to distance as the unit-effort. This was evaluated qualitatively by 

 comparing the frequency distributions of the tow times and the time- 

 based CPUE statistics for before and after October 1977. Since the 

 extant program (unchanged since 1977) was distance-based, these were the 

 data primarily used in the evaluation analyses. 



Several approaches were used on the trawl catch data to elucidate 

 the underlying probability distribution. The frequency distributions of 

 the distance based CPUE were evaluated visually. The highly skewed data 

 were suspected of having come from a negative binomial distribtuion and 

 this possibility was investigated. To do this the mean, m, and a parameter, 

 k, must be known (Poole 1974) . The mean was estimated by the sample 

 mean x, while k was estimated by the maximum likelihood method (Bliss 

 and Fisher 1953). Scores, Z, were computed from trial values of k until 

 Z=0 in the following equation: 



Z = Z ( A x ) - N log (1 + (x/fc)) (1) 



£ + x 

 where A = accumulated frequency in all units containing 

 more than x organisms 

 x = sample mean 



