• survey sampling design;
  • biological indicators;
  • monitoring;
  • status and trend estimates;
  • aquatic life use;
  • lake trophic state;
  • coho salmon;
  • aquatic ecosystems;
  • water quality

ABSTRACT: Despite spending $115 billion per year on environmental actions in the United States, we have only a limited ability to describe the effectiveness of these expenditures. Moreover, after decades of such investments, we cannot accurately describe status and trends in the nation's aquatic ecosystems or even those in specific regions. Why? This situation has arisen in part because we have excluded the fundamental principles of probability designs that are widely used in other fields and we have often ignored direct measures of biota, the subjects of greatest concern. To demonstrate the results of ignoring these powerful statistical and biological tools, we present four case studies. These studies compare estimates of aquatic resource status derived from using (1) a probability-based study design, often with biological measures of condition; and (2) a nonstatistical study design, often using chemical surrogates. In three of the four cases, the results derived from the nonstatistical perspective underestimate the degree of biological degradation.