Estimation of phosphorus flux in a regulated channel

During a nutrient loading study in the Okanagan River Basin, British Columbia, the regression relationship between phosphorus concentration and the Okanagan River discharge was shown to be not statistically significant. Phosphorus flux could, therefore, not be determined by conventional methods which rely on the concentration-discharge relationship. The reason for failure of conventional methods is the regulated nature of the Okanagan River flow. Two sampling strategies for chemical analysis of water samples were applied: (a) simultaneous, based on manual collection of sample replicates; and (b) sequential, using an automatic device to obtain a series of point concentration measurements. These sampling strategies provided a basis for nutrient flux estimation by the following methods: (a) partial load method, resulting from simultaneous sampling; and (b) flow interval method, based on sequential sampling. Load estimates and their confidence limits derived from the partial load method were compared with loads obtained by the flow interval method. There was a high level of correlation between the load estimates. Annual trends are more appropriately investigated by the flow interval method, cross-sectional load variability and confidence limits should be assessed by the partial load method.