FISHERY BULLETIN: VOL. 83, NO. 2 



TABLE 5. — Percentage of northern anchovy females taken from 1977 to 1982' that were 

 classed in three atretic states that occurred in each of five reproductive classes. 



Reproductive state 



Atretic 

 state 



Percent yoltced 



oocytes 



with alpha stage 



atresia 



Postovulatory 

 follicles 



Hydrated 

 oocytes 



(%) 



Oday 



(%) 



1 day 



(%) 



Spawning No evidence 

 recent or of recent Total 



imminent^ spawning^ number 



(%) (%) of females 



' Calculated from data given in Table 4; only females with yolked oocytes are considered. 

 ^Females with either hydrated oocytes or postovulatory follicles ages or 1 d (the sum of the first three 

 columns). 

 ^Females with yolked oocytes but without hydrated oocytes or postovulatory follicles. 



This nonlinearity becomes obvious when the 

 end of the spawning season is extrapolated from 

 numbers of females classed in atretic state 2. For 

 example, of the 1,620 mature females taken dur- 

 ing the peak of spawning (28 January-18 March 

 1982) in southern California (Table 4), only two 

 were in atretic state 2 and 1,612 had yet to pass 

 through state 2. Since laboratory data indicate 

 that about 9 d are required to pass through atretic 

 state 2, it would require (1,612/2) x 9, or over 

 7,000 d for the entire population to become atretic 

 at the rates of atresia observed in February, which, 

 of course, is nonsense. Projections of the end of the 

 spawning season using higher rates of atresia 

 taken in April in southern California (24-27 April 

 1980) give a more realistic projection ((87/ 

 14) X 9 = 56 d). Such an arithmetic projection 

 may be inappropriate for collections which have a 

 very high rate of atresia such as those taken in 

 Monterey in March 1979 ((84/41) x 9 = 18 d), and 

 a geometric model might be preferable. The point 

 we wish to emphasize is that atretic rates are 

 nonlinear over the season with the rate increasing 

 markedly as the season closes. Thus only samples 

 taken near the close of the spawning season are of 

 value for forecasting the end of spawning for the 

 population. 



Seasonal Changes in Atresia 



Among Females of Different Lengths 



To evaluate how atretic rates change among 

 females of different lengths, we segregated our 

 data into two length classes (females ^ 10 cm SL 

 and those >10 cm SL) and calculated the percent- 

 age of mature females that had atretic ovaries 

 (atretic states 1-3 combined). Mature is defined 

 here as all females except those which have yet to 

 reach first maturity (small females with small 

 immature non-atretic ovaries). We also calculated 

 the fraction of females in each length class with 



1-d-old postovulatory follicles, a measure of the 

 percentage of females spawning daily (Hunter and 

 Goldberg 1980). 



In every case, regardless of cruise or season, 

 small females (^10 cm SL) consistently had a 

 higher rate of ovarian atresia than did larger ones 

 (>10 cm SL) (Table 6). This is a strong trend as the 

 probability of such an event (9 pairs of the same 

 sign) is (1/2)^. In addition, the difference between 

 pairs was statistically significant (chi-square test) 

 even when the levels of atresia were quite low. For 

 example, in February-March 1981, only 4.1% of the 

 small females and 1.9% of the large females were 

 atretic, yet this difference was significant at P < 

 0.05 using the chi-square test. As would be ex- 

 pected, the percentage of females with atretic 

 ovaries increased in both length classes as the 

 season progressed from January through June. 



The consistency of the differences in the inci- 

 dence of atresia between large and small females 

 indicates that the smaller ones must have a much 

 shorter spawning season than larger ones. 

 Females <10 cm long are typically about 1-yr-old 

 and are in their first spawning season whereas 

 those longer than 10 cm are predominantly 2-3 yr 

 old and have spawned during the previous sea- 

 sons. These data indicate that the first spawning 

 season of females may be quite short with signifi- 

 cant numbers of females leaving the spawning 

 population in early April, while the older fish con- 

 tinue to spawn. That the rates of atresia in young 

 fish were always higher even in the peak months of 

 spawning such as February and March indicates 

 that a small percentage of small females may only 

 spawn a few times during the season in contrast to 

 the older females which appear to be spawning at 

 about weekly intervals for months. The fraction of 

 small females spawning per day would be expected 

 to be less than larger females since the small 

 females have a higher incidence of ovarian atresia. 

 We calculated the fraction of females spawning per 



132 



