UNIVERSAL PARADIGM REFUTABILITY: HIGH-DIMENSIONAL SPACE, SCALE, MISLEADING METRICS, UNCERTAIN GOALS

Increasingly, geomorphologists and aquatic habitat specialists are asked to answer specific and general questions of how management activities affect headwater streams. It seems every answer to such questions can be refuted with literature examples. This is not surprising. The morphology and behavior of channels is affected by eight inter-related major factors: 1) climate, 2) geology and soils, 3) topography and network topology, 4) upland and riparian vegetation, 5) sediment loading, 6) flows, 7) large woody debris loading, and 8) time and type of last major disturbance. Assuming these are all continuous variables (which they aren't), a hypothetical cross-lansdscape study that varied each factor at three levels and assessed all combinations would require 3 to the eighth, or 6,561 streams. Such a study would be limited to first- or second-order headwater basins where some control over each variable could be achieved, thus routing effects and downstream issues could not be addressed. Therefore, a complete and robust data set from which to develop "universal" models of headwater stream response is not achievable. Developing answers to management questions requires extrapolating existing data along one or more of the above conceptual variable axes. Channel classification schemes have been used to lump streams into expected behavior classes, but experience suggests classification is only useful when limited to specific terrains and types of management questions. In the Pacific Northwest, management of larger streams has been simplified by an implicit acceptance of Coho salmon as the reference species, but no such unifying habitat paradigm exists for headwater streams. Habitat needs and population goals for creatures inhabiting headwater streams are poorly understood and often competing. This uncertainty leads to vague management goals. Application of stream geomorphology concepts to headwater stream management questions is complicated by temporal and spatial scale issues. Aquatic invertebrates have lifespans of single years, stream-dwelling amphibians have lifespans of 10-20 years, commercial forest stands have rotation lengths of 40-70 years, and debris flows have recurrence intervals in the range of centuries. The effects of riparian buffer strategies, for example, must be considered over all these scales.