Several researchers, working with a variety of warm-water fishes, have 

 reported that the acute response to ammonia was not affected when dissolved 

 oxygen levels dropped from saturation to approximately one-half or one-third 

 saturation, but below that resistance decreased (Wuhrmann 1952: Wuhrmann and 

 Woker 1953; Merkens and Downing 1957; Danecker lb»64; Vamos and Tasnadi 

 1967). Reports on rainbow trout generally agree that this species is more 

 sensitive than warm-water fishes to the combined effects of low dissolved 

 oxygen and ammonia, and that any reduction in dissolved oxygen or any reduc- 

 tion below two-thirds saturation will decrease rainbow trout tolerance to 

 amnonia (Allan 1955; Downing and Merkens 1955; Merkens and Downing 1957; 

 Danecker 1964), One of the findings reported by Downing and Merkens (1955), 

 who tested young rainbow trout in experiments lasting up to 17 hours, was 

 that a decrease in dissolved oxygen from 8.5 to 1.5 mg/liter shortened the 

 periods of survival at all ammonia concentrations tested; this decrease was 

 proportionally greatest at the lowest concentrations of ammonia. In longer 

 tests, lasting up to 13 days, these same researchers reported similar re- 

 sults (Merkens and Downing 1957). 



To explain the accelerated action of ammonia toxicity under reduced 

 oxygen conditions, Lloyd (1961) presented the argument that a given toxic 

 effect is produced by a specified concentration of toxicant passing across 

 the fish gill surface at a rate governed by the fish gill movement. At re- 

 duced oxygen concentrations the rate of movement increases, resulting in an 

 increased rate of gill exposure to the toxicant. He hypothesized that a re- 

 duction in CO2 excretion at the gill surface, resulting from reduced O2 in- 

 take, will raise the pH at the gill surface. Such an increase in pH will 

 favor the more toxic ammonia species (NH3) resulting in an even more accele- 

 rated toxic effect of ammonia than might be expected solely by an increased 

 rate of gill movement. However, CO2 loss at the gill surface is also con- 

 nected with the fish's ammonia excretion mechanism, and recent research on 

 the possible toxicity of NH4"'" suggests that a complete explanation may be 

 more complex. 



To examine the effect of dissolved oxygen on ammonia toxicity we con- 

 ducted two series of 96-hour flow-through bioassays, one of these (15 bio- 

 assays) on rainbow trout, and the other (10 bioassays) on fathead minnows. 

 Test conditions were similar to those described earlier, and test fish were 

 acclimated to the test oxygen level for at least 2 days prior to introduc- 

 tion of ammonia toxicant. The rainbow trout for all tests were from the 

 same stock, and the stock fish grew in size over the several weeks that the 

 tests were conducted so the average test fish size gradually increased from 

 2 to 10 g. The tests were not run in any particular sequence of dissolved 

 oxygen level, however, and subsequent statistical treatment showed that 

 there was no correlation between test result and fish size. Figure 4 shows 

 a plot of the 96-hour LC50 value (mg/liter NH3) for each test vs. the dis- 

 solved oxygen level at which the test was conducted. The correlation for 

 rainbow trout between LC50 and dissolved oxygen was striking (correlation 

 coefficient 0.9346, P = 0,00001); the lower the dissolved oxygen concentra- 

 tion, the greater the toxicity of ammonia. Although a regression line for 

 the fathead minnow tests was obtained, the slope of this line is not statis- 

 tically different from zero (P = 0,365), We conclude that there is most de- 



127 



