Exploring the Benefits of Paired Watersheds for Detecting Cumulative Effects

Adverse impacts of human activities on water quality are usually due to the cumulative effect of several different land uses and management practices. However, cumulative effects are often difficult to detect within a reasonable time frame and monitoring budget because of the high natural variability in the systems of interest. One partial remedy for this problem is the use of predictor variables, which are defined as hydrologic measurements related to the water quality variable of interest. Such predictive variables can help model or explain some of the natural variability and thereby facilitate the detection of change. The paired-watershed approach, implemented statistically via linear regression, can detect smaller changes than a single-watershed approach when there is a strong correlation between the variable of interest and a predictor variable. Since many monitoring studies have a poor likelihood of detecting a cumulative effect, the statistical power of detecting change should bean important consideration.