EFFECTS OF REEF SIZE ON COLONIZATION AND ASSEMBLAGE STRUCTURE OF FISHES AT ARTIFICIAL REEFS OFF SOUTHEASTERN FLORIDA, USA

Fifty standard concrete modules were deployed on a sand bottom to make 16 replicated artificial reefs of 1 to 8 modules.reef(-1). Fish assem blages were monitored for species composition, abundance, and fish siz e, and compared to two natural control sites. We censused 127 species (107,168 fishes) from artificial reefs, 93 species (16,495 fishes) on natural control reefs, and 17 species (1,040 fishes) on sand bottom fr om July 1987 to March 1989. Artificial reefs supported a diverse, abun dant, and dynamic assemblage of fishes that were a mixture of species found in surrounding sand and natural reef habitats. Colonization (num ber of species, individuals, and biomass) was very rapid. Abundance va ried seasonally with recruitment episodes tending to occur in the spri ng and summer followed by losses for the remainder of the year. Assemb lages were quite variable on and between similar sized reefs. Fish and biomass densities were higher at artificial reefs than on sand and na tural reefs. Resident fish biomass varied less than resident fish numb ers, because individual growth compensated for mortality after recruit ment episodes. Reef size significantly influenced total numbers of spe cies, individuals, and biomass. Smaller reefs had greater fish density while larger reefs had higher biomass density from larger but fewer, individuals. Multiple small reefs supported more individuals and more species than one large reef of equal material. Fishes recruited by lar val settlement accounted far 36% of the total resident abundance but o nly 2% of total biomass. As reef size increased, older juvenile or adu lt colonists comprised a greater percentage of total biomass (94% to 9 9%). Assemblage importance percentages (based on abundance, biomass, a nd frequency) were divided between residents (64%), visitors (20%), an d transients (16%). Economically important species comprised 61% of th e biomass and 55% of the individuals, among which settlers accounted f or 94.3% of individuals but only 5.7% of their total biomass. The most highly valued species were visitors or residents that utilized the re efs after first settling elsewhere. These results showed that data on artificial reef assemblages based solely on the abundance of resident species are biased. Data on visitors, transients, frequency-of-occurre nce, and biomass are important in evaluating bias. Results provided a partial test and support for a model predicting the importance of attr action over production for artificial reefs located in areas with high reef availability.