and the wound packed with foam rubber to 

 minimize bleeding. Electrocardiogram (ECG) 

 leads were mounted subcutaneously on the ven- 

 tral body surface and an 18-gage hypodermic nee- 

 dle, with a thermistor bead mounted in its tip, was 

 pushed through the dorsal musculature to the ver- 

 tebral column. The ECG signal was displayed on 

 an oscilloscope and the red muscle temperature 

 determined by balancing a wheatstone bridge cir- 

 cuit containing the needle mounted thermistor 

 and a 5-decade resistance substitution box. 



Oxygen concentration of the water upstream 

 and downstream of the fish was monitored simul- 

 taneously with two Yellow Springs Instruments 

 (model 51 A) dissolved oxygen meters equipped 

 with Clark-type, polographic electrode, oxygen- 

 temperature probes. Water flow through the res- 

 pirometer was maintained at approximately 3 1 

 kg ' (body weight) min"', and was measured by 

 recording the time required to fill a 1 1 graduated 

 cylinder. The source of the seawater was the same 

 as that which supplied the holding tanks. No at- 



tempt was made to control water temperature 

 which ranged from 23.5° to 25.5° C. 



Dissolved oxygen levels, water flow rate, heart 

 rate, red muscle temperature, and water tempera- 

 ture were determined every 10 min and measure- 

 ments were continued until the fish's metabolic 

 rate remained relatively stable for at least 1 h. 



Results and Discussion 



The SMR of each fish was determined by finding 

 the minimum predicted metabolic rate based on a 

 second degree polynomial fitted to observed 

 metabolic rate measurements. This method is an 

 acceptable approximation for asymptotic curve 

 fitting (Snedecor and Cochran 1967). To illustrate 

 this technique, the observed and predicted 

 metabolic rate, body temperature, and heart rate 

 are presented in Figure 1. 



Heart rate was monitored only as a check on the 

 healt+i of the animal. Experiments were termi- 

 nated if the heart rate became erratic or slowed 



200- i 



<i OBSERVED METABOLIC RATE 

 • RED MUSCLE TEMPERATURE 

 « HEART RATE 



PREDICTED METABOLIC RATE 



ELAPSED TIME (MIN) 



Figure l. — Changes in metabolic rate, heart rate, and red muscle temperature of a 1.456 kg skipjack 

 tuna during an experiment to determine standard metabolic rate (SMR), The predicted metabolic rate is 

 based on a least-squares fitted second degree polynomial. The minimum predicted metabolic rate (i.e,, 

 SMRl is .50.5mgO2h ', 



495 



