FISHERY BULLETIN: VOL. 75, NO. 2 



fecundity by a gravimetric method, and found that 

 his results could be represented by the relation: 

 F = (4.8556 x 10 15)^6.33454^ w hereL = fork length 

 in millimeters. 



Snytko's (1971) fecundity observations on 171 

 specimens were the most extensive made to date in 

 the Washington-Oregon region. Snytko collected 

 his fecundity samples during November-March 

 1967-68, in the "Vancouver-Oregon region" (lat. 

 40°-50°N). The ovaries were collected before 

 fertilization of the oocytes had occurred and 

 fecundity was determined gravimetrically by 

 counting the oocytes present in 0.5- to 1.0-g sub- 

 samples of the ovaries (Snytko and Borets 1972). 

 Snytko ( 1971) presented his data in terms of mean 

 fecundity at a given length and regression of log F 

 on logL indicates that they can be represented by 

 the relation: F = (0.13103 x 10 5 )L 49883y , where 

 L = length in centimeters. 



Length-fecundity relationships for Pacific ocean 

 perch off Washington-Oregon, as predicted by 

 Westrheim (1958), Snytko (1971), and myself are 

 shown in Figure 22. There was substantial varia- 

 tion in the results obtained by different workers 

 and this is to be expected in view of the differences 

 in the timing of ovary collection, techniques used 

 to subsample and count oocytes, and the wide ex- 

 panse of time (1951-73) covered by the studies. 

 There is also a strong possibility that length- 

 fecundity differences exist between substocks 

 within the Washington-Oregon region and could 

 have contributed to these differences. 



The variability in the results of different work- 

 ers reflects only the difficulties in estimating the 

 number of maturing oocytes a given fish will pro- 

 duce and leaves a larger question unanswered. 

 What we would really like to estimate is the 

 number of viable larvae that fish of a given length 

 or age will give birth to during the embryo-release 

 period, and yet we are totally ignorant of the rela- 

 tionship between the estimated number of matur- 

 ing oocytes and the number of larvae that will 

 result from them. 



Preliminary estimates of the number of larvae 

 that will be released at each age can be made, 

 however, if it is assumed that all oocytes present 

 immediately after fertilization will develop into 

 viable larvae. It should be kept in mind that even 

 though this assumption is patently false, the re- 

 sulting estimates are still well-suited to be- 

 tween-area comparisons if oocyte-larval mortality 

 does not differ between areas. 



It will be recalled that fecundity observations 



394 



applied to fish collected during August- 

 September, while estimates of mean length at 

 each age applied to the September-December 

 period for Queen Charlotte Sound and to July in 

 the case of the WVI stock. The estimate of the 

 number of larvae released during March of any 

 given year of life (Table 17) was consequently ob- 

 tained by combining the age-length and length- 

 fecundity relationships pertaining to the previous 

 July-December. For example, the estimated 

 number of larvae released by 11-yr-olds in Queen 

 Charlotte Sound was estimated from predicted 

 mean length at age 10 (Table 3), and the length- 

 fecundity relationship appropriate to that stock {F 

 = 0.12240 x 10- 6 L 5 - 51258 ). 



RESPONSE OF PACIFIC OCEAN 

 PERCH STOCKS TO FISHING 



Methods Used to Examine the Effects 

 of Fishing 



In the past, management recommendations for 

 Pacific ocean perch in the INPFC Vancouver area 

 have been developed by arriving at some estimate 

 of the fishing mortality (F) that the stock can with- 

 stand, then applying this value to the best avail- 

 able estimate of stock biomass to arrive at a quota. 

 Much discussion has consequently focused on 

 what levels of F can be sustained. 



In this section, the effects of different levels of 

 fishing intensity on a hypothetical cohort of fish 

 will be examined, with an approach similar to the 

 yield per recruit analysis that is commonly used in 

 stock assessment. In contrast to conventional 

 yield per recruit analysis, however, I have at- 

 tempted to look at the costs involved in exerting 

 high levels of fishing intensity on a population, as 

 well as the benefits of increased yield. In particu- 

 lar, the decline in exploitable biomass (CPUE/g, 

 where q is the catchability coefficient) and popula- 

 tion fecundity that go hand in hand with increases 

 in yield have been evaluated quantitatively. 



The basic computations used to accomplish this 

 are shown in Table 16. Data required included 

 age-specific schedules of instantaneous natural 

 mortality, vulnerability to fishing, mean weight, 

 and fecundity (Table 17). The mean weight 

 schedule represents average values for the entire 

 year, while the fecundity schedule applies to the 

 embryo release period at the beginning of the year. 

 Vulnerability and mortality were assumed to be 

 constant throughout the year. 



