concerns the small, partially-yolked oocytes 

 (<0.65 mm) that potentially could be recruited to 

 the advanced mode but were excluded from fe- 

 cundity estimates. Using the average lower limit 

 of advanced oocyte diameters (from only those 

 seven ovaries with clearly separated, advanced 

 modes) as the minimum size of oocytes destined 

 for release, it was assumed that further contri- 

 bution of smaller oocytes would be unlikely and 

 minimal in the current season. 



A compensatory increase in fecundity follow- 

 ing a reduction in population density was 

 suggested by Bagenal (1973). We did not detect 

 significant changes in length-specific fecundity 

 for Dover sole despite tripled landings and the 

 suggested decrease in age-specific length since 

 1977 (Demory et al., fn. 2). However, a decline 

 in reproductive output is unlikely because our 

 data indicate a compensatory shift in size at 

 first maturity. Dover sole now matui'e at signif- 

 icantly smaller sizes than reported by Harry 

 (1959). We found all fish >320 mm to be mature 

 (the smallest mature fish was 240 mm). Al- 

 though only 15% (or 245 specimens) of those fish 

 in Harry's study were smaller than 380 mm, 

 they were 330 mm at first maturity, 380 mm 

 when 50% mature, and 420 mm when 100% ma- 

 ture (Fig. 3). Similar sizes at maturity were re- 

 ported by Hagerman (1952) for Dover sole col- 

 lected in northern California in 1949. The 45 

 females (5% of total sample) <330 mm were all 

 immature. 



The evident change in size at maturity during 

 the interim 35 years may reflect differences in 

 the criteria used for assessing stage of maturity, 

 and in the time of year and size of the fish used 

 for maturity assessments. Harry collected sam- 

 ples between May and October, which excludes 

 the peak spawning months of December and 

 January. Postspawning, inactive, mature 

 ovaries are difficult to differentiate from imma- 

 ture ones during the early summer months using 

 anatomical or histological criteria (Hunter, fn. 

 4). Classifying postspawners as immature would 

 lead to a greater size at 50% maturity than that 

 estimated from samples collected within the 

 peak spawning period. Size at first maturity, 

 however, should be relatively independent of 

 season, since it seems unlikely that stage of ma- 

 turity would be incorrectly identified in all 

 smaller fish. Also, as egg development becomes 

 more evident with the approach of spawning 

 season, the mature ovaries become easier to 

 identify (at least in Harry's August-October 

 samples). Despite these potential problems with 



data comparisons, a decrease in length at ma- 

 turity is clear and supported by evidence that 

 Dover sole >370 mm collected in the Columbia 

 area during the 1980-81 spawning season were 

 all mature (Demory et al., fn. 2). 



We cannot assess possible changes in the age- 

 maturity relationship, because comparable age- 

 ing methodology was not used by Harry (1959). 

 It cannot be inferred, however, that Dover sole 

 are maturing at younger ages concurrently with 

 smaller sizes, because length at age has also de- 

 cHned. 



Although the decrease in length at maturity 

 implies that individual reproductive potential is 

 improved, lack of historical information on ac- 

 curate age-specific maturity and fecundity 

 makes it impossible to detect any net change in 

 reproductive output per individual. To identify 

 changes in the reproductive output of the Dover 

 sole off Oregon, a reliable assessment of spawn- 

 ing biomass, size and age structure of the popu- 

 lation, and both length and age at maturity is 

 required. 



Increased landings and effort may have af- 

 fected the duration and timing of the spawning 

 season, which could influence year class 

 strength. Size-selective exploitation removes 

 relatively more of the larger and older Dover 

 sole (Best 1961). Larger individuals are not only 

 more fecund, but appear to have a longer and/or 

 later spawning period than small, young fish. 

 The reproductive status and population dy- 

 namics of Dover sole, as influenced by factors 

 discussed here, can be further understood with 

 continued, appropriate, long-term research. 



Acknowledgments 



This work was supported by the National Mar- 

 ine Fisheries Service, Northwest and Alaska 

 Fisheries Center, Seattle, WA through the Co- 

 operative Institute for Marine Resources 

 Studies at the Hatfield Marine Science Center, 

 Newport, OR (contract #85-ABH-00025), and 

 the Oregon State University Sea Grant College 

 Program (Grant No. NA85AA-D-SG095, Project 

 R/ES-7). We thank .J. Rackowski for technical 

 assistance, D. Erickson and R. Demory for field 

 collections, R. Demory and W. Barss for consul- 

 tation on ageing criteria, and G. Taghon and R. 

 Greene for use of and instruction on the image 

 analyzer. We appreciate the helpful comments of 

 G. Boehlert, R. Demory, J. Emlen, J. Harvey, 

 J. Hunter, D. Markle, and W. Pearcy on an 

 earlier draft of this manuscript. 



994 



