Environmental Observations in support of Physical and Biological Modeling of Aquaculture Sites

James D. Irish*, David W. Fredriksson, Dale A. Kiefer, Jack Rensel, and Frank OBrien

University of New Hampshire, Jere Chase Ocean Engineering Laboratory
24 Colovos Rd., Durham, NH 03824
jirish@whoi.edu Optimum design of an aquaculture operation depends on environmental conditions at the site, and requires a combination of observations, hydrodynamic modeling, and biological modeling. To develop and demonstrate this approach, a study was conducted for an aquaculture site in the Western Gulf of Maine about 12 km off the coast of New Hampshire in a depth of 50 m of water. The University of New Hampshire has been conducting open ocean aquaculture research at this site for nearly 10 years (http://ooa.unh.edu/). The research included measuring physical, chemical, and biological parameters important for aquaculture operations. Moored time series were made for about 9 months a year, and discrete water samples were taken monthly from spring to fall at the site starting in 1999 and continuing into 2008. Also, the UNH Coastal Ocean Observing Center has been monitoring the ecosystem in the region with monthly cruises. Their results supply additional information on the biological and chemical properties at the site. All these data are used for initializing and forcing the hydrodynamic model (ADvance CIRCulation = ADCIRC) and biological model (AquaModel).

The moored instrumentation measured water current profiles, temperature and salinity at the surface, mid-water (the 22 m depth of the fish cages), and near the bottom. Also, moored mid-water and bottom observations were made of dissolved oxygen, chlorophyll-a and turbidity. These observations provided information on oscillatory tidal currents (0.1 m/s oriented perpendicular to the coast) that help disperse waste products. Weather forced currents that advect material off site were quite variable with maxima of 0.5 m/s oriented along shore. These current observations were used to validate the ADCIRC tidal model (whose results were used to force AquaModel), and the currents also directly forced AquaModel.

The temperature and salinity follow a yearly cycle of warming in the spring-summer, and cooling in the fall-winter. Fall storms cause full water column mixing from early winter through early spring when thermal stratification is established, preventing vertical mixing. Temperatures cool to 2 to 4C in the winter and bottom waters warm to 10C in the summer while surface waters reach 22C. River runoff freshens the water in spring into summer to about 31 PSU with storm runoff spikes in surface waters reaching the low 20s. With the winter mixing, the water column salinity increases to above 33 PSU. Oxygen profiles are saturated or supersaturated from surface to bottom in the winter, but with the onset of spring stratification, oxygen below the pycnocline steadily decreases from fully saturated to about 70% saturated by the end of summer. The mid-water oxygen remained saturated into the summer, then also decrease into the 70% range by the end of summer. Suspended sediment and chlorophyll time series and discrete sample show river runoff sediments, spring and fall chlorophyll blooms. Red-tide blooms are also observed consuming oxygen. These water quality results were summarized into weekly averages and used to initialize and drive AquaModel studies of the site.