Smith et al.: Distribution and biology of Zaprora silenus 



173 



30 70 110 150 190 230 270 310 



Bottom depth (m) 



350 390 430 470 



Figure 3 



Average prowfish catch per unit of effort (CPUE; no/km 2 ) within 20-m intervals over the 

 range of trawl depths of 20-479 m in the Gulf of Alaska (GOA) in 1996 (A) and 22-474 m 

 in the Aleutian Islands (AD in 1997 (B). The mean of interval midpoints, each weighted 

 by interval average CPUE, was 163.8 m for the GOA and 150.3 m for the AI. 



between sexes were not significant for either the GOA 

 (P=0.83) or the AI (P=0.76). Although the weighted mean 

 for both sexes combined was 61.0 cm (range: 11-90 cm) in 

 the GOA and 51.9 cm (range: 25-87 cm) in the AI, the differ- 

 ence in length between regions was not significant (P=0.1 1 ). 

 Grouping of prowfish of similar size within hauls was highly 

 significant in both the GOA and the AI (P«0.01 ). 



Weight-length relationship 



In the between-sex ANCOVA comparison of the linearized 

 (i.e. log-transformed) weight-to-length relationships based 

 on prowfish caught in the Gulf of Alaska, the slopes were 

 not significantly different between sexes (P=0.38). How- 

 ever, the difference in intercepts was significant (P=0.044 ). 

 Thus the best fitting model varied by sex with two regres- 

 sion lines of equal slope but with sex-specific intercepts. 

 The equivalent functions in terms of the untransformed 

 variables (Fig. 5) were 



and 



W males = 0.0164 xL2922 ; 



W /mwles = 0.017x7^22. 



The model indicated that adult females are, on average, 

 3.7% heavier than males of the same length. 



Age and growth 



Readable otolith specimens were produced for 138 prowfish 

 (71 males, 67 females) of the 172 from which samples were 

 collected. Production of readable specimens did not appear 

 related to fish size or age. The likelihood ratio test for a 

 difference between males and females in the relationship 

 of length to age was not significant (P=0.53), indicating 

 that there was no difference in growth between sexes. The 

 best-fit von Bertalanffy function (Fig. 6) had the following 

 parameters (with 95% confidence intervals): L a = 89.33 

 ±6.5 cm; k = 0.18 ±0.05/year; and t () = -0.55 ±0.12 year. 



Female maturity 



The proportions of females that were mature were highly 

 significant logistic functions of length and age (P<0.005; 

 Fig. 7). The fitted functions of length and age were 



and 



P mo ,= l/(l+e 37114 - 6 - 51L ); 



1/(1 +e 9.66-1.90(). 



The theoretical length and age at which 50%' of females 

 were mature, with respective 95% confidence limits, were 

 57.0 ±0.4 cm and 5.1 ±0.7 years. 



