wildlife refuge along the Illinois River at Havana, killed 15% of the 

 test fish in 96 hours; deeper sediments, taken at the 30,5-45.7 cm (12- 

 to 18-inch) depth, killed 25%. Fish mortality correlated (R = 0.71, P < 

 0.01) with the concentration of un-ionized ammonia released from the 

 sediment and both ammonia and fish mortality increased upstream toward 

 Chicago. The Long Term Research Monitoring Station (LTRM) at Havana 

 started measuring ammonia concentrations in Anderson Lake, a floodplain 

 lake of the Illinois River and a state fish and wildlife area, on 1 May 

 1990, 2 days after a fish kill. The total ammonia nitrogen 

 concentration was 0.90 mg/1 and the un-ionized ammonia nitrogen was 

 calculated to be 0.36 mg/1 at the temperature of 16.6° C and pH of 9.34. 

 NH3-N concentrations of 0.32 mg/1 at 3-5° C and 1.35 mg/1 at 24-25° C 

 were acutely lethal to bluegill sunfish, Lepomis macrochirus (Reinbold 

 and Pescitelli 1990). The fish kill might have been caused by ammonia, 

 if the un-ionized ammonia had peaked at higher concentrations before our 

 samples were taken. 



Elevated un-ionized ammonia concentrations might be triggered by 

 resuspension of sediments or episodes of elevated pH resulting from 

 phytoplankton blooms. Plants remove carbon dioxide from the water, in 

 the form of carbonic acid and bicarbonate, and thereby elevate the pH of 

 the water, which in turn increases the proportion of ammonia existing in 

 the toxic, un-ionized form. The Havana LTRM station (unpublished data) 

 measured pHs as high as 10.12 in backwater lakes of the Illinois River 

 in July 1990 and values between 9.0 and 10.0 occur fairly often. 

 Episodes of acute ammonia toxicity thus may be occurring sporadically in 

 places other than just the upper Illinois River, and it takes only one 

 brief episode per year to kill or reduce populations of invertebrates or 



50 



