Artificial reef structures and coral transplantation: fish community responses and effects on coral recruitment in North Sulawesi/Indonesia

Abstract

Artificial reefs alone and in combination with coral transplants (branching Acropora and Pocillopora species) were deployed in 100 m2 plots at three locations (Gangga, Meras and Bunaken) in North Sulawesi/Indonesia in order to study the effects on the associated reef fish community and on coral recruitment dynamics. Control plots covered with coral rubble were monitored for comparison. The study was carried out between May 2005 and July 2007. Coral recruitment was studied using limestone settlement plates, which were exchanged every three months, and responses of the fish community were monitored with monthly underwater visual census. In order to assess the natural fish community in the ambient reef, an additional one-time visual census was conducted in two 100 m2 plots in the natural reef at each location, and the substrate composition in these plots was measured. Fish community data were compared to census data from the experimental plots taken at the same time. Correlation analysis of the natural fish community, substrate composition, and depth at the three locations revealed that the factors depth, coralline algae, branching Acropora spp., foliose coral, soft coral and leather coral best explained the composition of the natural fish community at the three locations. Throughout the experiment, fish abundance, species richness and biomass remained low in the control plots. All three variables, as well as species diversity (H'), were higher in the plots with artificial structures. Coral transplantation led to the highest increase of all variables at Gangga, the shallowest location with low natural hard coral cover (<5 %). The highest fish abundance (796 à ± 36 individuals/100 m2, mean à ± SE), number of species (71 à ± 2 species) and biomass (107 à ± 22 g m-2) were observed in the plot containing structures and corals at Gangga. Integrated over the entire duration of the experiment, fish biomass in the plot with coral transplants at Bunaken did not differ from that in the plot containing only structures. At Gangga, fish abundance, species richness, biomass and species diversity was higher in the plot with coral transplants than in the ambient natural reef, while the values from the natural reef plots were higher than or similar to those in the experimental plots at Meras and Bunaken. Observations on the condition of the ambient reef, species composition and fish recruitment indicate that the effect of coral transplantation on the fish community is strongly dependent on the reef context.Fish community composition differed significantly between the experimental treatments, and between study locations. At all three locations, multivariate dispersion of the fish community and species turnover were lowest in the plot with coral transplants and highest in the control plots. Furthermore, coral transplantation led to increasing Bray-Curtis similarity of the fish community. The observed reduction in variability appears to be indicative of an alleviation of habitat degradation.Coral recruitment at each location was found to be seasonal, and the highest number of recruits observed in a three month interval (n = 2391, approximately 8.7 recruits per 100 cm2) occurred at Bunaken. There was no congruent relationship between coral transplantation and coral recruitment. Results obtained during the last sampling interval indicated that recruitment may be enhanced by coral transplantation over time scales longer than the study period. The species transplanted differed in their performance, with fragments of Pocillopora verrucosa displaying mortality rates between 75 and 80 % within the first year, while Acropora brueggemanni showed about 20 % mortality after one year. Mortality was linked to both disintegration of the transplantation substrate, which resulted in the loss of attachment, and environmental conditions such as sedimentation.