GOODING ET AL.: RESPIRATION RATES AND LOW-OXYGEN TOLERANCE IN SKIPJACK TUNA 



^0 45 ^j^^ thus, fit Weihs' model almost perfectly. 

 We conclude by noting that our discussion of 

 optimum swimming speeds for covering distance 

 relates only to skipjack tuna swimming at con- 

 stant depth (as those in our respirometers were 

 required to do). Weihs (1973a) has calculated that 

 negatively buoyant fishes like the skipjack tuna 

 could achieve an energy savings of 20% (compared 

 with swimming at constant depth) by alternately 

 gliding downward at an angle of about 11° (to the 

 horizontal), then actively swimming upward at an 

 angle near 37°. 



Resistance to Low Oxygen 



In areas of the world ocean with surface waters 

 not stressfully warm for skipjack tuna there 

 is always available air-saturated water that over- 

 lies oxygen-depleted strata (Barkley et al. 1978). 

 Therefore, the 4-h exposure period we adopted in 

 this study would seem to include all intervals of 

 low-oxygen exposure that skipjack tuna ever need 

 endure at sea. 



The data suggest for skipjack tuna a threshold 

 of response to hypoxic stress at about 4.0 mg O2/I 

 (Figure 7); this value is at or below that represen- 

 tative of fishes (Davis 1975). In our experiments, 

 the skipjack tuna's response to low oxygen was an 

 increase in swimming speed; this would seem 

 adaptive in that increased swimming speed initi- 

 ated by hypoxic stress would facilitate return 

 of fish from deep, oxygen-depleted water to air- 

 saturated surface water. 



The 4-h median tolerance limit to low oxygen 

 was also near 4.0 mg O2/I (Figure 6). This value, in 

 keeping with the skipjack tuna's exceptionally 

 high metabolic rate, is apparently higher than 

 that of any other fish yet investigated (Doudoroff 

 and Shumway 1970). 



Angular Acceleration and 

 Excess Body Temperature 



Compared with other studies of fish metabo- 

 lism, our experiments with skipjack tuna involved 

 two unusual elements: 1) The fish were forced, by 

 the relatively small size of the tanks, to swim 

 a curved path, and 2) they probably had core 

 temperatures up to several degrees higher than 

 the temperature of the surrounding water. 



Weihs (1981) has suggested that our continuous- 

 ly turning fish expended more propulsive energy 



than they would in swimming a straight path at 

 the same speed. A turning tuna must counter 

 centrifugal forces by "banking" with its pectoral 

 fins to produce a component of lift directed in- 

 wards along the turning radius. Therefore, our 

 results may overestimate the oxygen-uptake rates 

 and perhaps also the lower lethal oxygen concen- 

 tration for skipjack tuna at sea. However, we 

 doubt that the magnitude of the overestimate can 

 be very great, for fish in the large and small 

 respirometers (with radii of typical swimming 

 paths about 2 and 0.8 m) respired at similar rates 

 (Table 1). Furthermore, metabolic rates of fish 

 in our experiments were consistent with those 

 inferred from weight and energy "loss" rates of 

 starved skipjack tuna living in tanks 7.3 m in 

 diameter (Kitchell et al. 1978). 



Oxygen-uptake rates of our test fish also com- 

 pare well with theoretical estimates of the amount 

 of energy consumed by similarly sized fish swim- 

 ming a straight course at the same speeds (Figure 

 10). The observed oxygen-uptake relationship 

 (milligrams 02/hour) was extrapolated from 

 to 8.5 L/s for four skipjack tuna ranging in 

 weight from 800 to 3,800 g (dashed lines). (Recall 

 that mean speeds of our fish were between only 0.9 

 and 2.2 L/s.) Superimposed on the empirical 

 relationship are theoretical projections of energy 

 consumption based on estimates of drag force. 

 Theoretical energy uptake — in keeping with the 

 reasoning of Webb (1975), Sharp and Francis 

 (1976), Sharp and Vlymen (1978), and Dizon and 

 Brill (1979) — was computed according to the fol- 

 lowing rationale: 



1. Total power required is the sum of the power 

 required for nonswimming processes (P2) plus 

 power required for thrust (Pi), the latter divided 

 by an estimate of total aerobic efficiency =0.2 

 (Webb 1975). 



2. Power required for nonswimming metabolic 

 processes (the standard metabolic rate of a fasted 

 fish from Brill (1979)), 



P2 - 1.53 • W^^^ 



where P2 = power (watts), 



W = weight (kilograms). 



Brill's (1979) relation is used despite some 

 doubts about the validity of the exponent because 

 it provides for skipjack tuna the only estimate of 

 P2 independent of our data. 



45 



