NEILSON and GEEN: CHINOOK SALMON OTOUTH INCREMENT FORMATION 



ered during this study and reported by Neilson 

 and G«en (1982), one otolith growth increment 

 was formed each day. This supports the hypothesis 

 that an endogenous rhythm influences growth in- 

 crement formation. Earlier, Neilson and Geen 

 (1982) reported that multiple feedings within a 

 24-h period resulted in the formation of >1 

 increment/24 h. We suggested that this resulted 

 from the interaction of an endogenous diel rhythm 

 of increment production and some regularly re- 

 curring environmental event. Data presented 

 here are consistent with that view, as increased 

 feeding frequency, exposure to a warm/cool tem- 

 perature cycle twice in 24 h and an enforced in- 

 crease in fish activity were all associated with an 

 increased rate of increment formation. The effects 

 of at least some of these environmental events on 

 otolith microstructure may be mediated through 

 activity-induced modification of fish metabolism, 

 which often follows a circadian rhythm (Matty 

 1978). If otolith growth increment production fol- 

 lows a circadian rhythm that is sometimes over- 

 lain by environmental events, it seems reasonable 

 to assume that fish may produce one or more 

 growth increments but not less than one every 24 

 h. In our studies, O. tshawytscha alevins and fry 

 produced one or more growth increments every 24 

 h, a result consistent with most earlier studies. 

 Even when fish were exposed to light and tempera- 

 ture stimuli with periods >24 h, Campana and 

 Neilson (1982) reported that only one increment 

 was formed every 24 h. 



Diel water-temperature fluctuations were not 

 required for otolith increment production in O. 

 tshawytscha. However, cyclic changes in tempera- 

 ture with a 24-h periodicity apparently result in 

 differences in the appearance of otolith growth 

 increments (Fig. 2). The deeply etched portion of 

 the increments is significantly wider in otoliths of 

 fish taken from a diel water-temperature regime 

 than those from fish held in water of constant 

 temperature. Mugiya et al. (1981) concluded that 

 the deeply etched portions of goldfish, Carassius 

 auratus, otoliths have a relatively high concentra- 

 tion of protein relative to calcium carbonate. 

 Degens et al. (1969) suggested that the deposition 

 of the organic matrix is not readily modified by 

 environmental events. If these results are appli- 

 cable to salmonids, the greater contrast in otoliths 

 of fish reared under a diel temperature regime 

 may result from changes in the rate of calcium 

 carbonate deposition. However, the presumed 

 change in composition and structure of daily 

 growth increments produced under various en- 



vironmental conditions does not affect the incre- 

 ment width-fish growth relationship illustrated in 

 Figure 8. 



Interactions between ration level and time of 

 feeding with respect to the 24-h temperature cycle 

 affected mean increment width. Ration level as a 

 single factor influencing increment width was not 

 significant. However, the interaction between 

 temperature and ration on increment width was 

 not significant suggesting higher calcium car- 

 bonate deposition on the otolith when temper- 

 atures were elevated at time of feeding. In a 

 two-way comparison with ration level and feed- 

 ing frequency (water temperature was constant), 

 increment width was affected by feeding fre- 

 quency but not by ration level. This agrees with 

 the results of Neilson and Geen (1982) who showed 

 that the rate of increment production is affected by 

 feeding frequency. 



Mean increment width reflected fish growth 

 under a variety of water-temperature and ration 

 regimes (Fig. 8). However, different equations de- 

 scribed increment width-growth relationships 

 under conditions that produced 1 increment/d or 

 >1 increment/d (Fig. 8). The extent to which in- 

 crement width data can be used to predict instan- 

 taneous growth rates in natural populations re- 

 mains to be examined. 



Increment widths can provide an indicator of 

 environmental changes and consequent alteration 

 of growth rates. However, such changes, at least 

 under laboratory conditions, did not occur rapidly 

 (Fig. 6). These data suggest that at least 3 wk 

 would be required before the change in increment 

 width would be statistically detectable. 



Our data indicate that otolith weight-fish 

 weight regressions are similar under a range of 

 experimental conditions suggesting that otolith 

 growth in salmon fry is closely coupled to fish 

 growth. Marshall and Parker (1982) also reported 

 that differences in ration and water temperature 

 did not significantly affect slopes of otolith size- 

 fish size regressions among fed sockeye salmon, O. 

 nerka, fry. Exceptions to the isometric growth re- 

 lation between fish size and otolith size have only 

 been observed in recently hatched salmonid ale- 

 vins (Neilson unpubl. data) and in starved O. 

 tshawytscha fry. Fry deprived of food for 19 d con- 

 tinued to form daily growth increments. Assum- 

 ing fish dry weight did not increase over this 

 period, then the slope of the otolith weight-fish 

 weight regression would probably be greater than 

 for fed fish. Marshall and Parker (1982) also re- 

 ported continued otolith growth in O. nerka fry 



99 



