110 



Fishery Bulletin 90(1). 1992 



Figure 3 (below) 



Seasonal change in four indices of reproduction in Dover sole Microstomus pacificus. Indices 

 are plotted as a function of elapsed time since 1 November; data are combined from different 

 years; California (solid circles) and Oregon (open circles) data were combined to fit trend lines; 

 numbers are the sample size of females, (upper left) Percentage of Dover sole with active 

 ovaries; trend line is a weighted Weibull model (see text Eq. 1). (upper right) Percentage of 

 females with active ovaries which had one or more spawning states; trend line is logistic model 



ga + bL 



P = , where a = - 5.678 and b = 0.036. (lower left) Mean number of spawning states 



l + e"*'''' 



in active ovaries; bars are two standard errors of the mean, (lower right) Percentage of females 

 with inactive ovaries identified as postspawning; trend line is logistic model where a = 8.495 

 and b = 0.042. 



for detection of yolked oocytes in 

 the ovary. An exact criterion, 

 such as oocyte diameter, would 

 be more accurate but would be 

 impractical for production work 

 on the ship. Misclassification of 

 highly atretic ovaries as active is 

 also expected, since a-atretic ad- 

 vanced yolked oocytes are diffi- 

 cult if not impossible to see with 

 the unaided eye. As highly-atre- 

 tic advanced ovaries were rare in 

 this study, our failure to detect 

 them was a minor systematic 

 error. Under environmental con- 

 ditions unfavorable to reproduc- 

 tion, however, this could be an 

 error of consequence. 



Changes in ovarian 

 condition during tine 

 spawning season 



The fraction of females anatom- 

 ically classed with active ovaries 



70 



iJJ 60 



< 



ff 50 



UJ 



> 

 I- 



o 



50 JOO 150 



DAYS ELAPSED SINCE NOVEMBER 1 



FEB 

 MONTH 



50 100 150 



DAYS ELAPSED SINCE NOVEMBER 1 



JAN FEB 



MONTH 



