AN EVALUATION OF PHYSICAL STREAM HABITAT ATTRIBUTES USED TO MONITOR STREAMS¹

Abstract

The last few decades have seen an increased reliance on the use of stream attributes to monitor stream conditions. The use of stream attributes has been criticized because of variation in how observers evaluate them, inconsistent protocol application, lack of consistent training, and the difficulty in using them to detect change caused by management activity. In this paper, we evaluate the effect of environmental heterogeneity and observer variation on the use of physical stream attributes as monitoring tools. For most stream habitat attributes evaluated, difference among streams accounted for greater than 80 percent of the total survey variation. To minimize the effect that variation among streams has on evaluating stream conditions, it may be necessary to design survey protocols and analysis that include stratification, permanent sites, and/or analysis of covariance. Although total variation was primarily due to differences among streams, observers also differed in their evaluation of stream attributes. This study suggests that if trained observers conducting a study that is designed to account for environmental heterogeneity can objectively evaluate defined stream attributes, results should prove valuable in monitoring differences in reach scale stream conditions. The failure to address any of these factors will likely lead to the failure of stream attributes as effective monitoring tools.

Animal Waste, Water Quality and Human Health

“…A wide range of methods are available and provide a tangible resource for decision making, management and restoration plans (Johnson et al 2001, Roper et al 2002. The Bank Erosion Hazard Index (BEHI) and near-bank stress (NBS), developed by Rosgen (1996Rosgen ( , 2001) are some of the most common methods used (Somervielle & Pruitt 2004).…”

Monitoring Fine Sediment; Grande Ronde and John Day Rivers, 2001-2003 Final Report.

“…These authors concluded that alternative methods need to be developed to adequately describe stream substrate and reduce the level of variation due to streams and observers. Rather that simply imply that stream substrate measures are impractical (Roper et al 2002(Roper et al : p. 1644), these authors should have discussed the obvious biases introduced in their sampling methodology as described by Bunte and Abt (2001b), whom they cite. Burton (2000) used a 100-intersection grid to follow trends in fine sediment (<6.4 mm) in streams on the Boise National Forest, but did not specify grid spacing.…”

“…This grid method (Roper et al 2002, Archer et al 2004) relied on only three random tosses of a grid frame within a pool tailout. Even though the toss was reportedly random, three tosses are not sufficient to ensure representation of the entire channel cross-section, given its potential spatial heterogeneity.…”

“…Although fine sediment is widely known to be a major impediment to salmon survival and recovery, the attitude toward sediment monitoring in streams is often that it is too variable, too inaccurate, and subject to excessive operator error (McDonald et al 1991, Roper et al 2002, Archer et al 2004. Even if sediment monitoring presents technical challenges, choosing not to monitor an environmental parameter that is known to be a key limiting factor is unacceptable.…”

Alternative representations of in‐stream habitat: classification using remote sensing, hydraulic modeling, and fuzzy logic

“…Conventional stream classification methods, however, suffer from several fundamental limitations (Poole et al 1997, Goodwin 1999, Roper et al 2002. Typical approaches include process-oriented categorizations (Whiting andBradley 1993, Montgomery andBuffington 1997), visual qualitative schemes (Bisson et al hydrodynamic simulations, remotely sensed data, and fuzzy classification techniques to explore potential alternative representations of in-stream habitat and characterize spatial patterns of classification uncertainty.…”