Clear et a\. Validation of annual increments in otoliths of Thunnus maccoyii 



37 



pretation are consistent with this age. Therefore, we 

 beheve that the interpretation of Itoh and Tsuji (1996) 

 and Rees et al.^, that 50-cm-FL fish are one-year-olds, 

 is most Ukely correct. 



The identification offish that we considered to be one- 

 year-olds as two-year-olds was made by Thorogood in 

 his 1987 study. However, we have found no evidence of 

 two increments in the otoliths of new recruits. The early 

 zones on all axes of otohth growth are difficult to read 

 on some otoliths, and the increments in these areas 

 are less distinct than those deposited later. In some 

 fish a poorly defined "band" is also present very close 

 to the primordium (within 2 mm along the postrostral 

 axis). Although these two factors may confuse an 

 inexperienced reader, Thorogood makes no mention of 

 difficulty in reading the first increment. An alternative 

 explanation for Thorogood's interpretations may be 

 that his readings were influenced by the findings of 

 Hynd (1965) and Yukinawa (1970) that were based on 

 scales, because their estimates of age at recruitment 

 were entrenched within the dogma of SBT population 

 dynamics current in the 1970s and 1980s. 



It has been hypothesized that more than one trans- 

 lucent zone forms per year in the otoUths of mature 

 Atlantic bluefin tuna, Thunnus thynnus (Berry et al., 

 1977; Lee et al., 1983). In females, one translucent zone 

 may correspond to a winter slow-growth period, the 

 other to a spawning period (Lee et al., 1983). In the 

 two large, tagged fish examined in our study, only one 

 opaque and one translucent zone were deposited per 

 year throughout hfe. The outer increments (i.e. those 

 assumed to be deposited after sexual maturation) were 

 consistent in both their width and optical density and 

 were visually equivalent to the increments described by 

 Lee et al. ( 1983), comprising a wide opaque region and 

 a narrow translucent area that, on a black background, 

 appears dark under reflected light. Occasionally, there 

 appeared to be two translucent zones closer together 

 than normal and, if these bands coalesced at the margin, 

 they were counted as part of the same increment. These 

 may be equivalent to the bands described by Berry et 

 al. ( 1977 ) who hypothesized that a pair of these paired 

 bands represented an annual increment. The close 

 agreement between otolith increment counts and bomb- 

 radiocarbon age estimates for mature SBT up to 34 

 years old (Kahsh et al., 1996) supports our hypothesis 

 that one increment, comprising one translucent and 

 one opaque zone, continues to form per year, as does 

 the consistency of the width and optical density of 

 increments deposited after sexual maturation in the 

 otohths aged by Kalish et al. (1996). In summary, there 

 is no significant evidence to suggest that mature female 

 SBT deposit two translucent zones per year. In this 

 regard our findings are similar to those of Hurley and 

 lies (1983) and Hurlbut and Clay (1988) for T. thynnus; 



they found, albeit in the absence of direct validation, 

 that a single translucent zone is laid down per year in 

 medium- and giant-size classes. 



The use of strontium chloride to mark otoliths of 

 large fish 



This study has shown that intramuscular injection of 

 strontium chloride leaves a distinct mark on the oto- 

 liths of SBT that is clearly visible as an SEM back- 

 scatter image in the Robinson detector. In the 20 oto- 

 lith sections from Sr-injected fish that we examined, 

 95% had detectable marks. On this basis, we conclude 

 that the compound is an efficient marker. Success of 

 OTC as a marker at this rate of detection (95%) leads 

 to high mortalities (McFarlane and Beamish, 1987). 

 The high detection rates and lack of evidence of mor- 

 tality for SrClg are not surprising. This mineral occurs 

 naturally in sea water, the mean concentration being 

 3.8-8.2 ppm (Carriker et al., 1991) or 0.09 mM/kg 

 (Bruland, 1983), and both Sr and CI are major con- 

 stituents of the otoliths of SBT (Gunn^). When SrCU 

 is injected into the muscle it is taken up into the blood 

 stream and incorporated in the endolymph and then 

 the otolith, substituting for Ca within the CaCOg por- 

 tion of the aragonite. The combined weight fraction 

 of Ca and Sr ( approximately 42% ) within the otolith 

 does not change as a result of the injection. However, 

 the Ca:Sr ratio changes fi-om 250:1 before injection 

 to as low as 5:1 during the period over which the Sr 

 spike induced by the injection is metabolized. At a dis- 

 tance of 6 pm beyond the injection mark, the Sr lev- 

 els have dropped to about 50% and at 15 pm to 10% 

 of peak values (Fig. 6). These distances correspond to 

 time periods in the order of 2 and 5 days, respectively, 

 based on median growth rates of around 3.0 pm/day 

 estimated along this axis (Rees et al.^). 



Detecting Sr marks on whole otoliths by using 

 EDS was possible because the growth plane of tuna 

 otoliths lies near the surface of the distal face. In 

 otoliths of young fish, etching will expose the growth 

 plane, so that sectioning is not required. Although this 

 method of detection was slightly less successful (85%), 

 it had two advantages. First, the preparation time was 

 around half that required to prepare sections suitable 

 for the Robinson detector. Second, the age estimate 

 and measurements of increments were made along 

 the postrostral axis on whole otoliths from the smaller 

 fish (up to six years old), and the method of locating 

 the strontium mark by EDS meant that the position 

 of the strontium mark was measured along the axis in 

 the same plane. With the Robinson detector, the same 

 axis was measured but in cross section. 



In fish older than about 6 years, the increments de- 

 posited on the margin can be unclear on whole oto- 



