed in novel feeding situations. Thalassoma rüppelli had the shortest latency to first feed, Chaetodon paucifasciatus had the longest latency, while Sufflamen albicaudatus, Coris aygula and Lethrinus sp. had intermediate latencies. The mean number of T. rüppelli feeding at the beginning of experiments was higher than that of the other species. The difference decreased rapidly and disappeared within 90 s of the beginning. Latencies to first feed, and the number of fish feeding, were not correlated with the number of fish of each species in the study area. Aggression was directed predominantly by S. albicaudatus and by C. aygula towards Lethrinus sp. and T. rüppelli. It is suggested that these species which suffer more disturbances during their feeding and which receive more aggression are more likely to approach and feed more quickly in novel feeding situations.     

Diel patterns of abundance of presettlement reef fishes and pelagic larvae on a coral reef

Most presettlement reef fish settled at night at One Tree Island, Great Barrier Reef. Fish were sampled day and night using channel nets located on the reef crest, and a plankton-mesh purse-seine net in the lagoon (1992–1994). Catches of fish at night were generally tens to hundreds of times greater than those taken during the day. Preflexion fish, as well as postflexion and pelagic juveniles, were taken in greater numbers at night. Preflexion forms were a combination of those that had hatched from demersal eggs and later stages that had been transported over the reef crest. Highest numbers of postflexion and pelagic juvenile forms of Apogonidae, Blenniidae, Gobiesocidae, Gobiidae, Labridae, Lutjanidae, Mugiloididae, Mullidae, Pomacentridae, Scaridae, Serranidae and Tripterygiidae were found at night. Observations, while SCUBA diving, and purse-seine samples in the lagoon indicated that the only resident larvae were of the genera Spratelloides and Hypoatherina; most of the fishes caught in nets, therefore, were immigrants. Patch reefs, sampled for new settlers early in the morning and late in the day, indicated that the majority of apogonids (Apogon doederleini, >95%) settled at night. Although greater numbers of pomacentrids were found in morning counts (e.g. Pomacentrus wardi), if data were converted to an hourly rate, many pomacentrids showed a similar hourly rate of settlement day and night. Depth-stratified sampling in waters near One Tree Island (to 20 m) indicated that some taxa rise to the surface at night. This behaviour, perhaps combined with avoidance of diurnal predators may explain on-reef movement of potential settlers soon after dark. Studies on settlement cues, therefore, need to focus on night-related phenomena. Received: 3 March 2000 / Accepted: 20 June 2000     

Tolerance to high temperatures and potential impact of sea warming on reef fishes of Gorgona Island (tropical eastern Pacific)

Knowledge of upper thermal-tolerance limits of marine organisms in the tropical eastern Pacific (TEP) is important because of the influence of phenomena such as El Niño and global warming, which increase sea temperature. Laboratory experiments were conducted to determine the critical thermal maximum (CTM) of reef fishes from the TEP. In 15 reef fishes of Gorgona Island (TEP) the CTM was between 34.7°C and 40.8°C. None of these CTMs was exceeded by sea temperature in the TEP during any of the strongest El Niño events in this century (32°C during El Niño 1982-1983 and 1997-1998), which indicates that all species studied here may tolerate El Niño maximum temperatures. In addition, the CTM of the least-tolerant species was 8°C above the current mean sea temperature in a wide range of latitudes in the TEP. This suggests that fishes live far from their upper thermal tolerance limits and that the current global-warming trend is still unlikely to be dangerous for these species. If sea temperature continues to increase at the current rate, in about a century sea temperature could exceed the thermal tolerance of some reef fishes and threaten them with extinction. Such risk, however, might occur sooner if the sea temperature during El Niño also increased in step with the global warming, but also because other processes involved in maintaining population, such as reproduction, can be affected at lower temperatures. The possible ability of reef fishes to adapt to increases in sea temperature is discussed.     

