KalisiT Otolith microchemistry to distinguish salmonicl progeny 



663 



Discussion 



The results present evidence that the Sr/Ca ratios of 

 otolith primordia formed during embryonic develop- 

 ment are directly influenced by the strontium content 

 of the individual's yolk and that yolk composition is also 

 influenced by the composition of the waters where 

 vitellogenesis took place. These findings are important 

 to the study and management of salmonid species 

 where anadromy is a faculative behavior. They may 

 also be important to studies of other diadromous spe- 

 cies, but it remains to be seen whether the relatively 

 small nutritive contribution of the yolk in these fishes 

 is adequate to influence otolith composition. However, 

 since otolith primordia develop very early in the ontog- 

 eny of fishes (Brothers 1984). there may be detectable 

 differences in otolith strontium content within other 

 species that display facultative diadromy. In many 

 cases, the analysis of otolith primordium composition 

 should make it possible to investigate the relative im- 

 portance of parentage and environment in determin- 

 ing diadromous behavior. 



The basis for the differences in otolith primordium 

 composition among anadromous and non-anadromous 

 fishes is such that there is probably little, if any, effect 

 of race or population. Both wild and hatchery salmonids 

 involved in this study displayed similar ranges of ele- 

 mental ratios, depending on habitat (freshwater or 

 marine) indicating that different salmonid species may 

 incorporate into their otoliths a similar proportion of 

 the ions present in the endolymph. Furthermore, this 

 indicates that the relationship between endolymph and 

 blood plasma composition is regulated in a similar man- 

 ner in each of these species. Kalish (1989) showed that 

 there was a tendency for both endolymph and otolith 

 Sr/Ca ratios to be similar among individuals of a spe- 

 cies, whereas differences were greatest among species. 

 On the basis of the data presented here and in Kalish 

 (1989), it seems likely that freshwater and marine 

 salmonid otolith Sr content couFd be predicted on the 

 basis of endolymph composition. 



The results of this study confirm that differences in 

 the composition of freshwater and seawater can be 

 reflected in the composition of fish otoliths. These find- 

 ings are similar to those obtained by Casselman (1982), 

 Radtke et al. (1988). and Kalish (1989). Wavelength- 

 dispersive electron microprobe analyses of strontium 

 content in diadromous fish otoliths can provide infor- 

 mation on the rates of migration to and from the sea, 

 residence times at sea and in freshwater, the age at 

 which migrations take place, and confirmation of data 

 obtained from scales. 



Confirmation of data obtained from scale reading is 

 important because of the possibility of scale resorption 

 (Bilton 1974) and the equivocal nature of data obtained 



from scales in the discrimination of adult anadromous 

 and non-anadromous salmonids. Bagenal et al. (1973) 

 used the strontium content of whole scales to distin- 

 guish between brown trout and sea trout, and Moreau 

 and Barbeau (1979) used a similar rationale to dis- 

 criminate anadromous from non-anadromous whitefish 

 Coregonus clupeaformis. However, Castonguay and 

 FitzGerald (1982) found that this method was unreliable 

 for distinguishing between anadromous and non-anad- 

 romous brook char Salveliriusfontinalis, and Gausen 

 and Berg (1988) had similar results when investigat- 

 ing migratory and non-migratory Atlantic salmon. 

 Measurements of otolith Sr may provide a more reli- 

 able means of determining migratory behavior in these 

 species because otoliths are not resorbed during peri- 

 ods of stress (Simkiss 1974, Campana 1983), although 

 Mugiya and Uchimura (1989) present evidence for 

 otolith resorption in goldfish Carassius auratus dur- 

 ing extreme anaerobic stress. Also, otolith Ca and, 

 most likely, Sr are derived from ions taken up by the 

 gills (Simkiss 1974), and any variations in the Sr con- 

 tent of the diet would not be reflected in the composi- 

 tion of the otoliths. This is not the case for scales, and 

 differences in the Sr content of the diet as well as the 

 effect of different levels of discrimination against Sr 

 would be manifested in scale composition. 



It is important to point out the inadequacy of energy- 

 dispersive (ED) electron microprobe analysis for the 

 determination of Sr at the levels typically found in 

 marine fish otoliths and particularly in the otoliths of 

 freshwater species. Under optimal operating condi- 

 tions, the minimum detectability limit for a wavelength- 

 dispersive (WD) electron microprobe is approximately 

 10 times lower than that of an ED microprobe for all 

 elements (Geller 1977, Goldstein et al. 1981). General- 

 ly, the minimum detection limits attainable with an ED 

 spectrometer are on the order of 0.10% wt (1000 ppm), 

 and these conditions are frequently not realized due to 

 various factors. The range of Sr concentrations mea- 

 sured in this study was approximately 300-5000 ppm, 

 and all marine fish otoliths investigated to date appear 

 to contain Sr in excess of 1000 ppm (Radtke 1987, 

 Edmonds et al. 1989, Kalish 1989). Theoretically the 

 ED spectrometer should be capable of making all oto- 

 lith Sr measurements in marine fish and some of those 

 from freshwater fish. However, the detection and ac- 

 curate quantitative estimation of Sr creates special 

 problems for the ED microprobe. 



The X-ray line that is used to detect Sr at 1.806 keV 

 (L^) is affected by the silicon escape peak associated 

 with the large Ca K^ peak produced when measuring 

 specimens that are largely calcium (aragonite fish oto- 

 liths are approximately 38% wt Ca). The silicon escape 

 peak results from the production of Si K^ photons 

 following absorption of relatively high energy (> 1.841 



