Conclusion

For the water-quality constituents in the small Tualatin River basin streams presented in this report, automatic samplers were capable of collecting unbiased and representative stream samples. No major cross-contamination issues from sample to sample were observed. Sufficient care certainly must be exercised to keep the samples on ice and deliver them to the laboratory in a timely manner, but when used with a good quality assurance plan, autosamplers were a useful component of a sampling plan. Used in conjunction with water-quality monitors that can trigger sampling, autosamplers may become an invaluable component of future monitoring or sampling schemes.

The use of continuously measured field parameters to predict constituents of regulatory interest in streams could be helpful for understanding the effect of management strategies on water quality in the Tualatin River basin. Results of this study indicate that the potential to develop predictive relations is good. Additional data, including more water-quality data over a broader range of conditions along with co-located discharge monitoring, would increase the predictive ability of the resulting regressions.

These sorts of predictive regression equations may be used to quantify peak concentrations or annual loads of sediment or phosphorus moving through the system. The equations may be useful in suggesting certain types of occurrences during storms or other conditions that merit further study, thus aiding in our understanding of the water-quality dynamics of these streams. This method of using continuous water-quality monitors to predict the concentrations of unmeasured water-quality constituents is an underutilized technique that deserves more attention in the future.

This study was a reconnaissance effort to determine the transferability of techniques used elsewhere to tributaries of the Tualatin River basin. These techniques for predicting the unmeasured concentrations of selected water quality constituents from continuously monitored surrogates require site-specific correlations and relatively consistent upstream conditions. If successful, this effort could provide a foundation for development of more detailed and accurate correlations at the study locations and elsewhere, and for their use in near‑real time, potentially allowing evaluation of the efficacy of land-use and other management decisions.

First posted June 18, 2010