The U.S. Geological Survey and Bureau of Reclamation collected water-quality data from March 2012 to March 2015 at locations in the Lost River and Klamath River Basins, Oregon, in an effort to characterize water quality and compute a nutrient budget for the Bureau of Reclamation Klamath Reclamation Project. The study described in this report resulted in the following significant findings:

• Total phosphorus (TP), total nitrogen (TN), 5-day biochemical oxygen demand (BOD5), and 5-day carbonaceous biochemical oxygen demand (CBOD5) loads, calculated using the U.S. Geological Survey LOADEST software package at the upper and lower boundaries of the Klamath Reclamation Project, indicated higher loads at the upper boundary on the southern end of Upper Klamath Lake upstream of the Bureau of Reclamation A Canal diversion compared to the lower boundary on the Klamath River downstream of Keno Dam. Accounting for the diversion of loads down A Canal, BOD5 and CBOD5 loads decreased between these two sites during irrigation season, indicating that the Klamath Reclamation Project is not a large source of oxygen-demanding material and that much of the oxygen demand at study site FMT, the northern boundary of the study area, has been expressed by the time the same water passes through site KRK, the southern boundary of the study area.
  
• An evaluation of the nutrient balance along the Klamath River flowpath from sites FMT to KRK indicated that, during irrigation season in the 3 years of the study period (March 2012–March 2015), more loads of TP, TN, BOD5, and CBOD5 were being diverted from the Klamath River than were being added to the Klamath River from the combination of Klamath Straits Drain, regulated point sources along the Klamath River, and internal loading from the bottom sediments in the river. By contrast, during non-irrigation seasons, more loads were added to the Klamath River than were diverted through Ady and North Canals, and this difference primarily was due to additional loads to the river from the Lost River Diversion Channel.
  
• At the Lost River Diversion Channel, BOD5 loads were higher during irrigation season than non-irrigation season in all three study years owing to the high concentrations of oxygen-demanding cyanobacterial biomass from the seasonal blooms of Aphanizomenon flos-aquae in the Klamath River and Upper Klamath Lake. The difference between the two seasons was particularly large in years 2 and 3, when the low flows of these two drought years resulted in smaller nonirrigation period loads than in year 1. CBOD5 loads also were higher during irrigation season in years 2 and 3 than during non-irrigation season, indicating that the largest oxygen demand was coming from senescence of Aphanizomenon flos-aquae cells that are present in the Klamath River during the summer. However, during irrigation season in year 1, CBOD5 loads were lower than in the non-irrigation season, which may indicate that at times high concentrations of ammonia or cellular organic nitrogen leaving Upper Klamath Lake contribute a large nitrogenous oxygen demand as well.
  
• The smallest loads were computed for the farthest upstream sites in the Lost River Basin, suggesting that the upper Lost River Basin does not contribute substantial loads of TP, TN, BOD5, and CBOD5 to the Klamath Reclamation Project.
  
• Median concentrations of BOD5 and CBOD5 were lowest among the upper Lost River Basin sites and highest at site PPD (however, this comparison is based on only four samples collected at site PPD over the 3-year study). Median concentrations of BOD5 and CBOD5 also were elevated at sites KSDH (6.60 and 4.70 milligrams per liter [mg/L], respectively) and KSD97 (4.47 and 3.45 mg/L, respectively). The highest maximum BOD5 and CBOD5 concentrations were reported at the Lost River Diversion Channel (39.0 and 26.5 mg/L, respectively) when water was flowing from the Klamath River toward the Klamath Reclamation Project, and site FMT (25.0 and 23.9 mg/L, respectively), the study site at the southern end of Upper Klamath Lake. Carbonaceous oxygen demand, as represented by CBOD5, typically dominated the composition of the samples at all sites.
  
• The highest concentrations of dissolved organic carbon were present at sites KSDH (the headworks of Klamath Straits Drain) and KSD97 (Klamath Straits drain before it enters the Klamath River), and PPD (outlet of Tule Lake).
  
• Median concentrations of TN and TP at the upper Lost River Basin sites in years 1 and 2 were variable, but site MCRV showed a smaller range of values in those years compared to the other upper Lost River Basins sites, and an overall lower median concentration during irrigation seasons in years 1 and 2, suggesting that Gerber Reservoir does not contribute high concentrations of nutrients to the Lost River during irrigation season.
  
• Total Maximum Daily Load (TMDL) load allocations for TP and TN in Klamath Straits Drain were exceeded in all three study years. BOD5 load allocations were exceeded in years 1 and 2, but not year 3.
  
• TMDL load allocations for TP were exceeded in the Lost River Diversion Channel for all 3 years. Load allocations for TN were exceeded in year 1, but not in years 2 and 3. BOD5 loads were less than the TMDL load allocation for all three study years.
  
• The dearth of samples collected at the Klamath Straits Drain just downstream of the Lower Klamath National Wildlife Refuge did not allow for direct assessment of the Klamath Straits Drain acting as a nutrient source or sink.
  
• TP, TN, BOD5, and CBOD5 loads estimated during the study period likely were smaller than long-term average conditions because of persistent drought conditions in the Upper Klamath Basin. The study results, therefore, fail to characterize loads from the Klamath Reclamation Project to the Klamath River that could be present in typical years, and suggest the need for load assessments during average or aboveaverage streamflow years.