Marker technique for investigating gut throughput rates in coral reef fishes

Artemia sp. shells were evaluated to determine their accuracy for tracing the passage of algal filaments through the gut of the damselfish Pomacentrus amboinensis Bleeker, 1868 (family Pomacentridae), an omnivorous coral reef fish. An automatic faeces-collection apparatus enabled the quantitative collection of markers and faeces in the laboratory. Defecation rates were similar for light and heavy doses of Artemia sp. shells and controls, indicating no detrimental effects of Artemia sp. shells on the gut throughput rate of P. amboinensis. In addition, similar rates and patterns of the passage of Artemia sp. shells and the algal markers Enteromorpha sp. and Lyngbya sp. indicated that Artemia sp. shells provide a reliable representation of the throughput rate of algal filaments. The mean throughput time of P. amboinensis was 4.6 h ±0.3 SE, with a modal recovery time of 4 h. Laboratory throughput estimates were validated by comparing the distribution patterns of Artemia sp. shells in the dissected gut of specimens administered markers in the laboratory and field. In addition, the retention of markers in the stomach of P. amboinensis suggested a likely site of prolonged processing. Received: 24 October 1996 / Accepted: 18 March 1997     

Ultraviolet photosensitivity and feeding in larval and juvenile coral reef fishes

The ability of young coral reef fishes to feed using solely ultraviolet-A (UV-A) radiation during ontogeny was examined using natural prey in experimental tanks. Larvae and juveniles of three coral reef fish species (Pomacentrus amboinensis, Premnas biaculeatus and Apogon compressus) are able to feed successfully using UV-A radiation alone during the later half of the pelagic larval phase. The minimum UV radiation intensities required for larval feeding occur in the field down to depths of 90–130 m in oceanic waters and 15–20 m in turbid inshore waters. There was no abrupt change in UV sensitivity after settlement, indicating that UV photosensitivity may continue to play a significant role in benthic juveniles on coral reefs. Tests of UV sensitivity in the field using light traps indicate that larval and juvenile stages of 16 coral reef fish families are able to detect and respond photopositively to UV wavelengths. These include representatives from families that are unlikely to possess UV sensitivity as adults due to the UV transmission characteristics of the ocular media. Functional UV sensitivity may be more widespread in young coral reef fishes than in the adults, and may play a significant role in detecting zooplanktonic prey.     

Patterns of use of space in a guild of territorial reef fishes

There exists on Heron Reef, at the southern end of the Great Barrier Reef, an 8-species guild of ecologically very similar, territorial, herbivorous reef fishes. All individuals of these species maintain territories on rubble substrata throughout juvenile and adult life. Territories are defended from all other guild members. Three rubble patches, each containing residents of 3 guild species: Pomacentrus apicalis, P. wardi, and Abudafduf lachrymatus, have been monitored for 12 to 18 months. This paper examines the patterns of use of space as shown by the antecedent and subsequent histories of sites chosen by 43 new colonists, and sites vacated through the loss of 34 residents on the rubble patches. In addition, variation through time is examined in the total amount of space held in territories on each rubble patch. The 3 species show similar preferences for space as colonists, although adult and juvenile colonists behave differently. Numbers of colonists detected on rubble patches are not proportional to the resident populations of the 3 species. Residents of the 3 species are equal in their abilities to enter spaces vacated through mortality, although they differ slightly in methods used to enter sites. They do not respond preferentially to sites previously occupied by any particular species. P. wardi shows a higher rate of mortality than the other species, and residents of this species are more often dislodged by new colonists. The total amount of space held on any rubble patch did not vary during the year from September, 1972 to October, 1973. The data are interpreted as supporting the hypothesis that the 3 species do not differ in their requirements for space on rubble patches, and that they are in competition for a short supply of such space. Differences exist in their strategies for obtaining and holding such space. These differences are important for explaining the continued presence of P. wardi on rubble patches. P. apicalis and a. lachrymatus are both specialists in holding territories on the upper reef slope. No differences have been detected in their requirements or competitive abilities. P. wardi is a fugitive species on the upper reef slope, coexisting because it maintains a refuge from competition by occupying some un-preferred sites off rubble patches.     

