The effect of irrigation drainage on the water quality and wildlife of the Klamath Basin in California and Oregon was evaluated during 1990-92 as part of the National Irrigation Water Quality Program of the U.S. Department of the Interior. The study focused on land serviced by the Bureau of Reclamation Klamath Project, which supplies irrigation water to agricultural land in the Klamath Basin and the Lost River Basin. The Tule Lake and Lower Klamath National Wildlife Refuges, managed by the U.S. Fish and Wildlife Service, are in the study area. These refuges provide critical resting and breeding habitat for waterfowl on the Pacific flyway and are dependent on irrigation drainwater from upstream agriculture for most of their water supply.

Water-quality characteristics throughout the study area were typical of highly eutrophic systems during the summer months of 1991 and 1992. Dissolved-oxygen concentrations and pH tended to fluctuate each day in response to diurnal patterns of photosynthesis, and frequently exceeded criteria for protection of aquatic organisms.

Nitrogen and phosphorus concentrations were generally at or above threshold levels characteristic of eutrophic lakes and streams. At most sites the bulk of dissolved nitrogen was organically bound. Elevated ammonia concentrations were common in the study area, especially downstream of drain inputs. High pH of water increased the toxicity of ammonia, and concentrations exceeded criteria at sites upstream and downstream of irrigated land. Concentrations of ammonia in samples from small drains on the Tule Lake refuge leaseland were higher than those measured in the larger, integrating drains at primary monitoring sites. The mean ammonia concentration in leaseland drains [1.21 milligrams per liter (mg/L)] was significantly higher than the mean concentration in canals delivering water to the leaseland fields (0.065 mg/L) and higher than concentrations reported to be lethal to Daphnia magna (median lethal concentration of 0.66 mg/L). Dissolved-oxygen concentrations also were lower, and Daphnia survivability measured during in situ bioassays was correspondingly lower in the leaseland drains than in water delivery canals.

In static laboratory bioassays, water samples collected at the primary monitoring sites caused toxicity in up to 78 percent of Lemna minor tests, in up to 49 percent of Xenopus laevis tests, in 17 percent and 8 percent of Hyalella azteca and Pimephales promelas tests, respectively, and 0 percent in Daphnia magna tests. In situ exposure at the sites caused mortality in more than 83 percent of Pimephales tests and in more than 41 percent of Daphnia and Hyalella tests. Much of the observed toxicity appears to have been caused by low dissolved oxygen, high pH, and ammonia. Although water in the study area was toxic to a variety of organisms, no statistically significant differences in the degree of toxicity between sites were observed above or below irrigated agricultural land in any of the bioassays.

Pesticides were frequently detected in water samples collected at the monitoring sites during the 1991 and 1992 irrigation seasons. Among the most frequently detected compounds were the herbicides simazine, metribuzin, EPTC, and metolachlor and the insecticide terbufos. All the insecticides detected were at concentrations substantially below acute toxicity values reported for aquatic organisms.

The herbicide acrolein has been used extensively in the basin to manage aquatic plant growth in irrigation canals and drains. The concentration of acrolein was monitored in a canal near Tule Lake after an application in order to evaluate the potential for the pesticide to be transported to refuge waters. Although acrolein concentrations were toxic to fish in the channels adjacent to Tule Lake, very little of the canal water entered the refuge during the monitoring period.

Organochlorine pesticide concentrations in 25 surficial sediment samples collected in 1990 were below baseline levels commonly found in soils and sediment. Seventeen sediment samples were analyzed for chlorophenoxy acid herbicides and two samples were analyzed for organophosphorus and carbamate insecticides in 1992. No pesticides were detected in any of these samples.

Residues of the trace elements selenium, mercury, and arsenic in algae, invertebrates, fish, and avian eggs revealed no bioaccumulation problems. Concentrations of organochlorine compounds, especially of p,p' DDE, were associated with a mean 11-percent eggshell thinning in white-faced ibis. However, ibis populations appear to be increasing, and some eggs of ibis were relatively low in DDE concentration. DDE concentrations in eggs of western grebes were not as high as in the eggs of ibis. Concentrations and types of organochlorine compounds detected in grebe and ibis eggs were highly variable, indicating that the birds were exposed to these compounds outside the basin.

Fish and invertebrates inhabiting drainwater were representative of pollution-tolerant species assemblages. The aquatic communities retained little of their historic ecological structure. Extensive hydrologic modifications and hypereutrophic conditions in Klamath Basin waterways have degraded the quality of aquatic habitat and altered aquatic communities.