A water quality study was performed in the mainstem Klamath River from Keno, Oregon to Seiad Valley, California during 1996 through 1998. Four sites within the study area were continuously monitored using multiparameter recorders. Water quality sampling was also performed at these four locations in 1996 and 1997. Additional water quality sampling sites were added in 1998 for a total of 8 locations between Keno and Seiad.

Temperature ranged from near zero °C to >25 °C with cooler temperatures in early spring and fall, and maximum temperatures occurring in july and August of each year. Dissolved oxygen concentration ranged from near zero mg/L to >13 mg/L with highest DO occurring in early spring and fall and lowest DO occurring in mid-summer. Air temperature was generally highly correlated with water temperature with r values ranging from 0.8 to 0.9 during the study period from 1996-1998. Water temperature in the study area exceeded chronic (>16°C) and acute (>22°C) criteria for salmonids during the summer months. Although chronic DO «7 mg/L) criteria were exceeded throughout most of the study area during the summer, in the free-flowing river below Iron Gate Dam the acute DO «5.5 mg/L) criteria were not exceeded.

Nonpoint source pollution in the form of agricultural return flows, industrial, or sewage effluent entering the stream may have resulted in higher ammonia and total organic nitrogen concentrations at the upstream locations in the Klamath River study area (Keno and J.C. Boyle Powerplant). Nitrification of ammonia and organic nitrogen seemed to result in higher concentrations of nitrate in the downstream Klamath River (Iron Gate Dam). Total phosphorus concentration stayed relatively stable through the reservoirs in the study area, but decreased in the downstream direction between Iron Gate Dam and Seiad. Ortho-phosphorus concentrations increased longitudinally through the reservoirs, then decreased in the downstream direction between Iron Gate Dam and Seiad. An increase in ortho-phosphorus concentration can indicate internal cycling occurring in the reservoirs as well as photosynthesis.

On an annual basis total phosphorus loading increased longitudinally from up- to downstream between Keno and Seiad. The increase was statistically significant (p =.03) indicating that the reservoirs in series in the Klamath River study area do not function as a nutrient sink. However, during the summer there was no statistically significant difference in total P loading when Keno, Iron Gate and Seia~ locations were compared, therefore, the reservoirs may act as a nutrient sink seasonally.

The Klamath River study locations were generally nitrogen limited, although at Keno, a regUlar change from N limitation to P limitation occurred during the fall of all three years of the study. When the Klamath River annual nutrient loading values are compared to other rivers in the vicinity, the Carson, Truckee, and Long Tom Rivers also appear to be nutrient enriched. The Carson and South Yamhill Rivers seem to be N limited systems and the Wood, Long Tom, Snake and Truckee Rivers seem to be P limited systems.

Implementing management strategies for reservoir operations to improve water quality and reduce nutrient concentration or loading in the Klamath River study area to benefit anadromous fisheries may be difficult and expensive. However, improving the thermal regime in spring to benefit YOY salmonids may be possible as is short-term relief in late summer for oversummering species. Decreases in nutrient concentration or loading accomplished through best management practices in the water shed may allow general protection of water resources in the Klamath Basin for future needs.