Klamath Basin has historically received the largest concentration of migratory waterfowl in North America, as many as 5 to 6 million ducks and geese.  The basin is also heavily used by threatened bald eagles (Haliaeetus leucocephalus).  A large fraction of these bird populations forage in Lower Klamath and, to a lesser extent, Tule Lake National Wildlife Refuges (NWRs).  Two endangered species of fish, the Lost River (Deltistes luxatus) and shortnose (Chasmistes brevirostris) suckers, are also restricted to the Klamath Basin.  The smallest known population of these suckers, estimated at 250 adult fish, reside in Tule Lake sumps 1A and 1B, in the Tule Lake NWR.  Tule Lake populations of these suckers are largely isolated from other populations as a result of the Anderson-Rose Dam to the north and irrigation pumps at the outlet of Tule Lake.  The low population size of the endangered suckers at Tule Lake has been attributed to a number of factors, but primarily habitat loss, specifically the loss of deep water habitat. Tule Lake has been reduced from an historical 40,470 to 5,383 hectares to provide for irrigated farm acreages, and average depths have been reduced from 3 to 10 meters to approximately 1 meter as a result of both lake draining and siltation.  In addition to the loss of this habitat, fish passage to historic spawning areas is blocked by the Anderson-Rose Dam, although some spawning activity has been documented immediately below the dam.  Also, the introduction of nonnative fish species has probably contributed to reduced sucker populations. Both poor water quality upstream of Tule Lake NWR and the shallow water conditions of Tule Lake have contributed to poor water quality, including low dissolved oxygen and high concentrations of ammonia.  More than 50 different pesticides are used by refuge farmers.  Additionally,  more toxic pesticides, such as aldicarb, are used on private agricultural lands adjacent to and upstream of the Tule NWR.  The effects of these pesticides on aquatic species is unknown; previous pesticide studies have not established any clear link with acute or chronic effects in aquatic systems (Sorenson and Schwartzbach 1991; Dileanis et al.1996).  In two instances, applications of acrolein (Magnicide-H) upstream or in an adjacent canal preceded fish kills.  The pesticide, acrolein,  is used by the irrigation district  to clear canals of submergent and emergent vegetation.  It is an extremely volatile compound with a half-life ranging from about 14 to 92 hours in water depending on pH, water temperature, and turbulence.  Its unstable nature produces several metabolites, some of which are themselves either volatile or common in natural waters and therefore difficult to assess. Acrolein is an extremely hazardous substance with chronic and acute toxicity concentrations to freshwater aquatic life occurring in the parts-per-billion range. Despite, the potential toxicity of acrolein, most studies in refuge waters have not demonstrated toxicity to aquatic life (Winchester 1994; Dileanis et al. 1996).

This study did not disclose any acrolein in refuge waters and patterns of Daphnia magna and fathead minnows mortality were not consistent with expectations from any acrolein-induced mortality.  If acrolein had been responsible for observed mortalities, consistently higher mortality should have been observed in refuge border sites, closer to the acrolein application sites, rather than at Tule Lake.  Moreover, fathead minnows would be predicted to experience mortality more readily than Daphnia magna at the refuge border sites.

No other pesticides were found in significant concentrations (Dr. John Moore, Patuxent Analytical Control Facility, pers. comm.).  However, the study did not include analysis of glyphosate.  The herbicide, Rodeo (active ingredient, glyphosate), was applied to control loosestrife on refuge lands immediately adjacent to Tule Lake sump 1A occurred on August 14.  This application, on the west side of sump 1A [not near the study site] may have contributed to the observed fish mortalities both in the in situ containers, where 100% mortality was observed by August 17 (96 hours).

Also, between August 14 and August 17, approximately 25 to 35 fish (fathead minnows, tui chubs, Sacramento perch) were observed floating dead at Pump 10 on Tule Lake.  During this period, numerous other fish (fathead minnows) were observed flashing, an indication of distress.  However, no dead fish were observed at two other Tule Lake sites during the the fish kill, suggesting that poor water quality alone at our study site (including dissolved oxygen levels of 0.0 mg/L) is the more likely explanation for the observed mortalities. In the future, to avoid even the possibility of Rodeo impacts, we recommend applying this herbicide in the spring, when dissolved oxygen concentrations in the lake are acceptable.  

Learn more by reading the following full following report:

Snyder-Conn, Elaine, Acrolein On Aquatic Ecosystems In Tule Lake National Wildlife Refuge  US Fish and Wildlife Service Klamath Falls Fish and Wildlife Office, 1997

Return to the Klamath Falls Field Office Reports

Visit the: Pacific Region - Environmental Contaminants-Investigating and Monitoring

Visit the: USFWS- Pacific Region Ecological Services Home Page

Visit these sites for more information on this topic:

Tracking Suckers on Tule Lake Refuge, Summer 1999- Klamath Basin NWR

Endangered Sucker Fish Update: Klamath Experiment Station- OSR Agricultural Research