126 



Fishery Bulletin 97(1), 1999 



Limburg, 1995; Gallahar and Kingsford, 1996), and 

 anthropogenic pollutants (Kalish, 1995). All of these 

 factors may cause variation in levels of strontium in 

 wild fish. Hence, if strontium tagging is to be suc- 

 cessful in a region, the magnitude of natural varia- 

 tion should be investigated at appropriate spatial and 

 temporal scales. 



Strontium tagging has been used for freshwater 

 and anadromous species (Table 3), but investigations 

 of saltwater species are limited. Hurley et al. ( 1985) 

 marked the statoliths of the short-finned squid lllex 

 illecebrosus and Gallahar and Kingsford (1996) 

 marked the rock blackfish Girella elevata. Because 

 seawater has some 200-400 times the strontium con- 

 tent of freshwater (Guillou and de la Noue, 1987), 

 the range of concentrations available for tagging (i.e. 

 before saturation and precipitation ) is reduced for 

 marine fish. This limited our ability to batch mark 

 separate groups offish with different concentrations 

 of the same element, especially considering the un- 

 known factors affecting long-term persistence, such 

 as dilution due to growth. Multiple immersions in 

 strontium have, however, created multiple, discrete 

 bands on the otoliths of salmon, allowing the pro- 

 duction of codes for batch marking (Schroder et al., 

 1996). Another attractive possibility is to mark 

 batches of fish with different combinations of sev- 

 eral elements. A characteristic of ICP-MS analysis 

 is the ease in identification of a number of isotopes 

 with little corresponding increase in preparation 

 time. Schroder et al. ( 1996) found that rubidium was 

 incorporated into the calcified tissues of juvenile 

 salmon. Caesium, which has a similar ionic radius 

 and the same +2 valence as calcium, may also prove 

 to be a likely candidate for consideration. Pollard'^ 

 investigated iron chloride ( FeCl3-6H,,0 ) together with 

 strontium chloride, however the iron was not success- 

 fully incorporated into the dorsal spines of snapper. 



Advantages, limitations, and applications of the 

 technique 



The strontium immersion technique allows hundreds 

 offish to be tagged in a single batch without the need 

 for individual treatment. Other more labor intensive 

 methods may introduce problems of stress from fish 

 handling and are unsuitable for large numbers, es- 

 pecially the very large numbers involved in reseed- 

 ing programs. Stock enhancement in Japanese wa- 

 ters exceeds 15 million juvenile snapper each year 

 (Tsukamoto et al., 1989). Strontium tagging may be 

 applied to very small fish and even larval fish (e.g. 

 Behrens-Yamada and Mulligan, 1987). Fast body 

 growth and the difficulties of attachment make the 

 conventional tagging of small fish extremely diffi- 



cult. Some alternatives to batch tagging with stron- 

 tium and with comparable advantages include 

 microwire tags (Beukers et al., 1995), thermal mark- 

 ing (Schroder et al., 1996), oxytetracycline (Francis 

 et al., 1992; Lang and Buxton, 1993), alizarin 

 complexone (Lang and Buxton, 1993; Secor and 

 Houde, 1995) and other chemical fingerprints (e.g. 

 Gillanders and Kingsford, 1996). It may be useful 

 for some purposes to use a combination of techniques 

 simultaneously. 



The costs associated with the analysis of samples 

 on ICP-MS are relatively high compared with those 

 of many conventional tagging techniques, however 

 they may compare favourably with the costs of some 

 other chemical, electronic, or genetic methods of iden- 

 tification. The ICP-MS has very low detection lim- 

 its, generally 100 to 1000 times more sensitive than 

 inductively coupled plasma-atomic emission spectro- 

 scopy and inductively coupled plasma-atomic fluo- 

 rescence spectrometry (Horlick and Shao, 1992). 

 Sample preparation for ICP-MS is labor intensive; 

 around 100 fish samples require up to 2 days prepa- 

 ration and a half day for analysis. Strontium marks 

 however may also be detected by using other ana- 

 lytical techniques with lower costs, such as energy- 

 dispersive x-ray spectrometry or wavelength disper- 

 sive spectrometry (Schroder et al., 1995). Many other 

 analytical techniques, such as backscattered electron 

 microscopy (e.g. Schroder et al., 1995), or laser abla- 

 tion ICP-MS (e.g. Fowler et al., 1995), are also de- 

 signed to measure microconstituents at specific loci 

 across a section of calcified tissue, allowing the ap- 

 plication of strontium marking for the validation of 

 aging and batch marking of groups offish. 



Otoliths are commonly used for chemical marking 

 and are the most widely used structures for age de- 

 termination (Campana and Neilson, 1985). Otoliths 

 comprise discrete, directly comparable samples and 

 are recognized for their elemental stability (Edmonds 

 et al., 1989). However, fish must be sacrificed for 

 otolith sample collection. By contrast, spine and scale 

 removal are nondestructive techniques, allowing 

 catch and release techniques to be used for tagged 

 fish. Spines are not shed or replaced during growth, 

 unlike scales which may be lost, resorbed, or regen- 

 erated during fish ontogeny (Coutant and Chen, 

 1993). The small size of juvenile snapper scales also 

 creates counting, weighing, and manipulation difli- 

 culties (Pollard*). The dissection of otoliths or other 

 internal skeletal structures from large groups offish 

 is a labor intensive activity. The ease of removal of 

 spines or scales allows for the routine sampling of 

 these tissues regardless of the original purpose of 

 fish collection. Removal for sampling could take place 

 before the distribution of fish for commercial pur- 



