FISHERY BULLETIN: VOL. 78, NO. 3 



weight (Table 6). These two expressions of relative 

 fecundity were calculated to allow comparisons 

 with results of previous studies and to illustrate 

 how each method might bias fecundity estimates. 

 Mean relative fecundity based on total body 

 weight of fish in ovarian stages 6-8 and 10 was 

 similar. But mean relative fecundity based on 

 ovary-free body weight in these two groups dif- 

 fered by nearly 100 oocytes. Because oocyte hydra- 

 tion substantially increases ovary weight in ripe 

 fish, the best (least biased) estimate of mean rela- 

 tive fecundity, when individual estimates are 

 based on fish captured both prior to and after oo- 

 cyte hydration, is ooctyes per unit ovary-free body 

 weight. Yet either expression of relative fecundity 

 can be biased if there are changes in somatic 

 weight associated with the reproductive cycle or 

 changes in condition, i.e., the length-weight rela- 

 tionship from year-to-year or among geographical 

 regions (Bagenal 1967). 



Linear and exponential (based on logjo trans- 

 formed variates) expressions yielded similar fits to 

 the relationship between fecundity (TF) and total 

 body weight (W) and between TF and standard 

 length (SL) with r values ranging from 0.73 to 0.82 

 (Table 7). The linear equations describing the re- 

 lationship between TF and ovary weight (OVW) 

 also yielded high r values, 0.92 for stages 6-8 fish 



Table 6. — Mean fecundity (± SE) oiEngraulis mordax collected 

 off the Oregon-Washington coast. Tbtal fecundity = number of 

 advanced oocytes in ovaries; relative fecundity = (1) number 

 advanced oocytes per gram total body weight (ovary weight 

 included) and (2) number advanced oocytes per gram ovary-free 

 body weight. 



Classification 



Ovarian 



stages 6-8 



(11 fish) 



Ovarian 

 stage 1 

 (10 fish) 



Ovarian 

 stages 6- 1 



(21 fish) 



Total fecundity 

 Relative fecundity (1) 

 Relative fecundity (2) 



17,335±2,525 

 726 ±65 

 782 ±75 



16.267 ±1,881 

 712±43 

 873 ±63 



16,826 ±1,563 

 720 ±40 

 826±49 



Table 7. — Functional (geometric mean) regression equations, 

 sample size (N), standard error of the regression coefficient (SE), 

 and correlation coefficient (r) for the relationship between total 

 fecundity {TF) and total body weight (IV), standard length iSL), 

 and ovary weight ( OVW) in Engraulis mordax collected off the 

 Oregon-Washington coast. 



Equations 



SE 



Ovarian stages 6-10: 



TF= -13,889.91 + 1,339.57W 

 LogioTF =1.91-1- l.69logiolV 



TF = -76,286.31 -i- 738.99 SL 

 Logio TF = -6.80 + 5.23 logio SL 

 Ovarian stages 6-8: 



TF = -1,181.63 + 11,506.73 OVW 

 Ovarian stage 10: 



TF = -2,654.20 + 4,637.89 OVW 



and 0.96 for stage 10 fish. The apparent difference 

 between the slopes of these two equations, 

 11,506.73 for stages 6-8 and 4,637.89 for stage 10, 

 can be explained by the substantial increase in 

 ovary weight in stage 10 fish caused by oocyte 

 hydration, and not by an actual decrease in 

 number of oocytes in these fish. Fecundity would 

 be underestimated if an equation relating TF to 

 OVW in fish with hydrated oocytes was used to 

 predict fecundity. The relationship between TF 

 and OVW, if based on fish captured both prior to 

 and after oocyte hydration, would yield a low cor- 

 relation coefficient (r). 



Spawning Frequency 



The number of times a female anchovy in the 

 northern subpopulation spawns during the year 

 could not be determined directly with available 

 data. Oocyte observations, however, provided 

 some information pertinent to the question of 

 spawning frequency in these fish. 



Degenerating, immature, yolked oocytes (those 

 with abnormal-looking nuclei) were found in ripe 

 northern anchovies during the spawning season. 

 Although relative numbers of these oocytes were 

 not determined, their presence suggests that oo- 

 cytes in early stages of vitellogenesis in July may 

 eventually degenerate and be absorbed. Higham 

 and Nicholson (1964) also found disintegrating in- 

 termediate and maturing oocytes in the ovaries of 

 recently spent Atlantic menhaden, Breuoortia 

 tyrannus, indicating perhaps that this species may 

 also absorb immature, yolked oocytes after spawn- 

 ing. 



The presence of a distinct, intermediate mode of 

 oocytes, indicating simultaneous maturation of a 

 new batch of oocytes while a group of advanced 

 oocytes is still in the ovary, is considered to be 

 strong evidence of multiple spawning (Clark 1929; 

 MacGregor 1976). In the oocyte size distributions 

 of mature northern anchovies an intermediate 

 mode of yolked oocytes never became distinctly 

 separate from the smaller, yolkless oocytes (Figure 

 5). There was some indication of continued growth 

 among intermediate-sized, yolked oocytes (0.56- 

 0.66 mm) in five stage 10 fish (both unovulated and 

 ovulated) before spawning. Yet oocytes in this size 

 range were absent in three spent fish (Figure 5). 

 The mode of intermediate-sized, yolked oocytes at 

 0.50 mm in these spent fish was in essentially the 

 same position as in stage 10 (ovulated) fish, indi- 

 cating little additional oocyte growth for some un- 



612 



