ALBERT

Notes: Abstract The slicks of tidally induced fronts and Langmuir circulations were studied near Bowden Reef, Great Barrier Reef, in December 1987. There were two components to the study: (1) a description of physical oceanography adjacent to the reef; (2) sampling for planktonic organisms, designs being stratified according to hydrology. Tidal fronts extended 1 to 2 km from the reef outside the lagoon. Fish of a variety of developmental forms and zooplankton were most abundant in slicks of fronts. Sixteen to 81% of fish that were captured were presettlement reef-fishes. There was an unclear relationship between the presence of slicks of Langmuir circulations and abundance of fish in the lagoon. This relationship was largely because well developed atherinids and recently hatched pomacentrids showed no predictable relationship with the presence of windrows. Jellyfish,Aurelia aurita, were found in extremely high concentrations in the slicks of Langmuir circulations; small carangids were associated with jellyfish. Movements of fronts were influenced by the tide and wind. In some conditions we observed fronts to change position as the tide reversed direction. Fronts that were orientated offshore (1 to 2 km), were observed to rotate and align with the reef. Because of a “halo” of oceanographic features such as tidal fronts, the reef may be a larger target for presettlement fishes than its topography would suggest. It is argued that some organisms are advected into slicks, while others respond to high concentrations of zooplanktonic food and remain in slicks. Localised oceanographic features may also affect the settlement patterns of fishes on reefs.

Notes: Abstract Estimates of daily feeding rates were obtained for two groups of herbivorous labroid fishes, one confined to cold water and the other to tropical reef environments. These were the family Odacidae, represented by Odax pullus from New Zealand waters, (Goat Island Bay: Latitude 36° South; on the northeastern coast of New Zealand) and the family Scaridae, represented by Scarus rivulatus, S. schlegeli and S. sordidus from the northern Great Barrier Reef (Lizard Island; a mid-shelf reef at 14° South latitude). Observations on the odacid were made in 1984 and in 1992, and on the scarids in 1984 and 1988. O. pullus displayed a diurnal feeding pattern in which the rates (expressed as bites min-1) are greatest early in the day. The mean combined feeding rate for three size groups (juveniles, subadults and adults) peaked (average of 2.9 bites) from 06.00 to 08.00 hrs and declined fourfold to a combined average of 0.7 bites min-1 by midday. The greatest mean feeding rate recorded was 3.7 bites min-1, with an overall mean of 1.8 bites min-1. For subadults and adults there were consistent trends in feeding, with subadults feeding at a greater rate than adults and both groups displaying a decline in feeding rate during the day. The change in feeding rate with time of day was statistically significant in both groups. The pattern for juvenile O. pullus was different from that in the two larger size groups in that juveniles did not show a uniform decline in feeding with time of day. For scarids, the daily feeding rate varied by site, but the pattern was similar for all species, characterised by initial low rates increasing to higher but variable levels by midday. The influence of both site of feeding and time of day on feeding rate was confirmed by analysis. The overall mean values for each species were 20.1 bites min-1 for S. rivulatus, 19.7 bites min-1 for S. schlegeli and 14.9 bites min-1 for S. sordidus. For scarids, the peak feeding rates varied from 19.3 to 32.8 bites min-1, with overall rates from 14.9 to 21.1 bites min-1. Estimates of activity and movement patterns during feeding were obtained for O. pullus. Distance moved per unit time was highly variable, 0.1 to 47.5 m min-1, with a mean of 8.5 m min-1 (SD=9.9). Trends in movement among sexes and size classes were obscured by the variable movement patterns of individual fishes.

Notes: Abstract The diet of the temperate marine herbivorous fish Odax pullus (Pisces: Odacidae) was examined using gut-content analysis followed by principal-component analysis and analysis of variance. Fish were collected near Leigh, on the northeast coast of New Zealand, from February 1983 to September 1984. The data were categorised by size of fish, season, and state of tide. A major ontogenetic shift in diet was observed. Juveniles fed on animal material (crustaceans and gastropods), epiphytic rhodophytes, and some phaeophyte material. Adults fed almost exclusively on two phaeophyte taxa, the laminarian Ecklonia radiata and the fucoid Carpophyllum spp. Dietary selection was evident at the level of particular plant components; the diet of larger fish in spring samples was dominated by fucoid reproductive receptacles. There was no direct evidence of a tidal influence on diet composition of O. pullus, although the relative proportions of E. radiata and Carpophyllum spp. ingested by adult fish appeared to vary with tidal state. The volume of gut contents varied seasonally for all size classes, although the timing of peak annual food intake varied among size classes. Tidal state did not influence the volume of the gut contents. The ontogenetic and seasonal trends evident in the diet of O. pullus appeared to be related to a number of factors including relative gut length, gonad development, ontogenetic changes in feeding anatomy and seasonal changes in algal composition.

