FISHERY BULLETIN; VOL. 87, NO. 3, 1989 



tion between the end of MRN and the beginning 

 of AFT was named DAYTIME (DAY), and the 

 time duration between the end of DSK and the 

 beginning of DWN was named NIGHTTIME 

 (NIT). We examined the data collection of 15 

 years (1972-86) and estimated size- and time- 

 specific retention rates. Changes of these values 

 are due to the fish behaviors such as net frame 

 evasion and vertical migration. The rates were 

 calculated by taking ratios of catch/tow in each 

 time category and that of the NIT (Table 1). 

 Extrusion of larvae was examined as well, com- 

 paring observed numbers of fish with predicted 

 numbers calculated from a mortality curve (see 

 results section for detail). 



To reduce the seasonal effect of larval produc- 

 tion and mortality, we also considered the bias 

 on average catch/tow within a year from unpro- 

 portional sampling efforts among months. Net 

 tows were concentrated between February and 

 May; the percentage of juvenile fish in the total 

 catch was low in winter and increased in spring 

 and early summer (Fig. 2). Thus, we divided 12 

 months of the year into 4 seasons based upon the 

 abundance of saury larvae and juveniles: Jan- 



JFMAMJJASOND 



MONTH 



Figure 2. — Number of tows (columns) and percentage of 

 juvenile Pacific sauries (dots) in total fish captured by 

 months. 



uary to March as the main spawning season, 

 April to June as the late spawning season, July 

 and August as the off spawning season, and Sep- 



Table 1 . — Annual average of Pacific saury catch/tow (c/t) and size- and time-specific correction factors (CF) of 



size classes in 10 time categories. 



'SRS = SUNRISE; MRN = MORNING; DAY = DAYTIME; AFT = AFTERNOON; SST = SUNSET; ETW = EVENING 

 TWILIGHT; DSK = DUSK; NIT = NIGHTTIME; DWN = DAWN; MTW = MORNING TWILIGHT 



604 



