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



1.1 and 3.2 Lis,, the cost for skipjack tuna are, 

 respectively, 2.5 and 1.4 times those for sockeye 

 salmon. We can only conclude that 1.8 kg skip- 

 jack tuna swim at intermediate speeds less effi- 

 ciently than 1.8 kg sockeye salmon — this despite 

 the fact that, among fishes, the skipjack tuna 

 represents the apex of evolutionary engineering 

 for speed (Magnuson 1973; Stevens and Neill 

 1978). Presumably, the evolution of skipjack tuna 

 (like that of fast cars) has involved sacrifice of 

 energetic efficiency at low speeds in favor of in- 

 creased efficiency at high speeds, permitting a 

 dramatic increase in maximum attainable speed. 



Interrelation of Metabolic Rate, 

 Swimming Speed, and Body Weight 



Voluntary speeds {S, lengths/second) of skip- 

 jack tuna swimming in our laboratory respirom- 

 eters were inversely related to fish weight by the 

 relation S = 3.14 - 0.53 /logW. Magnuson 

 (1973), working from basic hydrodynamic rela- 

 tions, predicted minimum speed for steady-state 

 swimming in various tunas; his model for skip- 

 jack tuna yielded a speed versus fish-weight rela- 

 tion very similar in slope to that we observed 

 (Figure 2). The difference in means may be attrib- 

 utable to differences in condition factor and/or 

 body-water content (Kitchell et al. 1977) between 

 our captive fish and the wild skipjack tuna on 

 which Magnuson's calculations were based. 



Oxygen-uptake rates (Vbj. milligrams O2/ 

 grams per hour) of our laboratory fish were 

 influenced not only by swimming speed but also by 

 fish weight independent of speed: log Vo^ - 

 -1.20 + 0.19 logW + 0.21s. We have concluded 

 above that 1) the intercept value ("standard" rate 

 at any weight) is unusually large for fishes; 2) the 

 weight coefficient is opposite in sign from that 

 typical of fishes (and of organisms, generally); and 

 3) the interdependency of log Vbz and S on weight 

 is compensatory, resulting in no statistically 

 demonstrable difference among oxygen-uptake 

 rates for skipjack tuna of various weights (600- 

 4,000 g) swimming at their characteristic speeds. 



Conclusion (3) led us to explore the relation 

 between oxygen-uptake rate per unit distance 

 (V62, milligrams 02/gram per kilometer) and 

 swimming speed for skipjack tuna of different 

 sizes. Exponentiating the linear regression equa- 

 tion relating V02 in milligrams 02/gram per 

 hour to W and S, Vq^ = 0.063 • W"^^ • 10" "S 

 = 0.063  W°i» • e°^«^. Multiplying the last 



equation by 27.78 km ^ • S 



-1 



gener- 



ated an equivalent expression for ^62 in milli- 

 grams 02/gram per kilometer: 



Vo, = 1.75 • L 



W 



0.19 . _0.48S 



Finally, we used the exponentiated length-weight 

 relationship for experimental fish: logW = 

 -2.657 + 3.532 logL; thus, W = 0.0022L3^='2 

 to eliminate W: 



Vo, = 0.55 • L 



0.33 



1 . _0.48S 



Solutions of this equation for V62 at various 

 values of L and S are shown graphically in Figure 

 9. Small fish are less efficient (higher V62) at any 

 particular speed than are larger fish, but fish of all 

 sizes reach their particular minimum V62 at the 

 same relative speed — about 2.1 L/s. The relation 

 between this, the optimum speed (Sopt) for cover- 

 ing distance, and the value of the coefficient, 0.48, 

 for the exponential term in S is simple — each is 

 the reciprocal of the other: 



dVo. ^^^ _ ,33 0.48Se«^«^-e»''«« 

 0.55 • L "^^ per 



ds 



at 



ds 



s- 



= 0, 0.48S • e" ^«^ = e" '^^; 



therefore, Sopt = 



0.48 



2.08. 



For skipjack tuna between 30 and 60 cm length, 

 the characteristic speeds and Sopt = 2.08 corre- 

 spond with V62 rates that are maximally (for 60 

 cm fish) different by only 13% of min V02 (Figure 

 9). The question arises as to whether the observed 

 characteristic speeds, rather than "Sopt," might be 

 the (evolutionary) "design" speeds that minimize 

 VOz- The characteristic speeds agree remarkably 

 well (better than does "Sopt") with the optimum 

 speed predicted by Weihs' (1973b) model; Weihs, 

 reasoning from thrust and drag relations for 

 fishes, argued that speed is optimized (energy 

 expended per unit distance is minimized) when 

 "the rate of energy expenditure required for pro- 

 pulsion [and associated physiological work?] is 

 equal to the standard (resting) metabolic rate." 

 For our skipjack tuna, S at Vb2 equivalent to twice 

 the hypothetical standard rate was 1.43 L/s, a 

 value that falls midway in the range of speeds 



43 



