158 



Fist-iery Bulletin 88(1). 1990 



Figure 1 



Light micrograph of sagittal otolith of a 

 22 day-old (postfertilization) laboratory- 

 reared Atlantic menhaden larva showing 

 18 growth increments. The primordium 

 (p) is delineated hy an innermost protein- 

 rich layer. The nucleus (n) is delineated 

 by the first continuous growth increment 

 surrounding the primordium. The first 

 growth increment formed at hatching (h) 

 is characterized by a thick protein-rich 

 layer, and the first prominent growth in- 

 crement is formed at first feeding (fl"). 

 Scale bar represents 10 nm. 



in this experiment, but estimates in the literature in- 

 dicate that wild Atlantic menhaden transform to the 

 juvenile life stage at 25-30 mm (Lewis et al. 1972, 

 Nelson et al. 1977). 



Initiation and frequency of 

 increment formation 



The time to first increment formation and frequency 

 of increment deposition were estimated from linear 

 regressions of mean increment count on days after first 

 feeding. The regression intercepts of larvae from the 

 stock population and control and treatment groups 

 were not significantly different from zero, except for 

 the 1 -day starved treatment, indicating that the first 

 prominent increment is formed at or near the time of 

 first feeding (Table 1). 



Frequency of increment formation varied from 0.86 

 to 0.98 increments/day among experimental groups 

 (Table 1, Fig. 2). There did not appear to be any sys- 

 tematic effect of starvation on frequency of increment 

 formation. The rate of increment formation for both 

 the control and 1-day starved treatments were signif- 



icantly different from one increment/day, but the stock 

 population, 2-, and 3-day starved treatments formed 

 increments at a rate not sigTiificantly different from 

 one increment per day. Computations of statistical 

 power indicated low variability among the increment 

 count— age regressions and provided additional con- 

 fidence that growth increments are formed daily. The 

 test for homogeneity of slopes indicated a significant 

 difference between larvae from the control and pooled 

 treatments {p<0.05, n = 108). The difference between 

 the estimated slopes for individuals from the contrt)l 

 and 1-day starved treatment tanks accounted for this 

 difference. The estimated pooled slope for the control 

 and 2- and 3-day starved treatments was not signifi- 

 cantly different in the test for homogeneity among 

 slopes (;)>0.fi3, n =68). 



Response of increment width to changes 

 in larva feeding 



Since there were no significant differences in increment 

 width between duplicate containers within different 

 levels of treatments (/*> 0.05, ANOVA), measurements 



