658 



Fishery Bulletin 88(4), 1990 



the oocyte. Clearly, yolk deposition is a conservative 

 process in that the overall composition of the yolk is 

 more dependent on the genetic programming of the 

 female parent than the environment. However, in 

 many calcium-binding proteins, such as vitellogenin, 

 the calcium moiety of the complexed molecule can be 

 substituted by strontium due to the similar structural 

 features of Ca^ ^ and Sr+ + (Skoryna 1981). The rela- 

 tive degree of this substitution would be largely depen- 

 dent on the relative concentrations of calcium and 

 strontium in the ambient environment. Typical Sr/Ca 

 ratios in marine waters (salinity SS-Am) are 0.0087 

 (0.09 mM/kg Sr:10.3 mM/kg Ca) (Bruland 1983) and 

 average 0.0019 (0.00068 mM/kg Sr:0.35 mM/kg Ca) in 

 freshwater (Rosenthal et al. 1970). These differences 

 would be expected to affect the Sr/Ca ratio of the yolk 

 and, ultimately, the composition of the developing 

 embryo and its otoliths. 



In a study of otolith and endolymph composition, 

 Kalish (1989) showed that the quantity of strontium in- 

 corporated into the otolith was directly related to the 

 quantity of strontium present in the endolymph, and 

 that anadromous brown trout collected in an estuary 

 had higher levels of strontium in both the endolymph 

 and otolith than non-anadromous brown trout. There- 

 fore, it seems likely that differences in the elemental 

 constituents of the yolk and embryos of anadromous 

 and non-anadromous salmonids would result in differ- 

 ences in otolith composition. This hypothesis is sup- 

 ported by research that shows little or no exchange of 

 calcium between prehatch salmoniti embryos and the 

 environment (Hayes et al. 1946, Zeitoun et al. 1976). 

 This would be expected since calcium present in the 

 yolk of the developing embryo is probably present in 

 a protein-bound form and is destined for the tissues of 

 the fry. Craik and Harvey (1984) found that the cal- 

 cium composition measured in whole rainbow trout 

 eggs was indistinguishable from the calcium measured 

 in protein precipitate obtained from the egg. This, of 

 course, excludes calcium that would be present in the 

 fluid of the perivitelline space following fertilization 

 (Laale 1980, Alderdice 1988). Strontium would prob- 

 ably behave similarly and there would be minimal loss 

 of any seawater-derived protein-bound ions in the yolk 

 to the freshwater environment where the development 

 of salmonid embryos takes place. 



In this paper, I discuss variations in the elemental 

 composition of the otoliths of non-anadromous and 

 anadromous salmonids with emphasis on the composi- 

 tion of the sagittal otolith nucleus. Several "life his- 

 tory" transects (scans of elemental composition across 

 an axis of an otolith made with a wavelength-fiispersive 

 electron microprobe) are presented to indicate the 

 variety of forms that these data may take in salmonids 

 with differing life histories. I also present the results 



of an experiment designed to test the hypothesis that 

 otolith primordia of the progeny of anadromous rain- 

 bow trout contain higher levels of strontium than the 

 otolith primordia of non-anadromous rainbow trout. 

 These data are examined in view of their usefulness 

 in distinguishing the progeny of sympatric non-anad- 

 romous and anadromous salmonids and in investiga- 

 tions of diadromous behavior. 



Methods 



Brown trout (non-anadromous Snimo fruffn) and sea 

 trout (anadromous Salmo trutta) were collected by gill- 

 net from the Derwent River and estuary, southeast 

 Tasmania, Australia. Juvenile and adult rainbow trout 

 were obtained from both wild and hatchery stock. 

 Atlantic salmon were obtained from hatcheries. Oto- 

 liths were removed from fresh fish, cleaned, and stored 

 in glass vials. Details of otolith preparation and micro- 

 probe analyses are described below. 



Ripe ova were obtained from ovulating freshwater 

 rainbow trout and sea-farmed rainbow trout and frozen 

 in plastic bags for later analyses of calcium and stron- 

 tium. Whole rainbow trout eggs were used for the 

 determinations. Groups of 10 eggs from 4 freshwater 

 and 4 sea-farmed trout were used. Preparation of eggs 

 was modified from the methods employed by Craik and 

 Harvey (1984). All solutions were made up with Milli-Q 

 water (Millipore Corp.). Groups of eggs were weighed 

 wet and then oven-dried at 50 °C for 24 hours to esti- 

 mate dry weight. The dried whole eggs were ashed at 

 500°C for 24 hours and then digested in 2.0 mL Aristar 

 ultrapure hydrochloric acid at 100°C for 2 hours. 

 Samples were separated into two e(iual aliquots for sep- 

 arate calcium and strontium determinations. Sample 

 digests for estimation of calcium were diluted in 10 mM 

 lanthanum chloride to suppress interference due to 

 phosphate binding. Calcium concentrations were deter- 

 mined by flame atomic absorption spectrojjhotometry 

 using an air-acetylene flame. 



Strontium concentrations in eggs were determined 

 by the method of standard additions using gi'aphite fur- 

 nace atomic absorption spectrophotometry on a Varian 

 AA-1475 spectrophotometer equipped with a GTA-95 

 graphite furnace and an autosampler. Argon was used 

 as the purging gas in the graphite furnace. Samples 

 were diluted with a 0.25% solution of an ionic deter- 

 gent, Triton X-100, and 20 /jL of diluted sample were 

 injected by autosampler into a walled, pyrolytically 

 coated graphite tube. Furnace conditions were: drying 



Reference to trade names does not in)()ly endorsoinenl by the 

 National Marine Fisheries .Service, N()A.\. 



