FISHERY BULLETIN: VOL. 83, NO. 1 



et al. 1975). However, otolith nucleus dimensions 

 did serve to separate summer and winter races of 

 steelhead trout (McKern et al. 1974). Workers 

 proposing to use otolith nucleus dimensions as 

 stock identification criteria should consider rear- 

 ing fish under controlled conditions to establish 

 the extent of nucleus size variability in the stocks 

 in question. 



Otolith nucleus length is also influenced by 

 water temperature during embryonic develop- 

 ment. Our data showed an increase of about 25% 

 in length in fish reared at 9.5° or 15°C relative to 

 that observed in fish incubated at 6.5°C. The sen- 

 sitivity of otolith nucleus length to water tempera- 

 ture may allow separation of selected fish stocks 

 whose eggs are incubated at different water tem- 

 peratures. For example, O. tshawytscha juveniles 

 originating from Campbell River stock reared in 

 the Canada Department of Fisheries and Oceans 

 Quinsam Hatchery on Campbell River had sig- 

 nificantly greater otolith nucleus lengths (P < 

 0.01) than wild Campbell River O. tshawytscha 

 incubated in cooler waters (M. Bradford pers. 

 commun.^). Increased water temperature may in- 

 fluence nucleus length through a greater rate of 

 accretion of the calcium/protein matrix around 

 primordia, reflecting a faster rate of embryonic 

 development. 



The definition of otolith nucleus suggested here 

 can be consistently applied. With relatively simple 

 preparation techniques, otolith nucleus dimen- 

 sions can be measured from micrographs or by 

 using a light microscope equipped with an ocular 

 micrometer Previous workers have delimited the 

 otolith nucleus in relation to metamorphic or 

 nuclear checks. Such terms are ill-defined and 

 should be avoided since they imply that otolith 

 checks result from important developmental 

 events. While it seems likely that such events may 

 result in growth interruptions or checks, causal 

 links have not yet been demonstrated. 



The imprecise definition of the periphery of the 

 otolith nucleus may reduce the comparability of 

 measured dimensions derived in various studies. 

 While we have defined the nucleus as lying within 

 the first increment surrounding the primordia, 

 several checks occur during early otolith develop- 

 ment. Use of one of these checks to define the 

 periphery of the nucleus would result in inconsis- 

 tency between various investigations. For exam- 

 ple, nucleus lengths of steelhead trout used in this 



*M. Bradford, Department of Biological Sciences, Simon 

 Fraser University, Burnaby, B.C., Canada V5A 1S6, pers. com- 

 mun. November 1983. 



88 



study were generally <0.2 mm (Fig. 3). The mean 

 diameter of the otolith nucleus of summer and 

 winter steelhead reported by McKern et al. (1974) 

 were 0.348 and 0.436 mm, respectively. Differ- 

 ences between studies of this magnitude may be 

 racial in nature or may reflect differences in defi- 

 nition of the extent of the nucleus. 



Data on variation in primordia number and lo- 

 cation have not been reported previously although 

 the existence of primordia was described by 

 Radtke and Dean (1982) in mummichogs. McKern 

 et al. (1974) did not describe primordia in their 

 work involving the otolith nucleus in steelhead 

 trout. Their results were based on the use of X-ray 

 techniques. We were not able to detect primordia 

 using this method. 



It is likely that the otoliths of many fish species 

 are formed by fusion of multiple primordia. From 

 our observations, this is apparently the case in all 

 five species of Pacific salmon and the Pacific her- 

 ring, Clupea harengus pallasi. Radtke and Dean 

 (1982) noted multiple primordia in masou salmon, 

 O. masou; Arctic char, Salvelinus alpinus; brook 

 trout, S. fontinalis; and the sculpin, Cottus 

 nozawa. 



While both steelhead trout and O. tshawytscha 

 otolith nucleus areas were variable, otolith areas 

 in older fish (longer than 15 d after primordia 

 fusion) were less so as indicated by the decreasing 

 coefficient of variation of otolith area with increas- 

 ing age (Table 1). The decreased variation proba- 

 bly reflects the development of otoliths from an 

 indeterminant array of primordia to the otoliths of 

 adult fish, the latter considered a species-specific 

 characteristic (Fitch 1968; Morrow 1979). How- 

 ever, variation in otolith development in the 

 juvenile salmonids studied here do not present 

 difficulties for the interpretation of microstruc- 

 ture as neither the number nor width of growth 

 increments is significantly affected by nucleus 

 size variation. 



ACKNOWLEDGMENTS 



The cooperation of the staff at British Columbia 

 Department of the Environment fish hatcheries at 

 Abbotsford, Loon Lake, and Summerland is grate- 

 fully acknowledged. In particular, we wish to 

 thank John Cartwright, Dennis Graf, Chris Hous- 

 ton, Bob Land, Don Peterson, and Hugh Sparrow 

 for their exceptional support. Eldon Stone of the 

 Canada Department of Fisheries and Oceans 

 Capilano Hatchery provided the chinook salmon 

 used in this study. Mike Bradford suggested calcu- 



