In natural waters with low to moderate amounts of dissolved solids 

 (200-1000 mg/liter), this effect will slightly lower the concentration of 

 NH3. The magnitude of this effect will vary with the composition of the 

 water in question. For a water of high pH (8-9) and total dissolved solids 

 (TDS) of 500 mg/liter, which are predominantly calcium salts, the effect 

 on the fraction of NH3 present is approximately the same as if the tempera- 

 ture were lowered one degree. For waters of lower pH (5.5-6), but still 

 high in calcium, somewhat higher values of TDS (600-700 mg/liter) would be 

 required to produce a similar effect. For waters in which sodium chloride 

 is the dominant ionic species, approximately twice these amounts of TDS 

 would be necessary to produce a change comparable to a one-degree drop in 

 temperature. 



Toxicity of Ammonia to Fishes 



Concentration values for ammonia toxicity tests on fishes have been 

 variously reported as NH3, NH3-N, NH4OH, total ammonia, total ammonia 

 nitrogen, and formula weight for ammonium salts. Calculation of the per- 

 centage of total ammonia as un-ionized ammonia has also been made in a 

 variety of ways, sometimes incorrectly. Recalculation of reported values 

 is not always possible because of failure to report essential water 

 chemistry parameters. Nonetheless, certain trends have developed which 

 give some approximation of lethal levels of ammonia for salmonids and some 

 species of warm water fishes. 



In the case of short-term tests on rainbow trout [Salmo gcuJidmii) fry 

 and fingerlings, median lethal concentration (LC50) values as low as 0.2 

 mg/liter NH3 have been reported (Liebmann, I960; Danecker, 1964). Other 

 researchers have reported LC50 values ranging between 0.3-0.6 mg/liter 

 NH3 for tests of one day or less on rainbow trout (Lloyd and Herbert, 1960; 

 Herbert and Shurben, 1963, 1965; Ball, 1967; Lloyd and Orr, 1969; Smart, 

 1976), on brown trout [Salmo tnutta.) fry (Penaz, 1965), and on Atlantic 

 salmon [Salmo i>alaA) smolt (Herbert and Shurben, 1965). 



In the case of short-term tests on fishes other than salmonids, Hazel, 

 it at. (1971) reported 4-day LC 50 values of 1.4 mg/liter NH3for striped 

 bass [Mccone. i>axa£ilU) and 1.0 for sticklebacks [Ga^ttnoi,tza6 aculeatiu,). 

 Colt (1974) reported 4-day LC50 values ranging from 2.4-3.8 mg/liter NH3 

 for channel catfish [IctalunuA pm\n.cXatuu>) , and LC50 values ranging from 

 1.9-3.4 have been observed for fathead minnows [Puncphalm pnomelai) 

 (Thurston, unpublished data). A 17-hour LC50 of 1.3 mg/liter NH3 was re- 

 ported for Gambusia [GambuAia a^^lwU) (Hemens, 1966), and a 24-hour LC50 

 of 2.9 was reported for channel catfish (Robinette, 1976). Lower LC50 

 values between 0.35-0.50 mg/liter NH3 have been reported for 5- to 7-day 

 tests on bream [Ab'iamu biama) , roach [Rutilu.6 nutiluA), perch [Ptftca 

 {)luviatllAj>), and rudd [ScaAdivunuA zfiyth^ophthalmiu) (Ball, 1967). In a 

 longer test on rudd (Water Pollution Research, 1971) the LC50 value for 7 

 days was 0.5 mg/liter NH3 and for 95 days was 0.24. 



There are Title published data available on longer-term mortality 

 tests for fishes of any species. A three-month test on 200 rainbow trout 

 (Water Pollution Research, 1967) showed that 15 percent died at 0.22 



77 



