Thesis (Ph.D.)--Victoria University of Wellington, 2008. / Includes bibliographical references.

Throughout the Pacific, many species of echinoderms and molluscs have cultural value and are harvested extensively in subsistence fisheries. Many of these species are sedentary and often associated with distinct reef-top habitats. Despite the significance of reef habitats and their fauna for fisheries and biodiversity etc, little information has been available on the distribution of habitats and their influence on the reef-top fauna in the Cook Islands. This thesis developed a novel approach to assess the status of the shallow-water reef-tops of Rarotonga, Cook Islands, to provide critical information to fisheries and conservation managers. The approach used remote sensing (aerial photography with ground truthing) to map the spatial arrangement and extent of the entire reef-top habitats accurately, and historical wind data and coastline shape to determine the windward and leeward sides of the island. The benthic habitat maps and degree of wind exposure were used to design and undertake a stratified sampling programme to assess the distribution and abundance of the epibenthic macroinvertebrate fauna of the reef-top. I quantified the distribution and abundance of the epibenthic macroinvertebrates and how they varied with habitat, assessed the effectiveness of a traditional ra�ui (marine protected area) for conserving stocks of Trochus niloticus and other invertebrates, and investigated the reproductive biology and impacts of traditional gonad harvesting on Holothuria leucospilota. There were four major habitat types (rubble/rock, sand/coral matrix, algal rim and sand) identified, the most extensive being rubble/rock (45%) and sand/coral matrix (35%). The degree of exposure to winds was found to correlate with the reef development and habitat distribution. The assemblage composition of each major habitat type differed significantly from every other habitat. The rubble/rock habitat had the greatest substratum heterogeneity and structural complexity, and the highest number of species and individuals. The overall abundance of the fauna was dominated by holothurians (68%) and echinoids (30%), while Trochus niloticus and Tridacna maxima accounted for the remaining 2% of the total invertebrate assemblage. Clear habitat partitioning was also found for adult and juvenile Trochus niloticus and Tridacna maxima. In the traditional fishery for Holothuria leucospilota, the mature gonads of males are harvested by making an incision in the body-wall of the animal, removing the gonads and then returning the animal to the reef to allow regeneration. Monthly collections of H. leucospilota were used to describe the reproductive biology of this species. Gametogenesis and spawning were synchronous between the sexes and spawning occurred annually during summer, when water temperature and photoperiod were at their highest. Although the incision in the body-wall and gonad removal had no impact on the survival of H.leucospilota in experimental cages, their body weight, and general sheltering and feeding behaviors were affected. Gonads took at least 41 days to start regenerating, suggesting a considerable delay in the spawning of fished individuals. In 1998, five Rarotongan communities re-introduced the traditional ra�ui system of resource management, prohibiting all fishing and gathering from their reefs. The performance of the Nikao ra�ui, which had been put in place to allow trochus stocks to increase, was investigated. Comparisons of macroinvertebrate assemblage composition and species density were made between three fishing treatments, i.e. fished areas adjacent to the ra�ui, within the ra�ui after two years of protection, and in the ra�ui after it had been lifted for three weeks to allow a commercial trochus harvest. Analysis of variance on the count data for the twelve most abundant species, and non-metric multi-dimensional scaling indicated that there were no differences in the microhabitat or the invertebrate assemblage composition between the three fishing treatments. However, there were significant differences between the rubble/rock and sand/coral matrix habitat types. The results on the effectiveness of the Nikao ra�ui are equivocal, due to the small sample size, and the variability between samples which was highlighted by the wide confidence intervals. This study highlights the importance of habitat to the macroinvertebrate fauna of the reef-top and the need for accurate habitat maps to increase the cost-effectiveness of future resource surveys, to provide information to management, and for the design of Marine Protected Areas. The mapping and survey methods must be reliable and repeatable in terms of the limitations of time, and the availability of expertise, funding and resources. The results provide important information for fisheries and conservation managers of Rarotonga and other Pacific Islands to better design rigorous sampling programmes for monitoring the status of reef-top resources, and for evaluating and planning Marine Protected Areas.