Consequences of patch reef spacing for density-dependent mortality of coral-reef fishes

The spatial configuration of habitat patches can profoundly affect a number of ecological interactions, including those between predators and prey. I examined the effects of reef spacing on predator-prey interactions within coral-reef fish assemblages in the Bahamas. Using manipulative field experiments, I determined that reef spacing influences whether and how density-dependent predation occurs. Mortality rates of juveniles of two ecologically dissimilar species (beaugregory damselfish and yellowhead wrasse) were similarly affected by reef spacing; for both species, mortality was density dependent on reef patches that were spatially isolated (separated by 50 m), and density independent on reef patches that were aggregated (separated by 5 m). A subsequent experiment with the damselfish demonstrated that a common resident predator (coney) caused a substantial proportion of the observed mortality, independent of reef spacing. Compared to isolated reefs, aggregated reefs were much more likely to be visited by transient predators (mostly yellowtail snappers), regardless of prey density, and on these reefs, mortality rates approached 100% for both prey species. Transient predators exhibited neither an aggregative response nor a type 3 functional response, and consequently were not the source of density dependence observed on the isolated reefs. These patterns suggest that resident predators caused density-dependent mortality in their prey through type 3 functional responses on all reefs, but on aggregated reefs, this density dependence was overwhelmed by high, density-independent mortality caused by transient predators. Thus, the spatial configuration of reef habitat affected both the magnitude of total predation and the existence of density-dependent mortality. The combined effects of the increasing fragmentation of coral reef habitats at numerous scales and global declines in predatory fish may have important consequences for the regulation of resident fish populations.     

Behavioural flexibility in reef fishes responding to a rapidly changing wave environment

Flexible traits can be crucial in helping animals meet the challenges of a highly variable environment. We examined whether coral reef fishes systematically alter their swimming behaviour in response to temporal changes in hydrodynamic conditions. Exploiting site-based differences in wave energy impinging on reef habitats over calm and rough weather, we found many reef fishes changed their swimming behaviour alongside increases in wave-driven water motion. Most fishes tended to orient their bodies parallel to oncoming wave-driven flow and recruited additional stabilising fins during rough weather; such changes in swimming behaviour aid stability and minimise drag. However, not all species displayed such flexibility, with fishes possessing highly tapered pectoral fins and high levels of swimming performance continuing to successfully navigate reef habitats without significantly altering their swimming behaviour. We highlight how labile behaviours augment the morphology and performance of species to underpin their occupation of habitats subject to variable environmental conditions.     

Contrasting effects of habitat loss and fragmentation on coral-associated reef fishes

Disturbance can result in the fragmentation and/or loss of suitable habitat, both of which can have important consequences for survival, species interactions, and resulting patterns of local diversity. However, effects of habitat loss and fragmentation are typically confounded during disturbance events, and previous attempts to determine their relative significance have proved ineffective. Here we experimentally manipulated live coral habitats to examine the potential independent and interactive effects of habitat loss and fragmentation on survival, abundance, and species richness of recruitment-stage, coral-associated reef fishes. Loss of 75% of live coral from experimental reefs resulted in low survival of a coral-associated damselfish and low abundance and richness of other recruits 16 weeks after habitat manipulations. In contrast, fragmentation had positive effects on damselfish survival and resulted in greater abundance and species richness of other recruits. We hypothesize that spacing of habitat through fragmentation weakens competition within and among species. Comparison of effect sizes over the course of the study period revealed that, in the first six weeks following habitat manipulations, the positive effects of fragmentation were at least four times stronger than the effects of habitat loss. This initial positive effect of fragmentation attenuated considerably after 16 weeks, whereas the negative effects of habitat loss increased in strength over time. There was little indication that the amount of habitat influenced the magnitude of the habitat fragmentation effect. Numerous studies have reported dramatic declines in coral reef fish abundance and diversity in response to disturbances that cause the loss and fragmentation of coral habitats. Our results suggest that these declines occur as a result of habitat loss, not habitat fragmentation. Positive fragmentation effects may actually buffer against the negative effects of habitat loss and contribute to the resistance of reef fish populations to declines in coral cover.     

Variation in larval life-history traits among reef fishes across the Isthmus of Panama

