FISHERY BULLETIN: VOL 78. NO. 4 



where N = number of cruises in a season 



Z), = number ofdays represented by cruise 

 i. [Sette and Ahlstrom (1948) de- 

 fined this to be the days included 

 in the cruise plus one-half the days 

 since the previous cruise and one- 

 half the days to the next cruise; in 

 this study only one cruise was 

 made during the spawning season 

 and the niunber ofdays represent- 

 ed by the cruise was taken to be 

 the entire spawning season. Dura- 

 tion of the spawning season for 

 E. mordax off the Oregon-Wash- 

 ington coast was estimated from 

 monthly or bimonthly collections 

 of larvae taken off Oregon in 1969, 

 1971, and 1972 (Richardson 1973, 

 footnote 5, unpubl. data; Richard- 

 son and Pearcy 1977). The earliest 

 that northern anchovy larvae 

 were taken in those years was 19 

 June and the latest that small lar- 

 vae (<10 mm) were collected was 

 10 August. Peak abundance, >100/ 

 10 m2 or >1, 000/1,000 m^ at a 

 station, occurred only between 21 

 July and 6 August. The spawning 

 season was estimated to last ap- 

 proximately from 15 June to 15 

 August, or 62 days.] 

 ti = the time in days from spawning to 

 hatching of the egg determined 

 from the equation given by Zweifel 

 and Lasker (1976) for incubation 

 times for northern anchovy: 



It = hexp [mil - exp( -ftT))] 



(6) 



where Iq - 1861 



m = -5.4572 

 13 = 0.0626 



T = the mean temperature at 3 m depth 

 at stations where northern an- 

 chovy eggs were taken, 15.18° C 

 in 1975 and 16.09° C in 1976. 



The Sette and Ahlstrom Egg Method assumes a 

 constant egg production throughout the spawning 

 season. 



The Simpson Egg Method of estimating the 

 number of eggs spawned in a season (Eg) was that 

 given by Simpson (1959) as modified by Houde 



(1977). The cruise census estimate, C^ or C^r, was 

 used to determine the number of eggs produced per 

 day during the spawning season 



E. 



or 



^kr 



(7) 



where E^ — daily egg production 



t, is defined under Equation (5). 



The seasonal estimate of egg production (Es) 

 was then determined by plotting the daily egg 

 production against the middate of the cruise rep- 

 resentative of the spawning season. The area 

 under the resulting polygon (triangle in this case), 

 determined by planimetry, was then equated with 

 egg production for the entire spawning season. 

 This method assumes a high egg production mid- 

 season tapering off to low production at the begin- 

 ning and end of the spawning season. It ap- 

 proaches a normal distribution of egg production. 

 For species with a short spawning season as in 

 northern anchovy off Oregon and Washington, es- 

 timates obtained by this method are about one- 

 half as large as those obtained by the Sette and 

 Ahlstrom Egg Method. 



The Saville Egg Method of estimating seasonal 

 egg production {Eg) is based on Saville's (1956, 

 1964) approach. It assumes that egg production 

 follows a normal distribution throughout the 

 spawning season. A census estimate iCf, or C^^) of 

 eggs obtained from a cruise made during the 

 spawning season represents a proportion of the 

 area under a normal curve. If the duration and 

 peak of the spawning season are known, seasonal 

 egg production (Eg) can be estimated (Houde 1977) 

 as: 



Es = 



C,j, 



^rt 



or 



Ckry. 



Xitj 



(8) 



where y = the number ofdays in cruise i 



X, = the proportion of the area under the 

 normal curve represented by 

 cruise i 

 t^ is defined under Equation (5). 



Duration of the spawning season is assumed to be 

 62 days, lasting from 15 June to 15 August (see Z), 

 under Equation (5)). The spawning peak is as- 

 sumed to be the middate, 15 July. 



The Smith Larva Method of estimating spawn- 

 ing biomass iB) is modified from the method used 



860 



