FISHERY BULLETIN: VOL. 83, NO. 2 



i. i 



where xi is the adjusted mean egg count for re- 

 gion i and W, is the relative area weight for region 

 i. 



^ Sum is not equal to the total due to rounding error. 



Zero catch was assumed for regions where no 

 samples were taken because historical records 

 show those regions usually had low densities of 

 eggs and larvae. The weighted Xu's were also 

 corrected for extrusion through the mesh by mul- 

 tiplying the catch by the ratio of the catch in a 

 0.150 mm CalVET net to the catch in the net used 

 in a particular survey (r): r = 3.6 for 0.55 mm 

 mesh silk 1 m ring net 1 1951-68), r = 3.04 for 0.505 

 mm mesh Nitex' 1 m ring net (1969-76), r = 12.76 

 for 0.505 mm mesh Nitex bongo net (1978-present) 

 (Lo 1983). The 0.505 mm mesh bongo net seems to 

 catch 4 times that of a 1 m ring net. The reason is 

 unknown. (A field experiment was conducted in 

 April 1983 to reestimate the extrusion rate of 

 anchovy eggs from 0.505 mm mesh bongo net. The 

 data have not been analyzed at the time of writing. 

 Although the egg samples from bongo nets were 

 u.sed to compute the HEP, the bongo net is pri- 

 marily used for catching anchovy larvae, whereas 

 the CalVET net is the egg sampler The discrep- 

 ancy between bongo and 1 m ring net is not of 

 major concern for the current anchovy biomass 

 estimation. ) The standing stock of eggs per 0.05 

 m^ is then 



and 



mt, = Xu- r 



var (mil ' = var ixw )r + xfr var(r) 



where mt is the standing stock ofeggs( and larvae) 



Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



140 



up to age t days from fertilization. Here ti is the 

 duration of incubation. 



The size of standing stock of eggs depends on not 

 only egg production rate and mortality rate but 

 also the duration of incubation (or the incubation 

 time), which is a function of sea temperature. The 

 average temperature for all positive egg tows 

 (tows which contain one or more anchovy eggs) 

 over January-April in each year was used to 

 estimate incubation time {ti) using the equation 

 (Lo 1983) 



ti - (18.73 e"°^25^^"^P) 



where ti = incubation time in days, 



temp = temperature in degrees centigrade. 



Both the standing stock of eggs imii') and the 

 incubation time (ti) are essential in computing 

 the time series of daily egg production. The 

 temperature in January- April ranges from 11° to 

 19° C. The long-term average temperature from 

 January to April is 14.25° C, thus the average 

 incubation time is 3.15 d. 



Larval Data 



The anchovy larvae from all years were mea- 

 sured to the nearest 0.5 mm preserved length. For 

 the purpose of estimating mortality rate, larval 

 data were grouped into 2.5 mm, ranging 2-3.0 mm; 

 3.75 mm, 3.5-4.0 mm; 4.75 mm, 4.5-5.0 mm: ... for 

 larvae < 30 mm. Each preserved length was first 

 converted to a live standard length using a shrink- 

 age formula based on the tow duration (Theilacker 

 1980), and then converted to age (,t days) using a 

 two-cycle Gompertz growth curve. The first cycle 

 is from hatching to yolk-sac absorption, a temper- 

 ature-dependent growth curve, and the second 

 cycle is from yolk-sac absorption to 22 mm larvae, 

 a food-dependent growth curve (Zweifel and Hunt- 

 er footnote 2; Methot and Hewitt footnote 3; Lo 

 1983). Larval abundance by length (age) group 

 was estimated using a negative binomial weighted 

 model (Bissell 1972; Zweifel and Smith 1981) 

 which incorporates the "effective sampler size" 

 (relative sampler bias). All larval abundance data 

 were adjusted to conform to the following standard 

 conditions: no extrusion, no day-night difference 

 in avoidance, and a constant water volume filtered 

 per unit depth. These data were converted to daily 

 production (Pt) by dividing the total number of 

 larvae in each length group by the duration (the 

 number of days larvae remain within each length 



