Z0° * 1° C. and saturated with oxygen. Sea 

 water of 33 ±1 p.p.t. salinity was used in all 

 experiments. The test fish were not fed for 

 Z days before an experiment and were accli- 

 mated to the experimental conditions in the 

 respiration chamber for 24 hours before the 

 first oxygen sample was taken. 



Comparison of Respiration Rates 



Knowledge of the metabolish of organisms 

 may be used to estimate their food or energy 

 requirements in ecological field studies. An 

 estimate of metabolism may be obtained by 

 measuring the oxygen consumption of fish. 

 This value, in conjunction with the known weight 

 of the fish, makes it possible to calcxilate re- 

 gression coefficients for the relation between 

 metabolism and weight. This relation between 

 metabolism and weight has two possible inter- 

 pretations. One is that the values for a and k 

 obtained from the experiments are typical for 

 all fish and indicate a general relationbetween 

 metabolism and weight. The other interpre- 

 tation is that metabolism varies with size in 

 different species. The "basic" equation, based 

 on the first interpretation, states that in gen- 

 eral the relation between total metabolism and 

 weight for all fish may be approximated by the 

 equation Q = 0.3Vf^-^. 



In these experiments, the "basic" equation 

 and its modification both were used to calcu- 

 late regression lines for the experimental data. 

 These two lines were then compared to the 

 regression line calculatedby the least-squares 

 method from the data obtained in this study. 

 For each species of fish examined, the mean 

 rate of oxygen consumption was plotted against 

 total body weight on a double logarithmic scale 

 (figs. 20 and 21). 



A value less than unity for the slope (k) of 

 the regression line for each species of fish 

 indicates that oxygen consumption per unit 

 weight is greater for smaller than for the 

 larger fish (table 5). This finding is in general 

 agreement with reports on other species of 

 fish where, with few exceptions, oxygen con- 

 sumption per gram decreased with increased 



weight. The k values ranged from 0.874 

 (croaker) to 0.537 (black sea bass). The cal- 

 culated k value for croaker, pinfish, and toad- 

 fish differed only slightly from the k value of 

 the "basic" equation. The slope values for 

 black sea bass and mummichog, however, devi- 

 ated considerably from the "basic" equation. 

 The difference in number of fish of each spe- 

 cies examined, size distribution, and relatively 

 high individual variationof oxygen consumption 

 made statistical comparison of respiration 

 differences difficult. Also, fish in different 

 stages of development cannot be connpared. 

 For example, a 10-g. mummichog would be 

 considered a mature fish and not comparable 

 to a 10-g. juvenile toadfish. Statistical com- 

 parisons, therefore, were made only among 

 black sea bass, pinfish, and croaker, since a 

 substantial number of fish (table 5) over a 

 comparable size range (10- 100 g.) were avail- 

 able for these three species. 



The calculated regression line for each spe- 

 cies of fish, extrapolated to 1 g., is shown in 

 figure 22. The data were tested by analysis of 

 covariance to determine whether a true dif- 

 ference existed among the slope of the three 

 regression lines. The hypothesis thatthe three 

 lines are parallel was rejected at the 0.01 

 probability level; the relationbetween metabo- 

 lism and weight apparently is not the same for 

 all three species of fish. 



The most accurate estimate of a and k for 

 each species was obtained from regression 

 lines calculated from the experimental data. 

 Neither the "basic" equation nor its modifica- 

 tion gave satisfactory regression lines for all 

 five species of fish, but the modification was 

 consistently closer to the calculated regres- 

 sion line (figs. 20 and 21). In the experiments 

 where it was possible to compare respiration 

 in three species of fish at about the same stage 

 of development (fig. 22), a statistical difference 

 among the slopes of the three regressionlines 

 was demonstrated. This test indicates that 

 metabolism does vary with size among the 

 species examined, and therefore one equation 

 will not describe the relation between metabo- 

 lism and weight for all species of fish. 



37 



