| Most of Ontario's aquaculture production is from rainbow trout cage culture in Lake Huron. In this environment, it is difficult to distinguish farm-loaded nutrients from other sources. Extensive near-field (<30m distant) spatial and temporal variation of nutrient concentrations can also affect the accuracy of discrete samples to infer mean conditions. An improved understanding of nutrient loading dynamics and measurement methods is required to effectively manage and regulate these farms. To address these issues, diel (24 hour) fluctuations of soluble effluent from tank cultured rainbow trout (Oncorhynchus mykiss), near-field nutrient loading dynamics at a cage-site in Lake Huron, and a model to estimate cage-site outflow of phosphorous concentrations were investigated.; Diel effluent concentrations of soluble phosphorus (SP), total ammonia nitrogen (TAN) and dissolved oxygen (DO) from rainbow trout tank water varied in a non-linear (p < 0.05) pattern. The coefficient of variation for diel SP concentrations in tank water was 16% at 8°C and 23% at 15°C.; At the commercial field site, mean water quality and current dynamics from 6 multi-day periods were measured. Near-field, period mean concentrations of total phosphorus (TP) ranged from 4.5 +/- 0.9 to 11.7 +/- 6.7 mug L-1. The distribution of these concentrations was non-normal with standard deviation increasing with the means. Variation of TP, pH and DO increased with exposure to the source of nutrient loading, from the inflow, outflow to the site centre. These water quality parameters also varied together, the strength of their correlations also increasing with exposure.; For four of the sampling periods (totalling 14 days), site mid-depth TP concentrations, 30m distant in the site outflow, were modelled by coupling mass balance estimates of non-settleable TP with a one-dimensional flushing equation. Despite several uncertainties with current dynamics, particulate advection, fish distribution and feed conversion, the model fit reasonably well with empirical measurements (r2 = 0.82), although estimates were approximately 20% less.; This study's findings are potentially useful for assisting in the interpretation of near-field TP measurements at cage farms in the Great Lakes. Monitoring and modelling recommendations are made. |
