Sediment Processes on the Coral Reefs of Kahoolawe: A Rapid Field Assessment in 1993 (NODC Accession 0000883)

Sediment Sample collection and preparation Bulk Samples. Sediments were collected at 19 sites around the island at the same siteswhere coral community composition was measured. When possible, sediment samples were collected from two depths (3m and 10m). A hand-held plastic scoop was used to collect samples in areas along the transect lines that had sediment accumulation. These transect lines were also used in the fish and coral surveys (Cox et al., 1993). About four scoops (roughly 40-100grams) of unconsolidated surface sediment were taken at each station and placed into plastic bags (Whirlpak brand). These plastic bags were then sealed and stored in wet condition for later analysis by wet sieving(McManus, 1988). Core Samples: Core samples were taken from 5 randomly chosen sites along the northerncoastline of Kahoolawe. Observations from the first site visit in March 1993, suggested that the northern coast was heavily sedimented and an in-depth investigation of sediment loads impacting these areas was undertaken. Core samples were obtained in replicates of two at 50 m intervals, starting at the water line on the beach and moving seaward through the central partof each bay. Polyvinylchloride (PVC) pipes of 18 mm in diameter and 300 mmlength were used as sampling devices at each station. The PVC core samplerswere manually pushed into the sediment and both ends capped tightly in the water before being brought up to the surface. All the core samples werekept frozen prior to analysis. A modified core extraction method was used due to the small diameterof the cores (18 mm). Briefly, the core samples were allowed to thaw and then extracted from the PVC samplers by opening the lower end cap first and then slowly opening the top cap. The sediment cores came out slowly from the tubes after the top cap was removed. In a few cases when the cores were too sticky, a small glass-tip plunger was used to slowly push air through the PVC sampler and extrude the remaining sediment core. The length of the extracted cores was then measured, and the number and size of layers per core were noted. The type, texture and the color ofeach layer per core were also noted. Sample analysis: Size Fraction Determination Sediment samples were wet sieved through standard brass sieves (USA Standard Testing Sieve: A.S.T.M.E.-11 specifications with opening diameters of 500 um and 63 um) and categorized into 3 sizefractions: silt (<63\264m), fine sand (>63 um but <500 um) and coarsesand (>500 um). The wet samples were individually shaken and manually mixed while still in the plastic bags. The homogeneous mixture was thensub sampled (range 30-60 grams) for sieving. The sub samples were washed through the 500-um sieve into the 63 um sieve with filtered fresh water. Washings were done with a hand-held wash bottle and all the washings through the 63-um sieve were collected onto a brass pan. The sediment fraction remaining on each sieve was then washed through pre-weighed filter paper(Whatman # 114) and air-dried for about a week. These filtered samples were then weighed and the total weight of the sample per station from each transect site was determined. The percent by weight of each fraction was then determined by calculating the ratio of the different size fractions to the total sample weight (McManus, 1988). Size fraction determinations were also performed on the core samples. Cores with noticeable stratigraphic layers were cross-sectioned atthe demarcation point and sub samples from each layer (ranging from 20-50 grams) were taken for wet sieving. Chemical Composition Determination: Mineral composition of the sediment samples was determined by X-ray diffraction analysis (XRD) as described by Hardy and Tucker (1988).The analysis was performed by Mr. Clark Sherman of the University of Hawaii's School of Ocean and Earth Science and Technology (SOEST). Samples were ground up using a mortar and pestle and the powdered sediment was then placed on smear slides. These slides were then loaded into anautomatic sample loader and fed to the Scintag Pad V X-ray diffractometer connected to a solid-state Germanium (Ge) detector tuned to Copper (Cu)K radiation. Two runs per sample were performed. First, a general scanwas done to determine the overall mineralogy of the sediments. This run was performed with the machine set at a range of 2 deg 2f to 70 deg 2f witha rate of 5 deg 2f per minute. The second run was performed at a muchnarrower range of 22 deg 2f to 32 deg 2f with a rate of 1 deg 2fper minute to quantitatively determine the carbonate mineralogy of thesediments. Aragonites to calcite ratios were determined using the methods of Sabine (1991) and the mole % Magnesium (Mg) content of the calcite fraction was determined using the procedures established by Bischoff et al. (1983). The possible sources of calcite in each sample can be ascertained based on the mole % Mg content of the calcite fraction. Specifically, the mole % Mgis the amount of Mg atoms substituting the calcium (Ca) atoms at the Ca binding sites in the crystal structure of CaCO3. Calcareous organisms havedistinctive ratios of Mg to Ca substitutions in the calcite produced. Representative groups of organisms and their respective mole % Mg content were based on Table 5.2 of Morse and Mackenzie (1990). A subset of the total sediment samples collected from Kahoolawe was analyzedby x-ray diffraction. These samples were selected from sites that best represented the different and unique regions along the coast. Coral Reef Sampling Methods: At each site, visual inspections of the entire area by skin diving were first conducted. Stations for quantitative surveys were selected as areas with typical coral reef structure. Most sites included two stations, one at 10 m and one at 3 m depth. Coral cover was estimated in 5 contiguous 1 m2 quadrats haphazardly taken along a 25 m transectline at 3m and 10 m isobaths at each site (Jokiel and Maragos, 1978;Maragos and Jokiel, 1986; Jokiel and Tyler, 1992). One observer recorded visual estimates of percent cover of each species within the quadrat and notations of additional species observed outside of the sampled area were recorded. Species identifications were based on Maragos (1977).Visual estimates are more reproducible and more accurate than random-pointsampling for this type of rapid assessment work (Dethier et al., 1993).At several stations (Honokoa, Kaukamoku, Papakaiki, and Waaiki), data on the size distribution of small colonies located on vertical faces at the base of the reef structure were collected. At some sites, colony size was directly measured in bands 1 m by 0.5 m up from the bottom of the reef. At other sites, photographs of the quadrat frame were taken and subsequently analyzed. Colony sizes were estimated using the 10 cm gridof the quadrat frame for scale. An index of relative water motion and potential for impacts from major storm conditions was developed using summarized data on typical current patterns, wind speed records and wave patterns (EnvironmentalImpact Study Corp., 1979; University of Hawaii, Geography Department, 1983).Data Analysis: Coral community diversity was calculated using the modified (Loya, 1972) Shannon-Weaver diversity index (H?c = - ? pi ln pi) of the mean percent cover each species on the transect lines. Similarity of communities from the 33 sites sampled was assessed using a modified Sorensen Similarity index, after transformation of the data (angular transformation of percent data and square root transformation of average counts): Iab = ? 2Ma (Ma + Mb)-1;where Iab is the index for two sites (a and b) for each species(i = 1, 2, ...S); Ma is the lower cover or density for the ith species in the two site pair and Mb is the higher cover or density for that species. Cluster analysis was performed on the resulting matrix of similarity values using SAS procedures (SAS 1988).