Impact of Point and Non-point Source Pollution on Coral Reef Ecosystems In Mamala Bay, Oahu, Hawaii based on Water Quality Measurements and Benthic Surveys in 1993-1994 (NODC Accession 0001172)

Since the research design of the study was based on gradient analysis (effects of distance or time on data), the selection of stations was exceedingly important. This could have been done arbitrarily at regular and even distances away from point or non-point sources of pollution. However, other factors such as depth and substratum type are not uniformly or regularly distributed in Mamala Bay preventing this simple approach. Rather, an attempt was made to census the entire bay in order to obtain a qualitative understanding of the pattern of distribution of coral reefs and their condition, from Diamond Head to near Barber's Point. A tow survey was conducted in which a diver operated sled was used to census 24 kilometers of coastline at depths between 5 and 15 meters depth. Based on this survey, 16 stations were selected, 5 along the 6.5 m isobath and 11 along the 13 meter depth contour. It turned out that hard bottom areas naturally fell into two categories; low relief (<1.0 m) with very low coral cover, and high relief (2-4 m) with relatively higher coral cover. For this reason, stations representing both categories were selected for study (4 low and 12 high relief stations). The stations spanned a distance of 26 km from Diamond Head to west Ewa Beach. The following data were collected at every station; coral community structure (abundance and species composition), diversity of coral species, salinity, temperature, depth, phosphate, nitrate, ammonium, silicate, suspended solids, turbidity and chlorophyll-a. Sedimentation was measured at eight representative stations. Water quality data were collected on eight days at all stations; 3 collection dates were calm dry days (baseline), 3 were during significant wave events, and 2 were on significant wet (rain event) days. A more detailed description of how significant wave and rainfall events were defined can be found in Gould (1995). In brief, a significant event for both waves and rainfall was defined as the cut-off for the upper 5% of the data. Coral colonies of P. lobata up to 30 years in age were also collected at every station for sclerochronological analysis in the laboratory. Colonies of P. lobata were also collected at 13 m depth in Hanauma Bay and off Sunset Beach to serve as controls. Coral abundance (cover) and species diversity measures were obtained by conducting 50 m line transects at every station. The method involves placement of a 50m line, previously marked at 10 random points, on the bottom. A 0.7 m2 quadrat is then placed over each random point and photographed. In the laboratory, photographic slides are projected on a grid and coral cover for each species present is recorded. Water samples were collected at each station from the deck of a small vessel at the surface and one meter off the bottom with a messenger triggered Niskin open and closing bottle. Stations were relocated using a GPS (Global Positioning System) accurate to within 20 meters. After retrieving Niskin bottles on deck, subsamples for nutrients and chlorophyll-a were filtered through a 0.45 mm GF/F glass microfiber filter and stored in double washed polyethylene bottles. These subsamples were chilled and transported to the University of Hawaii where they were analyzed in the SOEST Analytical Services Laboratory. Subsamples were also taken for suspended solids and turbidity and also analyzed at the ANS at SOEST. A more detailed account of the methods used to measure various water quality constituents can be found in Gould(1995). Sedimentation was measured by deploying sediment traps at fixed locations on the bottom. The traps were attached to a meter tripod frame and placed on the bottom such that the top of the traps were 1.5 meters off the bottom. Traps were designed in order to sample effectively within longshore or wave generated currents between 0-25cm/sec. In order for turbulence caused by this flow to be dissipated within the trap, an aspect ratio (length to height) of 8:1 is required (US GOFS, 1989). Traps were designed with 3 inch PVC tubing cut into 24 inch lengths. This design would be expected to produce a tranquil layer at the bottom of the trap. Six small baffles (1' x 6") were inserted in the top of each trap to further damp turbulence around the mouth. A funnel with a removable test tube (volume calibrated vial) was affixed to the bottom of each trap. This allowed samples to be collected in situ at the end of the deployment periods by a diver without retrieving the whole trap. Deployment times averaged one week. NaCl was placed in the collection vials to serve as a fixative to prevent microbial activity. After collection, sediment samples were chilled and transported to the laboratory where sample wet volume, dry weight and percent organic weight were measured. Dry weight was determined after drying at 57 degrees Celsius for three days. Percent organic matter was determined by the difference in dry weight before and after combustion at 500 degrees Celsius for four hours. Annual growth of Porites lobata colonies was determined using sclerochronology and densitometry (Buddemeier et al, 1974). Using a rock saw, coral colonies were cut into 4 mm cross sections across the central axes of growth. Cross sections were then x-rayed in the laboratory. This procedure produced high quality contact x-radiographic negatives from which positive prints were developed. Annual growth rate was calculated by measuring the linear distance (extension) between annual bands (alternating couplets of porous and dense layers) in the skeleton. X-ray negatives were scanned using a densitometer in order to determine the average density of the skeleton within each annual growth band. Six values of density were recorded for every millimeter of growth on the contact negatives. Hence, the density value obtained for each year of linear growth (about 7-8 mm) was on the order of 40 to 50 readings. Linear extension for each year, multiplied by the average density of the skeleton corresponding to that year, produced estimates of gross calcification in units of Kg CaCO3/m2/yr. In total, 118 colonies of P.Lobata were analyzed in the above manner, representing in total 910 years of growth. One hundred forty five of those years represented control colonies collected at Hanauma Bay and Sunset Beach on the north shore of Oahu. Bioerosion for each colony of P. lobata was calculated by measuring the ratio of excavated area to total area of each cross section. This was accomplished using a hand operated electronic planimeter (Planix Model 7, Tamaya Digital Planimeter), on positive photographic contact prints.