any rate, that regardless of a probable higher metabolic rate in the smaller fish, they are not 

 markedly hindered in their ability to extract oxygen from water low in oxygen content. 



A small, but nonetheless statistically significant, difference was^ found between the 

 mean summer lethal value of oxygen (1.0 cc. /I. , avejage temperature 25. 6 C, ) and the mean 

 winter value (0.96 cc. /I. , average temperature 21.9 C. ). This difference is of about the expected 

 magnitude according to the relationship of temperature to the lethal value of oxygen (fig. 12). In 

 the absence of the appropriate data, however, it is difficult to determine the possible influence of 

 purely seasonal differences in the behavior of the fish. 



A few investigators (Gardner and King 1922; Moore 1942; Fry et aX. 1947; Graham 

 1949) have studied the effects of temperature on lethal values of oxygen in fish and found, as in iao, 

 that the lethal level of oxygen increases as the temperature increases. The curve of temperature 

 versus lethal values of oxygen (fig. 12) shows a somewhat steeper rise from 26 to 31 C., rough- 

 ly corresponding to the pattern shown in the relationship of metabolic rate to temperature (fig. 4). 

 Thus, with a greatly increased metabolic rate at the higher temperatures, the fish is less able to 

 cope with low values of dissolved oxygen. This observation may be of some importance to the live- 

 bait fishery. As mentioned earlier, the fishery is conducted primarily in the summer months, 

 when temperatures often rise above 26 C., particularly in calm anchorages. At the higher tem- 

 peratures, even a small increase in temperature may bring about a marked decrease in the resis- 

 tance of the fish to oxygen deficiency. 



The lethal values of oxygen reported here from laboratory experiments should be 

 checked by more extensive measurements taken directly from the live-wells of the fishing boats. 

 Such tests might possibly give lethal values intermediate between the average for iao (about 1.0 

 cc. /I.) and that for nehu (about 2.0 cc. /I.), since many bait loads are "mixed", containing both 

 nehu and iao, 



SUMMARY AND CONCLUSIONS 



1. Confinement of tuna baitfish in small live-wells pointed to the desirability of a study of the 

 oxygen requirements of the fish. Experiments were designed to deternnine lethal levels of 

 oxygen amd the rate of oxygen consumption under various conditions. In the oxygen consunnp- 

 tion experiments special emphasis was laid on the effects of crowding, flow-rates, and the 

 oxygen concentration of the water. 



2. A flowing-water method was used to measure oxygen consumption, dissolved oxygen being 

 determined by the unmodified Winkler procedure. For tests of lethal values of oxygen, where 

 the fish were asphyxiated in sealed jars, the permanganate modification of Winkler's technique 

 was employed. 



3. The results are summarized below: 



a. From serial observations, it was evident that oxygen consumption in the iao showed con- 

 siderable hour-to-hour fluctuation even when there were no outward signs of increased or 

 decreased activity on the part of the fish. No rhythmic cycle in metabolic rate could be 

 demonstrated, at least during the first 24 hours in the container. 



b. Oxygen consumption increased with increase in flow-rate, the effect being perhaps more 

 pronounced in the summer months. In addition, oxygen consumption per unit weight was 

 significantly lower in the tests with large numbers of fish. Possible reasons for the 

 latter finding were discussed. 



c. The "critical oxygen level", i.e., the point at which the rate of oxygen consumption be- 

 came dependent upon the oxygen concentration of the water, could not be precisely de- 

 limited but lay somewhere between 1. 5 and 2. 5 cc. /I. Below this level, both breathing 

 rate and oxygen consumption steadily decreased. 



