126 



Fishery Bulletin 94(1), 1996 



Age reading 



Microincrements were counted on transverse 

 sections of yellowfin tuna otoliths under a light 

 microscope (l,000x) with an Olympus micro- 

 scope BX 40 with a MPL lOOx dry objective. 

 Only 151 preparations were readable from the 

 170 fish because some were broken during 

 grinding or too deeply etched by EDTA. The 

 reading of microstructures was always made 

 on the external part of the transverse sections 

 along the ventral limb. 



Each transverse section was chosen ran- 

 domly and microincrements were counted at 

 three different times by the same reader with- 

 out knowledge of the sample identification or 

 previous counts. After the readings, 16 sections 

 were prepared for observation on a scanning 

 electron microscope (SEM). A photographic 

 series was made along each section (800x) to 

 obtain the whole reading area and to count 

 the number of microincrements. These results 

 were then compared with those obtained from the 

 light microscope reading. 



Growth 



Several growth models exist to describe the relation- 

 ship between fish size (FL in cm) and age it in year = 

 number of increments/365). Among the most used 

 models, three of them (von Bertalanffy, Gompertz, 

 and Richards) were tested. The equations of these 

 models are as follows: 



von Bertalanffy model: FL t = FL r ( 1 - e A '"- '» ') 



Gompertz model: FL t = FL exp(-ae~ Kt ) 



Richards model: 



FL=FL Al+e 



i -Kt + 6 1 \m 



is 

 15 " 

 12 

 9 h 



- 



n = 151 



1 



Ml 



60 90 



Fork length (cm) 



120 



150 



Figure 2 



Length frequencies of yellowfin tuna, Thunnus albo.ca.res, for which 

 otolith transverse sections were readable 



when FL t = fork length at age t; 



FL^ = asymptotic fork length; 



K = coefficient of growth; 



t Q = theoretical age for FL = 0; and 

 a, b, and m = parameters. 



Results 



The fork lengths of the fish that had readable otoliths 

 are shown in Figure 2. On the transverse section ( Fig. 

 3), we observed that the microstructures are more 

 visible in the layer ( 50-60 /urn of thickness) which is 

 located immediately under the otolith surface and 

 that interpretation on the internal face is difficult be- 

 cause the microstructures tend to be obscure (Fig. 4). 



Comparison of methods 



The relationship between the numbers of microin- 

 crements on the transverse section and on the ob- 

 lique section was highly significant (r=0. 992, P<0.05, 

 n=33) (Fig. 5). The slope of the regression was not 

 different from one (,=-0.943, P>0.05, df=31), and the 

 intercept was not different from zero </=0.266, 

 P>0.05, df=31). Therefore, transverse sections were 

 used for the age readings. 



Counts of microincrements on the transverse sec- 

 tion under a light microscope (400x) compared to 

 counts from SEM on the same section were signifi- 

 cantly different for fish larger than 100 cm fork 

 length (£=-4.643, P>0.05, df=18). To determine 

 whether the difference was the result of the magni- 

 fication used, we compared counts for 16 transverse 

 sections under a light microscope ( l,000x magnifica- 

 tion) and a SEM with a comparable magnification 

 (l,000x) (Fig. 6). The slope of the relationship was 

 not different from one ( £ =-0.426, P>0.05, df=14), and 

 the intercept was not different from zero ((=0.246, 

 P>0.05, df=14). These results suggest that these two 

 techniques give comparable counts. Therefore, sub- 

 sequent analyses were based on counts from the light 

 microscope ( l,000x). 



As the relationship between microincrement 

 counts from acetate replicas and transverse sections 

 could not be established for the same otolith, we plot- 

 ted the results (age versus fork length) obtained from 

 10 individuals for which age was determined by rep- 

 licas directly on the age versus fork length relation- 

 ship estimated on the 151 transverse sections (means 



