FISHERY BULLETIN: VOL. 79, NO. 1 



600 

 400 



2.00 



;= 1.00 



'p. 80 



° 60 



E 







a. 



3 



.02 



1.8 kg SKIPJACK TUNA AT 

 aA-C (REGRESSION OF 

 LABORATORY DATA ) 



1.8 kg SKIPJACK TUNA AT - 

 24 "C (JUST-CAUGHT FISH) 



SOCKEYE 



SALMON AT IS'C 

 AND GLASS 1973) 



2 3 4 



SWIMMING SPEED (length»sec-i) 



Figure 8. — Comparison between respiration-speed relations 

 for 1.8 kg skipjack tuna calculated from the present study and 

 for 1.8 kg sockeye salmon computed from equations given by 

 Brett and Glass (1973). Ranges of observed values are indicated 

 by the lines extending from the median value for just-caught 

 skipjack tuna. Length measures are fork length. 



we been able to accurately measure swimming 

 speeds in just-caught fish. The basis for our cau- 

 tious appraisal of such apparently good "fit" is 

 suspicion that the linear model, log Voj = a 

 + b  speed, which seems adequate for many 

 fishes (Brett 1972; Brett and Glass 1973; Webb 

 1975), cannot hold for skipjack tuna over the en- 

 tire range of swdmming speeds that they can sus- 

 tain. Personal observations on these fish and 

 Yuen's (1970) report of a school of skipjack tuna 

 (ca. 44 cm fish) that traveled 28 km in 107 min 

 (average minimum speed = 4.4 m/s) convince us 

 that 40-50 cm skipjack tuna can swim for at least 

 an hour at speeds near 10 Lis. If that is so, our 

 linear model predicts oxygen uptake in 1.8 kg 

 skipjack tuna (median size of just-caught fish) 

 at a maximum sustained rate of at least 33.0 mg 

 02/g per h. Active metabolic rate of skipjack tuna 

 may substantially exceed Brett's (1972) predicted 

 maximum for fishes, but we are confident it does 

 not do so by a factor of nearly 30. The most logical 

 interpretation of this conundrum is nonlinearity 

 in the relation between log Vq^ and speed; as 

 skipjack tuna swim faster, they must become 

 more efficient in their use of oxygen and energy. 



The same is probably true for other fast-swim- 

 ming fishes, such as Peterson's (1976) striped 

 mullet, Mugil cephalus. Even in the relatively 

 sluggish goldfish, Carassius auratus, oxygen- 

 uptake rate actually declines as the fish pass 

 from spontaneous activity at low apparent speeds 

 to induced swimming (against currents) at higher 

 speeds (Smit 1965). 



There is, of course, an alternative explanation: 

 Our laboratory experiments overestimated the 

 true coefficient for speed. In fact, taking the lower 

 95% confidence limit on the speed coefficient — 

 0.11 — yields a comparatively modest 3.31 mg 02/g 

 per h for predicted Vba at 10 Lis. But a true speed 

 coefficient as low as 0.11 is not only inconsistent 

 with the comparable coefficient in other fishes 

 (Fry 1971; Brett 1972) but also with other, inde- 

 pendent data (Chang et al.^) on metabolism-speed 

 relations in skipjack tuna. The speed coefficient 

 estimated from that study was 0.22, a value re- 

 markably similar to our mean estimate. 



To close our consideration of activity-related 

 metabolism in skipjack tuna, we offer a compari- 

 son between respiration-speed relations of a 1.8 

 kg skipjack tuna at 24° C and a 1.8 kg sockeye 

 salmon, Oncorhynchus nerka, at 15° C (Figure 8). 

 We chose the sockeye salmon because its active 

 metabolic rate "is one of the highest [for fishes] 

 on record, exceeding that determined for other 

 salmonids by 30% to 40%" (Brett and Glass 1973). 

 The sockeye salmon respiration-speed relation 

 was computed from equations given by Brett and 

 Glass (1973); 15° C-values were used because this 

 is near the sockeye salmon's thermal optimum 

 for fast swimming and several other vital func- 

 tions (Brett 1971). Skipjack tuna seem to swim 

 and metabolize at rates nearly independent of 

 temperature (Dizon et al. 1977; Chang et al. 

 footnote 4). 



At all speeds common to the two fishes, skip- 

 jack tuna have the higher metabolic rate — 3.7 

 times higher at 1.1 Lis (the skipjack tuna's mini- 

 mum speed) decreasing to 1.7 times higher at 3.2 

 Lis (the sockeye salmon's maximum sustained 

 speed). If the basis of comparison is the energy 

 cost of swimming (oxygen-uptake rate associated 

 with any particular speed minus standard up- 

 take), the difference between these fishes is less- 

 ened but the qualitative relation is unchanged: at 



"Chang, R. K. C, B. M. Ito, and W. H. Neill. Manuscr in 

 prep. Temperature independence of metabolism and activity 

 in skipjack tuna, Katsuwonus pelamis. Southwest Fish. Cent., 

 Natl. Mar. Fish. Serv., Honolulu, HI 96812. 



42 



