120 



Fishery Bulletin 90(1). 1992 



State 



of oocytes that were atretic was sig- 

 nificant and negative. According to the 

 equation, when 10% of the advanced 

 oocytes were atretic, total fecundity in 

 a 1 kg female was about 8% lower than 

 when no advanced yolked oocytes 

 were atretic. This analysis indicated 

 that atretic losses of potential annual 

 fecundity occurred, but on a popula- 

 tion basis such losses were negligible. 

 No relation between fecundity and 

 atresia existed for Oregon females, 

 probably because atresia was less 

 prevalent in Oregon, with only 6% of 

 females effected compared with 26% 

 in central California. 



The ovaries of many more females 

 were examined histologically for 

 atresia (N 2607) than were examined 

 using the anatomical method, because 

 we restricted the anatomical work to 

 the fish used to measure total fecun- 

 dity. Only 2% of all females examined 

 histologically (Table 2) had ovaries in 

 which 50% or more of the advanced 

 yolked oocytes were in a atresia, but 

 minor atresia was more common. 

 Minor atresia occurred in 52% of the 

 nonspawning California females and in 

 35% of the nonspawning Oregon 

 females (Table 2). 



The histological method was more 

 sensitive than the anatomical one. 

 Alpha atresia of advanced yolked 

 oocytes was detected at least twice as 

 frequently using histological tech- 

 niques. The histological method was 

 more sensitive because we could detect 

 more subtle changes in oocyte struc- 

 ture and because we scanned about 150 oocj^es per 

 ovary, compared with 30 oocytes in the anatomical 

 method. Despite the lack of sensitivity, the anatomical 

 method was valuable because the standing stock of 

 atretic oocytes could be easily estimated and directly 

 related to total fecundity. 



The histological evidence indicated that females with 

 a-atretic advanced yolked oocytes were more common 

 in central California waters than off Oregon. However, 

 season and locality were confounded because most 

 females from Oregon were taken prior to the spawn- 

 ing season while most females from California were 

 taken during the season. To determine if either season 

 or locality affected the relative frequency of atretic 

 females, we combined the minor-atresia and major- 

 atresia classes for California and Oregon and fit the 



Table 12 



Histological determination of number of Dover sole Microstomus pacificus, with 

 a atresia of advanced yolked oocytes expressed as a percentage of all females with 

 advanced yolked oocytes taken in central California and in Oregon, beginning 

 (November-December) and during (January-May) the spawning season. 



Spawning season 



Beginning 



During 



Beginning + During 



% 



N 



N 



% 



95% CI 



N 



Central California 59.2 71 51.9 617 52.5 47.9-57.2 688 

 Oregon 36.8 536 49.4 83 38.4 33.8-43.3 619 



stepwise logistic model 



l + eCo+PiXi+zJoXz 



(Eq. 11) 



to the data (Table 12), where P is the fraction of females 

 with atretic oocytes. The independent variables for 

 location (Xi) are - 1 for California and 1 for Oregon, 

 and for season (X2) are - 1 for prespawning and 1 for 

 during spawning. The estimates of coefficients for the 

 equation are /Jo = -0.183 (SE 0.056) and /3, = -0.288 

 (SE 0.056) {P2 is not given because effect of season 

 was not significant in an early regression analysis); the 

 estimate of the atresia rate P for California is 0.525 

 (95% CI 0.479-0.572; Carter et al. 1986) and for 

 Oregon is 0.384 (95% CI 0.338-0.433). 



