806 



Fishery Bulletin 88(4), 1990 



Figure 1 



otolith (a), vertt'bral section (b), and scale (c) from a L! + year-oM 

 l)luefish, 375 mm FL, captured in May. Distal edges of age marks 

 are indicated by Roman numerals. F = focus; T = translucent edge. 



slides and impressions were viewed on an Eherliach 

 scale reader. The most legible unregenerated scale was 

 examined for age marks. 



On scales and scale impressions an age mark was con- 

 sidered to be a band of widely spaced circuli. usually 

 witli broken circuli in the anterior field and/or ana- 



Figure 2 



Percent of otoliths with opaque edges liy month from (iulf of Mex- 

 ico bluefish, 1978-82 (»). and U.S. South Atlantic coast, 1981-82 ( + ). 



stomosis (crossing over) in the lateral field, followed 

 by a series of closely spaced circuli (Fig, ]c). Measure- 

 ments were made on the projected image from the 

 focus along a radius in the center of the anterior field 

 to the distal edge of the marks and scale edge. 



Two investigators each made one reading of a set of 

 otoliths, vertebrae, and scales from 100 bluefish. All 

 examinations were made without reference to fish 

 length or interpretations of the other investigator. 



Computer analyses and plots were accomplished 

 using SAS (Ver. 6,03) software, Backcalculations of 

 length-at-age were accomplished with a program writ- 

 ten by the author using means weighted by multi])lica- 

 tion of number of samples. Gulf and Atlantic bluefish 

 were analyzed separately. Separate analyses by sex in- 

 cluded regressions of fork length and natural log of fork 

 length on otolith length. Least-squares regressions 

 were used to backcalculate length-at-age for males 

 and females separately (Ricker 1975). Multiple regres- 

 sion of sex and age was run to determine difference 

 in length-at-age by sex. Von Hertalanffy theoretical 

 growth curves were calculated using weighted-mean 

 backcalculated fork lengths. The growth etiuation (von 

 Bertalanffy 1938, 1957) was the following: 



L(i 



-k(t- 



...)) 



where 



1, = length at age. 

 'oo = asymptotic length, 

 k = growth coefficient, and 

 t|, = time when length would theoretically be 

 zero. 



