Radtke and Houngan: Age and growth of Nototheniops nudifrons 



565 



Figure 8 



Scanning electron micrographs of microincrements of sagittal 

 otoliths oi Nototheniops nudifro>is, as utilized for daily increment 

 enumeration, (a) Wide and narrow patterns of microincrements 

 from a 144-mmSL fish, (b) Narrow microincrements from a 

 70-mmSL fish. 



(A^=18), analysis of covariance indicated that the 

 slopes of the two lines did not differ between the sexes 

 (df=l, 26; F = 0.82, p>0.37). 



Relationship of fish size 



and otolith dimensions to fish age 



Stepwise multiple regression analysis relating fish size 

 (mm), weight (mg), and otolith morphometries (sagit- 

 ta length in mm and weight in ^g) to age (years), re- 

 sulted in the acceptance of the following model: 



Age (Yrs) = 3.892 + 0.856(sagitta wt) 



- 2. 206 (sagitta length) -i- 0.041(SL) 

 -0.025 (fish wt) 



Addition of the remaining variable (sagitta width) did 

 not improve the regression (N = 32, i?- = 0.967, 

 p<0.05; Table 3). The residuals were randomly dis- 

 tributed, which demonstrated that this multiple regres- 

 sion best explained the variability in age. Over 94% of 

 the variance was explained by the regression of the two 

 otolith measurements (weight and length). The vari- 

 ables in the model, such as otolith length, width, and 

 weight, are not strictly independent variables, nor are 

 length and body weight. This deviation from the as- 

 sumptions of the model may result in an artificially in- 

 flated R'^ value; however, this should not decrease 

 their practical value as predictors. Sagitta weight alone 

 accounted for 92% of the variance in the regression, 

 supporting the usefulness of otoliths as age-related 

 structures. 



The relationship between age and otolith morpho- 

 metries allowed the estimation of age for all the fish 

 in the sample (Fig. 11). The growth curve derived from 

 the multivariate model (Fig. 11) was comparable to that 

 derived from microincrement counts (Fig. 10), and both 

 were linear in form. Individuals in this population of 

 A'^. nudifrons grow at approximately 15 mm per year, 

 which is a slow growth rate. Analysis of covariance of 