Notes: Abstract   Specimens of the nominally herbivorous, closely-related, fish genera Girella and Kyphosus were collected from Australian waters in 1994 and 1995. The diet of three Girella species (G. cyanea, G. elevata, and G. tricuspidata) consisted mainly of chlorophytes and rhodophytes, with an animal component of␣15.9 ± 4.2% in G. tricuspidata. The diet of four species of Kyphosus (K. bigibbus, K. cinerascens, K. sydneyanus, and K. vaigiensis) included phaeophytes, chlorophytes and rhodophytes, and almost no animal material. Concentration of total short-chain fatty acids in the posterior intestine was 〈11.4 mM in the Girella spp. and 〉39.2 mM in the Kyphosus spp. These results suggest that microbial fermentation plays a role in algal digestion in Kyphosus spp., but not in Girella spp. Girellids and kyphosids appear to function quite differently as herbivores. Girellids should be considered as omnivores that complement readily-available energy from algae with protein from invertebrates. Kyphosids appear to be strict herbivores that can derive adequate nutrition from algae poor in easily assimilable energy, through microbial fermentation in the hindgut.

Notes: Abstract The aim of the study was to provide comparable estimates of abundance of herbivorous reef fishes at temperate and tropical localities using a standardized methodology. Faunas of herbivorous fish were sampled on the rocky reefs of temperate northern New Zealand and on the coral reefs of the northern Great Barrier Reef (GBR), Australia, and the San Blas Archipelago in the Caribbean. A pilot study established the most appropriate habitat setting and the scale and magnitude of replication for the sampling program in temperate waters. Herbivorous fishes, including members of families endemic to the southern hemisphere (Odacidae and Aplodactylidae), were most abundant in turbulent, shallow water (0 to 6 m) and had patchy distributions within this habitat. A hierarchical sampling program using 10-min transect counts within the 0 to 6 m depth stratum examined abundance patterns at a range of spatial scales including mainland and island coasts, localities separated by up to 100 km and sites separated by up to 10 km. This program identified a characteristic fauna of seven species of herbivorous fishes with mean total abundances ranging from 23 to 30 individuals per 10-min transect. Species composition of the fauna varied between islands and coasts. A similar methodology was used to sample the major families of herbivorous fish in a number of sites in each of the tropical regions. These sampling programs revealed a fauna dominated by acanthurids and scarids in both the GBR and Caribbean localities. Estimates of abundance from these regions were similar, with a mean of 108 individuals recorded on the GBR and 129 per 10-min transect in the Caribbean. Species richness varied between each region, with 44 taxa recorded from the GBR and 11 from the Caribbean. Abundances of temperate water herbivores in New Zealand were found to be 75 to 80% lower than those recorded from shallow water habitats sampled on coral reefs. This was not related to species richness, since both New Zealand and the Caribbean locality had patterns of low richness. We suggest that the differences in abundance found by our study between temperate and tropical regions are not restricted to herbivorous fishes, but are representative of general latitudinal trends in reef fish faunas.

Notes: Abstract Abundances of the surgeon fish Acanthurus lineatus (Linnaeus) within a single reef-system were estimated using a hierarchical sampling design during 1982. Additional sampling was carried out during 1983 and 1984 at a restricted number of sites. This species is aggressive toward other herbivorous fishes and is strongly site-attached. On the study reef (Lizard Island; 14°40′S; 145°28′E), A. lineatus was restricted to reef-crest sites below mean low water, mainly on reefs not directly exposed to prevailing winds. Within this reef zone and depth-defined stratum, A. lineatus was distributed heterogenously with high abundance, (approximately 14 fish per 300 m2) at a single sample-locality covering 600 m of reef crest. Abundances of herbivorous species (members of the families Acanthuridae and Scaridae) at other localities did not correlate with abundance patterns of A. lineatus. Subsampling within study localities revealed considerable heterogeneity in the abundance patterns of herbivorous fishes, especially within the area of high A. lineatus density. A detailed behavioural study of interactions among herbivorous fishes at two adjacent sites within the locality of high A. lineatus abundances revealed a complex pattern of site-general and sitespecific features. A. lineatus excluded smaller scarids from its feeding territories at one site, but not at another. Scarids attaining large size (〉350 mm standard length) were present at one site and consistently fed within A. lineatus territories; large scarids were rare at the second site, even though the distances involved were small. In addition, the small surgeon fish A. nigrofuscus, a consistent target for A. lineatus aggression, was rare at one site but moderately common at the other. Finally, the abundant surgeon fish Ctenochaetus striatus was present at high densities at both sites and fed within A. lineatus territories. This species was not attacked by A. lineatus nor did it attack other herbivorous fishes within the vicinity. Small but consistent differences in reef structure were detected at each site. Local-scale heterogeneity in these interactions makes it difficult to develop generalizations concerning the role of territorial herbivores such as A. lineatus within reef systems. We hypothesize that very local differences in the within-habitat component of acanthurid and scarid abundances and distributions may reflect site-associated variability in recruitment patterns, post-recruitment mortality or behaviour that is independent of A. lineatus activities. Replicated removal experiments which include explicit tests for local site-effects and better descriptions of recruitment in larger herbivorous fishes are required before these interpretations can be evaluated.

