LAUTH: CABEZON IN PUGET SOUND. WA 



spawned females (Fig. 3VI). Three females with 

 Stage VI ovaries were collected in the immediate 

 vicinity of freshly deposited embryo masses, pre- 

 sumably after spawning them. Within these ovaries, 

 an incipient mode, which ranged in size from 0.70 

 to 0.95 mm and had a mean size of 0.79 mm, was 

 apparent. A relatively small number of large diam- 

 eter eggs (~1.5 mm) were scattered within all Stage 

 VI, Stage VII, and Stage VIII ovaries (Fig. 3VI- 

 VIII) and were presumably remnants of the recent 

 spawning event. In Stage V females the largest 

 mode would mask evidence of residual eggs so it was 

 not possible to ascertain whether they had already 

 spawned in the 1984-85 spawning season. 



As the ovaries progressed to Stage VII, the eggs 

 of the incipient mode were larger and distinct from 

 the intermediate mode. These yolked eggs appeared 

 to be hydrating in preparation for another spawn- 

 ing. Females exhibiting this condition had spawned 

 previously and since eggs of an intermediate size 

 were still present, these females were capable of 

 spawning again. The incipient mode for Stage VII 

 ovaries ranged from 0.95 to 1.50 mm with a mean 

 of 1.16 mm. The intermediate mode of Stage VII 

 ovaries ranged from 0.35 to 0.90 mm and averaged 

 0.51 mm. 



Stage VIII ovaries were similar to Stage I ovaries. 

 There was a single and small mode of eggs which 

 were deteriorating noticeably. Irregularly shaped 

 ova were translucent or transparent and devoid of 

 yolk. Unlike Stage I, Stage VIII ovaries had rem- 

 nant eggs (~1.5 mm in diameter) from at least one 

 previous spawning event (Fig. 3VIII). 



When the eight stages were plotted against the 

 date when females were captured, a general pro- 

 gression of ovarian development was seen (Fig. 4). 

 Stage III to V ovaries were only seen in females 

 caught between December and May. Stages VI and 

 VII were found from February through August. 

 Stage VIII females were caught both before and 

 after Stage III through VII females. The early Stage 

 VIII's were probably carry-overs from the previous 

 spawning season. Females in Stages I and II were 

 found prior to all other stages. 



The WGSI values for the eight stages of ovarian 

 development were in agreement with what might 

 be expected in a multiple spawner (Fig. 5). For 

 ovaries in the resting condition (Stage I), the WGSI 

 was at its lowest point. The WGSI gradually in- 

 creased to a maximum in Stage V when eggs were 

 hydrated and females were in spawning condition. 

 After the eggs were released there was an obvious 

 drop in the weight of the ovaries relative to the body 

 weight. The WGSI slightly increased in Stage VII 



and then fell in Stage VIII. Without the aid of 

 histological techniques, it was not possible to dis- 

 tinguish an intermediate stage between VII and 

 VIII; this stage would have been virtually identical 

 to Stage V, and had there been such a stage, it is 

 conceivable the WGSI would have reached another 

 maximum before finally declining in Stage VIII. 



Relation Between Batch Fecundity and 

 Weight and Total Length 



From ova diameter frequency plots, two basic 

 types of female spawners were evident: 1) those 

 which were going to spawn for the first time (un- 

 spawned; Stages III to V), and 2) those which had 

 already spawned at least once and had potential for 

 spawning another batch (spawned; Stages VI and 

 VII). The number of eggs to be spawned during 

 each spawning event (batch fecundity) was deter- 

 mined by estimating the number of eggs in the 

 largest mode for females possessing ovaries in 

 stages III to VII. 



Data for spawned and unspawned females was 

 pooled for regression analysis because the range of 

 values for comparable fish weights and lengths was 

 similar, and because separate regressions for spawn- 

 ed and unspawned females were not statistically dif- 

 ferent (P > 0.05). Furthermore, pooling spawned 

 and unspawned data provided analysis over a 

 broader size range of fish and considerably increased 

 the sample size. The resulting regressions of batch 

 fecundity on length and weight were significant at 

 P < 0.001, and the correlation coefficients were 0.69 

 and 0.73, respectively (Fig. 6). The regression of 

 batch fecundity on length predicted that females 

 from 500 mm to 775 mm would release between 

 57,000 and 137,000 eggs during a spawning event, 

 and the regression of batch fecundity on weight 

 predicted that females from 2.5 kg to 10.5 kg would 

 release between 66,000 and 152,000 eggs during 

 each spawning event. 



DISCUSSION 



Along the western U.S. coast, the length of the 

 spawning season for marine cottids varies from 1 

 month to year-round (Atkinson 1939; Jones 1962; 

 Marliave 1975; Tasto 1975; DeMartini 1978; DeMar- 

 tini and Patten 1979; Goldberg 1980; Garrison and 

 Miller 1982). Based on temporal embryo mass 

 abundance and ovarian condition, cabezon spawn- 

 ing in Puget Sound commences in late November 

 and lasts 10 months through early September of the 

 following year while peak spawning occurs from 



149 



