von Bertalanffy growth model 



Calculated -lengths-at-age from 1983 data were used to estimate the von Bertalanffy growth parameters 

 (Table 15). The growth models for females and males had good fits to the data and represented theoretical 

 growth of the population. Use of a nonlinear procedure for parameter estimation should have resulted in 

 the least biased and variable estimates in comparison with traditional linear methods used in most older 

 studies (Vaughan and Kanciruk 1982). The co parameter was used for comparisons of growth as suggested 

 by Gallucci and Quirm (1979). They suggested constructing a rectangle formed by two standard errors on 

 each side of point estimates of Lao and K. This was done for both sexes and because the rectangles did 

 not overlap, significant differences in growth were indicated between females and males. It was not possible 

 to make comparisons with other winter flounder populations because of the lack of published estimates 

 of variability. However, the co parameter of female Niantic River winter flounder was similar to those of 

 other stocks or geographical groups examined with the exception of Georges Bank, which has a racially 

 distinct population with much greater growth (Lux et al. 1970; Lux 1973; Howe and Coates 1975). The 

 value for males most closely corresponded to the Charlestown Pond stock (Berry et al. 1965); greater 

 asymptotic maximum length was achieved by winter flounder stocks in other areas to the east. 



The estimates of Loo were actually less than the lengths of some specimens examined in the Niantic 

 River. However, this should not be considered unusual as Loo represents the maximum length that an 

 average fish would achieve if it grew indefmitely (Fabens 1965; Ricker 1975). This could have been a 

 result of the particular sample used and inclusion of additional larger and older specimens could have 

 increased asymptotic length estimates. Nevertheless, since 1977 lengths of only 1.7% of 14,374 females 

 and 1.0% of 10,706 males larger than 20 cm exceeded the calculated values. 



The 1977-83 age-length data were also fit to the von Bertalanffy model (Table 15). Because sex of 

 age 1 and 2 and smaller age 3 fish was not ascertained in the field, these specimens were used separately 

 with both older females and males in fitting the model. Although growth of females and males was found 

 to be similar in early life, it is not known how the calculation of model parameters was affected by using 

 these data with each sex. Using the 1977-83 data, significantly larger estimates of Loo (453 mm for females; 

 397 for males) and smaller values of K (0.30; 0.34) were obtained in comparison with results using the 

 1983 calculated length-at-age data. Another error potentially affected results as the observed lengths were 

 taken over a period of several months each year and individuals could have been actually older or younger 



47 



