North Fish Creek is underused as a trout and salmon hatchery despite its excellent water quality. The shifting-sand streambed in the lower 9 miles of the stream inhibits successful spawning and is a poor habitat for macroinvertebrates, the primary food for juvenile trout and salmon. To provide data necessary for evaluation of potential sand-loading-control practices, the U.S. Geological Survey determined total-sediment transport, particle size, and loads for three sites, designated A, B, and C, on North Fish Creek during water years 1990-91.

At site C, the most upstream site, all sediment was transported as suspended sediment. The average annual total-sediment load during 1990- 91 was 479 tons. About 88 percent of the load was transported during periods of snowmelt or storm runoff. About 75 percent of the sediment load was silt- and clay-size particles; the remainder was sand.

Total-sediment discharge was calculated by the modified-Einstein procedure to determine total sediment transport-rate relations for site A, the most downstream site, and for site B. Annual totalsediment load was 11,960 tons in water year 1990 and 18,430 tons in water year 1991 at site B. About 97 percent of the total load was transported during periods of snowmelt and storm runoff. About 60 percent of the total-sediment load was sand-size particles.

Annual total-sediment loads were 20,690 tons and 33,100 tons in water years 1990 and 1991, respectively, at site A. About 67 percent of the total-sediment load was sand-size particles.

Annual average streamflow, as indicated by flow in the Bois Brule River, was about 16 percent below average in water year 1990, and about 4 percent above average in water year 1991.

There was little relation between watershed area and sediment loads for the three sites. The watershed of site C is about 41 percent of that of site A, but the sand load at site C was only 1 percent of that at site A. The watershed area between sites B and C is 40 percent of that above site A, but this area yielded 49 percent of the sand load at site A. Nineteen percent of the watershed above site A is between sites A and B, yet this area yielded about 50 percent of the sand load at site A.