FISHERY BULLETIN: VOL. 84, NO. 2 



Table 1.— Group data of anchovy eggs and larvae: Sample mean daily production (y t ) at age f,, days for regression esti- 

 mators and sample mean daily death (n,) for maximum likelihood estimators (MLEs) of egg and larval mortality. 



1980 



1981 



Re gression estimators 

 ' t, y, 



1 0.41666 1.57 



MLE 



Regression estimators 

 ' ', Vi 



MLE 



Vi 



0.91666 1.21 



3 

 4 

 5 

 6 



7 

 8 

 9 



10 

 11 

 12 



1.41666 

 1.91666 

 2.41666 

 3.05 



5.6 



9.06 

 11.40 

 13.98 

 16.00 

 18.63 



1.07 

 0.64 

 0.88 

 0.74 



0.37 



0.21 



0.16 



0.10 



0.086 



0.012 



1 0.41666 1.57 



0.91666 1.21 0.36 



1.41666 1.07 0.14 



1 0.41666 1.33 



0.91666 2.00 



1 0.6666 1 .665 



1 .6666 



4 

 5 



6 

 7 

 8 

 9 

 10 

 11 



2.1666 

 3.05 



5.6 



9.06 

 11.40 

 13.98 

 16.00 

 18.63 



0.76 0.31 

 0.74 0.02 



0.31 



0.21 



0.16 



0.10 



0.086 



0.072 



0.37 



0.16 



0.05 



0.06 



0.014 



0.014 



3 

 4 

 5 

 6 



7 



8 



9 

 10 

 11 

 12 

 13 18.56 



1.41666 



1.91666 



2.41666 



2 2.91666 



3.10 



5.98 



9.45 

 11.97 

 14.37 

 16.53 



1.19 

 1.47 

 1.11 

 0.60 



1.23 

 0.44 

 0.25 

 0.16 

 0.10 

 0.10 



2.41666 1.11 



5 

 6 



7 

 8 

 9 



3.008 



5.98 



9.45 



11.97 



14.37 



16.53 



0.44 

 0.25 

 0.16 

 0.10 

 0.10 



0.08 10 18.56 



n, 



1.33 0.335 



0.22 



0.915 0.195 



0.475 



0.19 



0.09 



0.06 







0.08 0.02 



Vi-, 



2Used in MLE only. 



avoidance of the net becomes a serious bias. 



The sample mean daily production of eggs and lar- 

 vae per 0.05 m 2 (^) with its age (^) constituted the 

 data base for regression estimates of IMRs of eggs 

 and larvae in MEM and SEM. Mean daily produc- 

 tion represent eggs for 0.17 d (4 h) < t { < 3d, and 



larvae for 3 d < t { < 20 d; i = 1, . . .K, where K is 

 the total number of age groups. The sample age 

 structure (y^ tj) reflects that of a single cohort 

 under the assumption of steady production over the 

 survey period (Seber 1980). The same data set was 

 also used to generate the sample mean number of 

 eggs or larvae lost per day between two adjacent 

 age groups (n^ = y i _ 1 - y % ). The statistics n/s were 

 used directly in the MLE. Normally, sample totals 

 were used instead of sample means in MLE. I used 

 n/s because anchovy eggs and larvae were sampled 

 with different nets and because the number of egg 

 tows was different from that of larvae. 



MULTI-EQUATION MODEL (MEM) 



In the current estimation procedures, separate 

 mortality curves are constructed (Equation (3)) for 



398 