We tested the hypothesis that regional differences in oceanic productivity have led to the evolution of predictable patterns of regional variation in life-history traits of pelagic larvae of tropical reef fishes. To do so we compared larval traits (egg and hatchling size, larval growth rate and duration, and size at settlement) among closely related reef fishes from the Atlantic and Pacific coasts of the Isthmus of Panama. This comparison provides a control for phylogenetic effects because those regions shared a common fauna prior to the rise of the Isthmus ˜3.5 million years ago, subsequent to which each fauna evolved independently under a very different productivity regime. We measured larval traits of 12 benthic-spawning damselfishes (Pomacentridae: Abudefduf, Chromis and Stegastes) and 13 pelagic-spawning wrasses (Labridae: Bodianus, Halichoeres and Thalassoma). These included members of each genus on each side of the Isthmus and four sets of transisthmian sister species of pomacentrids. Among the pomacentrids we found consistent transisthmian differences in hatchling size, but not in other larval traits. Essentially the reverse pattern occurred among the labrids – larval growth and duration differed consistently among congeners in the two regions, but without consistent differences in hatchling size or size at settlement. Neither relationship is predicted by the regional-productivity hypothesis. Most of the differences were quite small. Stronger phylogenetic effects on larval traits (inter- and intrageneric variation within regions) occur in both families and evidently overwhelm any effect of regional variation in productivity. Reassessment of data that takes into account such phylogenetic effects questions previous conclusions about the existence of regional differences in larval traits among damselfishes in the West Pacific and the Caribbean. Received: 19 January 2000 / Accepted: 26 September 2000     

Propagule redirection: habitat availability reduces colonization and increases recruitment in reef fishes

Increased habitat availability or quality can alter production of habitat-dependent organisms in two contrasting ways: (1) by enhancing input of new colonists to the new sites (the Field-of-Dreams Hypothesis); and (2) by drawing colonists away from existing sites (the Propagule Redirection Hypothesis), and thus reducing the deleterious effects of density. We conducted a field experiment on coral reef fishes in Moorea, French Polynesia, to quantify how differing levels of habitat availability (controlling for quality) increased and/or redirected colonizing larval fish. Focal reefs without neighboring reefs received two to four times more settlers than reefs with adjacent habitat, demonstrating that increased habitat redirected larval fish. At the scale of the entire reef array, total colonization increased 1.3-fold in response to a sixfold increase in reef area (and a 2.75-fold increase in adjusted habitat availability). Thus, propagules were both increased and redirected, a result midway between the Field-of-Dreams and Propagule Redirection Hypotheses. A recruitment model using our data and field estimates of density-dependent recruitment predicts that habitat addition increases recruitment primarily by ameliorating the negative effects of competition at existing sites rather than increasing colonization at the new sites per se. Understanding long-term implications of these effects depends upon the interplay among habitat dynamics, population connectivity, colonization dynamics, and density dependence.     

Variability in the chemical defense of the sponge Chondrilla nucula against predatory reef fishes

Chondrilla nucula is a common Caribbean demosponge that grows in a range of habitats, from coral reefs to mangrove swamps. On reefs, C. nucula grows as a thinly encrusting sheet, while in mangrove habitats it surrounds submerged mangrove roots as fleshy, lobate clumps. Previous feeding experiments using predatory reef fish revealed a high degree of variability in the chemical defenses of C. nucula. The present study was undertaken to determine whether a relationship exists between habitat, growth form, and chemical defense of C. nucula. Both laboratory and field feeding-assays of crude extracts confirmed that C. nucula possesses a chemical defense with high intercolony variability, but there was no significant variation in feeding deterrency between reef and mangrove habitats at either geographic location (Bahamas and Florida). Extracts of C. nucula collected during September and October 1994 from the Bahamas were significantly more deterrent than those collected during August 1993, May 1994, and May 1995 from Florida, and extracts of these spring and summer Florida collections were more deterrent than extracts of C. nucula collected in December 1994 and February 1995 in the same locations. There was no evidence that deterrent compounds were concentrated in the surface tissues of the sponge, or that chemical defense could be induced by simulated predation. Laboratory and field assays of the fractionated crude extract revealed that feeding deterrency was confined to the most polar metabolites in the extract. Field transplants were used to determine whether predation influenced the growth form of C. nucula. Uncaged sponges transplanted from the mangrove to the reef were readily consumed by spongivorous reef fishes. Lobate mangrove sponges became thinner after being caged on the reef for 3 mo, but encrusting reef sponges did not become thicker after being caged in the mangroves for the same period of time. Reef sponges that were caged for 3 to 15 mo thickened by only a small amount (<1 mm) compared to uncaged and open-caged (i.e. in cages lacking tops) sponges. Simulated bite marks on both reef and mangrove sponges were repaired at a rapid rate (0.8 to 1.6 mm d⁻¹). Fish predation has an important impact on the distribution and abundance of C. nucula, but the thin growth form common to reef environments may be more the result of hydrodynamics than of grazing by spongivorous fishes. Received: 6 October 1997 / Accepted: 19 March 1998