Throughout the Pacific, many species of echinoderms and molluscs have cultural value and are harvested extensively in subsistence fisheries. Many of these species are sedentary and often associated with distinct reef-top habitats. Despite the significance of reef habitats and their fauna for fisheries and biodiversity etc, little information has been available on the distribution of habitats and their influence on the reef-top fauna in the Cook Islands. This thesis developed a novel approach to assess the status of the shallow-water reef-tops of Rarotonga, Cook Islands, to provide critical information to fisheries and conservation managers. The approach used remote sensing (aerial photography with ground truthing) to map the spatial arrangement and extent of the entire reef-top habitats accurately, and historical wind data and coastline shape to determine the windward and leeward sides of the island. The benthic habitat maps and degree of wind exposure were used to design and undertake a stratified sampling programme to assess the distribution and abundance of the epibenthic macroinvertebrate fauna of the reef-top. I quantified the distribution and abundance of the epibenthic macroinvertebrates and how they varied with habitat, assessed the effectiveness of a traditional ra�ui (marine protected area) for conserving stocks of Trochus niloticus and other invertebrates, and investigated the reproductive biology and impacts of traditional gonad harvesting on Holothuria leucospilota. There were four major habitat types (rubble/rock, sand/coral matrix, algal rim and sand) identified, the most extensive being rubble/rock (45%) and sand/coral matrix (35%). The degree of exposure to winds was found to correlate with the reef development and habitat distribution. The assemblage composition of each major habitat type differed significantly from every other habitat. The rubble/rock habitat had the greatest substratum heterogeneity and structural complexity, and the highest number of species and individuals. The overall abundance of the fauna was dominated by holothurians (68%) and echinoids (30%), while Trochus niloticus and Tridacna maxima accounted for the remaining 2% of the total invertebrate assemblage. Clear habitat partitioning was also found for adult and juvenile Trochus niloticus and Tridacna maxima. In the traditional fishery for Holothuria leucospilota, the mature gonads of males are harvested by making an incision in the body-wall of the animal, removing the gonads and then returning the animal to the reef to allow regeneration. Monthly collections of H. leucospilota were used to describe the reproductive biology of this species. Gametogenesis and spawning were synchronous between the sexes and spawning occurred annually during summer, when water temperature and photoperiod were at their highest. Although the incision in the body-wall and gonad removal had no impact on the survival of H.leucospilota in experimental cages, their body weight, and general sheltering and feeding behaviors were affected. Gonads took at least 41 days to start regenerating, suggesting a considerable delay in the spawning of fished individuals. In 1998, five Rarotongan communities re-introduced the traditional ra�ui system of resource management, prohibiting all fishing and gathering from their reefs. The performance of the Nikao ra�ui, which had been put in place to allow trochus stocks to increase, was investigated. Comparisons of macroinvertebrate assemblage composition and species density were made between three fishing treatments, i.e. fished areas adjacent to the ra�ui, within the ra�ui after two years of protection, and in the ra�ui after it had been lifted for three weeks to allow a commercial trochus harvest. Analysis of variance on the count data for the twelve most abundant species, and non-metric multi-dimensional scaling indicated that there were no differences in the microhabitat or the invertebrate assemblage composition between the three fishing treatments. However, there were significant differences between the rubble/rock and sand/coral matrix habitat types. The results on the effectiveness of the Nikao ra�ui are equivocal, due to the small sample size, and the variability between samples which was highlighted by the wide confidence intervals. This study highlights the importance of habitat to the macroinvertebrate fauna of the reef-top and the need for accurate habitat maps to increase the cost-effectiveness of future resource surveys, to provide information to management, and for the design of Marine Protected Areas. The mapping and survey methods must be reliable and repeatable in terms of the limitations of time, and the availability of expertise, funding and resources. The results provide important information for fisheries and conservation managers of Rarotonga and other Pacific Islands to better design rigorous sampling programmes for monitoring the status of reef-top resources, and for evaluating and planning Marine Protected Areas.