Notes: Abstract The occurrence of unusual symbiotic microorganisms was examined in the intestines of a range of fish from the Great Barrier Reef, Australia. The fish taxa examined included 26 species of the family Acanthuridae, as well as representatives of phylogenetically related and herbivorous taxa. The microorganisms, referred to as protists, were only found in herbivorous and detritivorous members of the Acanthuridae. Protists were not found in planktivorous acanthurids, nor in any members of the families Kyphosidae, Pomacentridae, Scaridae, Zanclidae, Siganidae and Bleniidae we examined. In addition, protists were absent from the herbivorous acanthurids A. xanthopterus and A. nigricans. A range of protist forms, characterized by differences in size (8 to 417 μm), shape and mode of cell division (daughter-cell production and binary fission), was observed. The occurrence of these forms appeared to be correlated with host feedingecology. Large forms (〉100 μm) of the protists were only found in acanthurids which fed over hard-reef substrata. Smaller forms were found in sand-grazing and detritivorous species. One of the protist forms appears identical to protists previously reported from Red Sea acanthurids.

Notes: Abstract The distribution patterns of presettlement reef fish and how they were influenced by the proximity of reefs were investigated off the coast of Northland, New Zealand, from 1981–1986. We used ichthyoplankton tows and visual counts of fish. Distributions of presettlement fish of some species were influenced by the proximity of reefs, regardless of whether reefs were on the coast of the mainland or islands across the shelf. Presettlement fish of families that lay demersal eggs were most abundant near reefs: Gobiescocidae, Acanthoclinidae, Tripterygiidae, Eleotridae, and Gobiidae. The distribution of presettlement sparids, mullids (pelagic eggs), and blenniids and monacanthids (demersal eggs) was not determined in a predictable way by the proximity of reefs. High-frequency sampling over three days suggested that patches of presettlement sparids of 1 to 2 km in dimension may move quickly through a study area. High abundance of presettlement gobiescocids and tripterygiids were found in 0 to 2 m of water over rocky reefs at high and low tides. Presettlement eleotrids were associated with reefs in deeper water (3 to 20 m) and in some habitats with aggregations of mysids. The lack of general patterns of distribution for presettlement reef fish suggests that modelling patterns of drift of these fish as a single group is inappropriate; this concurs with evidence from tropical waters.

Notes: Abstract Cheilodactylus spectabilis (Hutton) is common over shallow reefs in north-eastern New Zealand. Replicated transect between-area differences in density and in size frequency. C. spectabilis is not nocturnally active. The between-area differences in density and size frequency remained constant over daylight hours, demonstrating that C. spectabilis does not undergo any systematic feeding migrations. Topographic complexity was shown to have a significant positive influence on fish density. Mean size was related to water depth at 5 localities investigated, with small (〈200 mm standard length, SL) individuals being restricted to shallow water. All sizes of C. spectabilis examined had fed on small invertebrates, with gammarid amphipods predominating. No evidence of size-related differentiation in feeding patterns was observed, although small individuals spent significantly more time feeding than large ones. Small (〈200 mm SL) individuals occupied feeding areas and shelter sites from which they excluded other small C. spectabilis. Large individuals showed no site-associated aggression, covered a greater area during daily movements and overlapped with other large and small C. spectabilis while feeding. No evidence of dawn or dusk peaks of feeding activity or movement was collected.

Notes: Populations of the estuarine glass fish, Ambassis vachelli Richardson, were used to study the relationship between somatic growth and widths of daily increments in the sagittal otoliths. Variations in the somatic growth of A. vachelli were induced by a series of experimental feeding regimes which included feeding to satiation with two food sources and a starvation treatment. After 33 days of exposure to the experimental feeding regimes significant differences in the mean wet weight of individuals amongst the feeding treatments were recorded. Fishes subject to a starvation treatment showed a significant reduction in wet weight compared to the pretreatment population and the two experimental feeding regimes. No changes in lengths of fishes were recorded.Validation techniques revealed that daily increments were laid down in the sagittal and asteriscal otoliths. Estimates of ring widths from samples of sagittal otoliths revealed significant treatment effects. The increments of fishes from the starvation treatment showed a significant decline in mean increment width relative to the feeding treatments. This difference was detected only after a 15 day period of experimental feeding. It is suggested that the gradual decline in increment width reflects the exhaustion of readily mobilized energy reserves.