1,369 ±148 ooc5^es/g total body weight in =4) 

 (Pike 1951), differs from our estimate of 720±40 

 oocytes/g total body weight in = 21). Our estimate 

 is more accurate than Pike's because it is based on 

 counts of only the most advanced oocytes in ripe or 

 nearly ripe fish. Pike's fecundity estimate was 

 based on the assumption that northern anchovies 

 spawn three equal batches of oocytes per year and 

 was calculated by dividing the total number of 

 oocytes >0.20 mm by three. His assumption of 

 spawning frequency, based on the number of 

 modal peaks in oocyte size-frequency distribu- 

 tions, is still unproven and could lead to erroneous 

 fecundity estimates depending on the actual 

 number of spawnings per fish. 



Northern anchovies in the northern subpopula- 

 tion off Oregon and Washington apparently have a 

 greater fecundity (based on ovary-free body 

 weight), 826 in = 21), than those in the central 

 subpopulation off California, based on estimates 

 by MacGregor (1968) and Hunter and Goldberg 

 (1980), 606 (n = 19) and 389 (« = 23), respectively. 

 An analysis of variance (single classification) in- 

 dicated the presence of a highly significant 

 (P<0.01) added variance component between in- 

 dividual relative fecundity estimates (based on 

 ovary-free body weight) of our fish and those 

 examined by MacGregor (1968). Hunter and 

 Goldberg's mean value, which is more directly 

 comparable than MacGregor 's with our estimate 

 because it was similarly based on ripe fish with 

 hydrated oocytes, was even lower than Mac- 

 Gregor 's, although statistical comparisons were 

 not possible (individual fecundity values were not 

 listed). The difference between the two California 

 estimates may have been due primarily to differ- 

 ences in the stage of ovarian maturation of the fish 

 used for fecundity determinations. Difficulty in 

 distinguishing the most mature oocytes from less 

 mature ones before hydration could have caused 

 the higher estimate obtained by MacGregor who 

 used only fish with unhydrated oocytes. 



The higher fecundity of northern anchovies off 

 Oregon and Washington may represent a true ra- 

 cial difference between fish in the northern and 

 central subpopulations. Racial differences in 

 fecundity have been demonstrated in many 

 species of fish with probable causes being either 

 environmental or genetic factors (Bagenal 1957, 

 1967). Bagenal (1966) speculated that geographic 

 differences in plaice fecundity were caused by dif- 

 ferences in food availability and population den- 

 sity 



FISHERY BULLETIN: VOL. 78, NO. 3 



Spawning Frequency 



Fish, such as the northern anchovy, with asyn- 

 chronous oocyte development have the potential to 

 spawn more than once during the season (de Vla- 

 ming 1974). Yet the actual number of times a 

 female northern anchovy spawns during a year 

 has not been conclusively documented in any of 

 the subpopulations. Pike (1951) estimated that 

 northern anchovies in the northern subpopulation 

 off British Columbia spawn three times during the 

 3-mo spawning season. MacGregor (1968) 

 suggested that at least some fish in the central 

 subpopulation spawn more than once during the 

 spawning season, which may include all 12 mo of 

 the year. Hunter and Goldberg (1980) estimated 

 the spawning frequency of northern anchovies in 

 the central subpopulation to be once every 6-7 d 

 during months of peak spawning. 



Pike's (1951) conclusion was based solely on the 

 presence of multiple modes in oocyte size distribu- 

 tions and ambiguous data on changes in the ratio 

 of immature to advanced oocytes during the 

 spawning season. These data alone cannot be used 

 to determine spawning frequency in fishes. Mac- 

 Gregor (1968) concluded that spawning later in 

 the year represented repeat spawning by some 

 northern anchovies because early in the spawning 

 season all mature females had well-developed 

 eggs or were recently spent. Christiansen and 

 Cousseau (1971), using histological techniques, 

 found that some female M. merluccius had the 

 physiological ability to recover more rapidly after 

 spawning than the rest of the population and that 

 these fish spawned a second time later in the sea- 

 son off Argentina. It seems likely, therefore, that 

 anchovies in the central subpopulation spawn 

 more than once during the protracted spavvming 

 season. But the number of spawnings per year may 

 be variable because environmental conditions 

 such as temperature and food supply, which are 

 known to influence reproductive cycling in fishes, 

 can vary from year to year (Bagenal 1966, 1969; de 

 Vlaming 1971, 1974; Hodder 1972', Tyler and Dunn 

 1976). Recently, Brewer (1978) suggested that food 

 availability may limit both the number of eggs 

 spawned and the number of spawnings per year by 

 northern anchovies in San Pedro Bay, Calif. 



Hunter and Goldberg's (1980) estimate of 

 spawning frequency was based on the mean per- 

 cent incidence of northern anchovies with 1-d-old 

 ovarian follicles in trawl samples taken during a 

 2-wk period in February. Their determination of 



614 



