214 



Fishery Bulletin 93(2). 1995 



3, bottom graph) corresponded with declining tank 

 temperatures (beginning in September; Fig. 4). How- 

 ever, age could still be estimated within about 3 

 weeks up to 300 days after hatching (230 days post- 

 mark plus about 70 days in age at marking; Fig. 3). 

 If this relationship were consistently repeatable, it 

 would still be a useful tool for estimating ages of older 

 juveniles, even though the age-ring count relation- 

 ship was well below 1:1. However, we suspect that 

 this change was not so much a function of age as it 



160 

 140 

 120 

 100 



1 00 1 50 



Days postmark 



Figure 3 



Estimated number of otolith growth increments against 

 elapsed time in days (known age) or days postmark 

 (oxytetracycline [OTC] or alizarin complexone [ALC]) of 

 juvenile Atlantic menhaden, Brevoortia tyrannus. Results 

 from oblique-transverse sections (open circles) and trans- 

 verse sections (closed circles) are pooled for the regression 

 lines shown. The regression coefficients are given in Table 

 1. (Results for the alizarin-complexone trial for days 161- 

 237, where increment formation rates were less than daily, 

 are shown on the upper right of the bottom graph. Regres- 

 sion parameters for the oblique-transverse (dashed line) 

 and the transverse (solid line) data sets respectively, are 

 r 2 =0.81 and 0.79, intercept=39.62 and 62.76, and 

 slope=0.73 and 0.52.) 



was a result of reduced growth rate, possibly in con- 

 junction with declining temperature, which affected 

 our ability to accurately estimate ages. Savoy and 

 Crecco ( 1987 ) also showed that reduced rearing tem- 

 peratures can reduce growth rate and subsequently 

 result in a count-age slope below 1.0 for larval Ameri- 

 can shad, Alosa sapidissima. 



The growth rates offish treated with ALC (through 

 day 131 post-ALC-mark) are comparable with obser- 

 vations for upper growth rates of juveniles in tidal 

 creeks, spring through fall (0.7 to 0.83 mmd" 1 ; 

 Kroger et al., 1974). The laboratory-reared fish had 

 lower growth rates, but their rates were still greater 

 than those for the ALC fish following postmarking 

 day 131. Therefore it appears that reduced growth 

 rate was a contributing factor for the higher vari- 

 ance of our estimates of the slope of counts versus 

 days for our known-age and OTC test groups. Simi- 

 larly, the variances for the ALC group were highest 

 during the period when increments displayed a less 

 than daily periodicity (Fig. 3). 



May 



July 



Sep. 



Npv. 



Dec. 



Figure 4 



(A) Mean fork length of juvenile Atlantic menhaden, 

 Brevoortia tyrannus (error bars represent ±1 standard 

 deviation ),and (B) tank water temperatures for the alizarin 

 complexone rearing trial (see bottom graph on Fig. 3). 



