FISHERY BULLETIN: VOL. 83, NO. 1 



UJ 



o 

 or 

 o 



(/) 



Z) 



o 



to 



3.0 ^ 



2.5-1 



2.0- 



1.5- 



1.0 - 



0.5 



0.0 



8% (warm) 



,4% (cool) 



4%X2 >^ 4% (warm) 



D 4% (constant) 



8% (cool) 



2% X 2 



4% + activity 



o 



8% (constant) 



— 1 1 1 1 



AC o Q ^ '^ 



MEAN INCREMENT WIDTH (urn) 



3.5 



4 



Figure 8. — Regressions of mean otolith increment width versus instantaneous growth rate (dry weight) for the various 

 experimental regimes. Solid line represents groups where fish produced one increment every 24 h on average; dashed line 

 represents treatments where fish produced significantly more than one increment every 24 h. 



cantly in both slope and y-intercept (analysis of 

 covariance and Mest, P < 0.01) from that of fish 

 fed once per day. 



Slopes of arithmetic mean linear regressions of 

 fish dry weight on experiment day indicated that 

 the average rate of growth offish fed 8% B.W./24 h 

 at the beginning of the warm portion of the diel 

 temperature cycle was significantly greater than 

 that offish fed at the beginning of the cool period of 

 the diel temperature regime or at the constant 

 water temperature (6°C) (analysis of covariance 

 and the Student-Newman-Keuls test, P < 0.01). 

 Similar analyses among treatments in which fish 

 were fed 4% B.W./24 h [4% (warm), 4% (cool), 4% 

 (constant)] or among fish that received two feed- 

 ings or one feeding coupled with an additional 

 bout of activity (2 x 4%, 2 x 2%, 4^^ + activity) 

 indicated no significant differences in growth rate 

 (P> 0.05). 



To determine whether otolith growth-fish 

 growth relationships were similar among treat- 

 ments, we calculated otolith weight-fish dry 

 weight regressions for data from all experimental 

 regimes. Analysis of covariance indicated that the 



98 



slopes of the predictive regressions among groups 

 of fish fed 8% B.W./24 h and exposed to different 

 temperature regimes did not significantly differ 

 from each other (P > 0.1). Nor were there signifi- 

 cant differences among treatments in which fish 

 were fed twice/24 h or fed once/24 h and exposed to 

 an enforced 10-min bout of activity. The slope of the 

 regression representing the otolith weight-fish 

 weight relationship for those fish receiving a ra- 

 tion of 4% B.W./24 h on the cool portion of the diel 

 cycle was significantly greater than the slopes of 

 regressions representing fish fed 4% B.W./24 h 

 (warm or constant) (analysis of covariance and the 

 Student-Newman-Keuls test, P < 0.05). However, 

 as mentioned earlier, the treatment where fish 

 received a ration of 4*^ B.W./24 h on the cool por- 

 tion of the diel temperature cycle was affected by 

 an interruption in water supply. Only five fish 

 survived to day 40 and may not have been repre- 

 sentative offish held under those conditions. 



DISCUSSION 



Under most environmental conditions consid- 



