FISHERY BULLETIN: VOL. 85. NO. 2 



northern and central California, from Port San Luis 

 to Crescent City. I outline the reproductive patterns 

 among species and annual variation within species. 



MATERIALS AND METHODS 



Data for this study were collected between July 

 1977 and July 1982 from three sources: 1) a coast- 

 wide survey for adult rockfish made in July and 

 August 1977 (Gunderson and Lenarz 1980); 2) an 

 ongoing cooperative program, initiated in 1977 by 

 the California Department of Fish and Game and 

 the National Marine Fisheries Service, to sample the 

 commercial (Sen 1984) and sport rockfish landings 

 in northern and central California; and 3) a 1980 ex- 

 pansion of the cooperative program, to include data 

 on gonad condition, prey items, seasonal fluctuation 

 of interstitial fat, and gonad volumes. Additional col- 

 lections were taken during cooperative survey trips 

 with the National Marine Fisheries Service, South- 

 west Fisheries Center, and the California Depart- 

 ment of Fish and Game. Fish were collected to sup- 

 plement rarely sampled species and subadults that 

 were not well represented in the three surveys. 



Information on maturity was gathered primarily 

 for seven species of commercial and sportfishing im- 

 portance. Data for 27 additional species are pre- 

 sented, but these were inadequately sampled for 

 statistical treatment of the data (Table 1). For each 

 fish sampled from 1977 to 1982 the species, sex, and 

 total length (mm) was determined in the field and 

 otoliths were collected for age determination. The 

 viscera from each of the 16,444 fish sampled from 

 1980 to 1982 were removed, preserved in 10% buf- 

 fered formalin, and sent to the laboratory for anal- 

 ysis. In the laboratory the sex was verified and the 

 maturity stage of the gonads was determined based 

 on the criteria defined in this paper. Total length 

 was measured from the most anterior part of the 

 jaw to the dorsal tip of the caudal fin. When neces- 

 sary, total lengths were converted to fork length to 

 compare this study with others (Echeverria and 

 Lenarz 1984). 



Ages were determined for all fish investigated in 

 this study. Most estimates of age were made from 

 the surface of whole otoliths immersed in 70% ethyl 

 alcohol, under a dissecting microscope at 12 x . Stan- 

 dard techniques for counting annuli on whole 

 otoliths were followed (Kimura et al. 1979; Shaw and 

 Archibald 1981). Certain species such as 5. diplo- 

 proa required thin sectioning techniques for count- 

 ing annuli (Beamish 1979). In some species it is dif- 

 ficult to determine what constitutes an annulus 

 (Table 1), and consequently those ages have not been 



validated. The "break-and-burn" method of age 

 determination (Chilton and Beamish 1982) is useful 

 and more accurate than surface ages when aging 

 fish older than 16 years (Tagart 1984). Ages were 

 not redetermined in this study because the age at 

 maturity occurs before 16 years. Maximum ages, 

 however, may be underestimated. 



The gonad conditions described for males are im- 

 mature, maturing, mature (peak spermatogenesis), 

 spent, and resting. For females, they are immature, 

 maturing, mature (fertilized), ripe (eyed larvae), 

 recently spent, and resting (Lyubimova 1965; West- 

 rheim 1975; Gunderson et al. 1980). Histological sec- 

 tions were examined to define seasonal maturation 

 and the sections were analyzed based on the criteria 

 established by Moser (1967b) for S. paucispinis and 

 by Lisovenko (1970) on 5. alutus for mature fish. 

 The development of germ cells into spermatozoa or 

 mature oocytes was examined to determine the 

 developmental sequence leading to maturity. Cel- 

 lular development was compared with the external 

 morphology (Westrheim 1975; Gunderson et al. 

 1980) to understand clearly the developmental se- 

 quence and to aid in the interpretation of maturity 

 stages in the field. Gonads were subsampled from 

 fish in all maturity stages. Whole gonads were sec- 

 tioned and stained with hematoxylin. Histological 

 sections were examined from 519 testes and 708 

 ovaries from 30 species (Table 1). Egg diameters 

 were routinely measured from histological sections 

 and checked against whole eggs using an ocular 

 micrometer. 



To determine age and size at maturity and repro- 

 ductive seasonality, the approximate age and size 

 when 50% and 100% of the males and females were 

 mature were estimated for each species. For those 

 commercial species for which large samples were 

 available, I checked the accuracy of the rough ma- 

 turity estimates by applying the method of Gunder- 

 son et al. (1980). These authors used the logistic 

 model 



1 + e"*-^* 



where P^ = proportion mature at size x, and a and 

 b are constants, to estimate the length at 50% 

 maturity for a sample. I applied a transformation 

 of this equation 



In 



1 



I = ax + b 



to obtain estimates of a and b for size and age by 



230 



