Hunter et al Fecundity, spawning, and maturity of Microstomus paaficus 



109 



oocytes or postovulatory follicles present; (2) early 

 yolked oocytes with /3 atresia; (3) early yolked oocytes 

 with only a atresia of the early yolked oocytes or no 

 a or ^ atresia; (4) unyolked oocytes with /5; (5) unyolked 

 oocytes with only a of the unyolked oocjrtes; and (6) 

 unyolked oocytes with no atresia. The sexual maturity 

 for females identified by criterion 1 is certain, but some 

 females may be excluded if only criterion 1 is used. 

 Criteria 2-5, if added to criterion 1, broaden the 

 maturity definition but increase the risk of misclassi- 

 fication. Criterion 6 is considered by definition to be 

 immature. We evaluated these criteria to determine the 

 optimal histological definition of maturity using a 

 regression analysis of the lengths of females identified 

 by each criterion. 



Length, weight, and 

 gonad weight relationships 



To enable the reader to convert from one measurement 

 to another, equations are provided to estimate fresh 

 wet weight from frozen wet weight and from length 

 for Dover sole taken in Oregon and central California 

 waters (Table 3). Analysis of covariance indicated that 

 the slope of the regression of the natural logarithms 

 of weight on length did not differ between sexes for 

 either state. The adjusted group mean for males 

 differed from that for females in Oregon (A^ 1421, 

 Fi, 1418 64.87, P< 0.005 for length range 225-522 mm) 

 but not in California. The slope of the regression of the 

 natural logarithms of weight on length did not differ 

 between central California and Oregon females but the 

 adjusted group means were different (N 2215, F-y 2212 

 79.18, P<0.005 for length range 188-547mm).' No 

 difference existed between states in the equations for 

 males. We do not attach too much biological importance 

 to these differences; they could be related to differ- 

 ences in the timing of annual reproductive cycle or our 

 sampling of it. Nonetheless, it seemed preferable to use 

 the relationship for a specific sex or region, so all are 

 listed. 



An exponential model was fit to these data sets using 

 a statistical program of weighted nonlinear regression 

 (Dixon et al. 1988) where the weighting factor was 

 the inverse of the variance of fish weight because the 

 variance of fish weight increased with fish length. To 

 compute the variance, fish lengths were divided into 

 several segments, chosen so that within each segment 

 the variance of fish weight was homogeneous. We pre- 

 ferred to obtain the estimates of coefficients directly 

 from the nonlinear fitting so that fish weight could 

 be directly estimated from the exponential model 

 (Table 3). 



Freezing of Dover sole caused a 9.47mm shrinkage 

 in total length, independent of fish length (Table 3). A 



sample of 251 Dover sole was measured just after cap- 

 ture and again, after thawing, four months later. The 

 slope of the regression of fresh total length on frozen 

 total length (after thawing) was not statistically differ- 

 ent from 1, but the intercept, 9.47mm, was significant. 

 Freezing of females, with ovary removed, resulted in 



about a 22% loss in wet weight (0.22 = 1 - ; 



1 29 

 see Table 3). ^"^^ 



We also provided equations to estimate ovary wet 

 weight (g) from female wet weight (g, without ovary). 

 This conversion is important if one wishes to express 

 fecundity as a function of the total weight of the female, 

 because all fecundity relations in this study are ex- 

 pressed as a function of female weight without an 

 ovary. As ovary weight is a function of the developmen- 

 tal state of the ovary as well as the weight of the 

 female, separate equations are provided for the pre- 

 spawning period (November- December) when ovaries 

 are less developed and for the spawning season when 

 they are more fully developed. We also provided multi- 

 ple regression equations to estimate ovary weight from 

 female weight and the spherical volume of the average 

 advanced yolked oocyte (computed from the mean 

 diameter). These equations are used in the discussion 

 to estimate ovary weight when an ovary contains an 

 entire complement of fully matured advanced yolked 

 oocytes. 



Reproductive condition 



Accuracy of gross anatomical classification 



We rarely misclassified inactive ovaries using gross 

 anatomical criteria. Of the 1272 females classified as 

 inactive, only 14 (1.1%) were identified as active using 

 histological criteria. This error rate is so low that dif- 

 ferences could be attributable to clerical errors alone. 

 A more common error in gross anatomical classifica- 

 tion was to misclassify females as having active ovaries. 

 One hundred and fifty-nine females (11.9%) were 

 visually classified as having advanced yolked oocytes 

 and were believed to be capable of spawning, while 

 histological analysis indicated that their ovaries were 

 inactive and future spawning was unlikely. The 159 

 females misclassified as active fell predominantly into 

 two classes: females with ovaries in the early stages 

 of vitellogenesis (40.8%), and females with advanced 

 yolked oocytes with high levels of atresia (30.1%) 

 (Table 4). 



Misclassification of the early stages of vitellogenesis 

 as active is expected because the gross anatomical 

 criterion, "yolked oocytes visible," is not exact; 

 observers are bound to differ on whether to include or 

 exclude females that fall near the visible threshold 



