LO: EGG PRODUCTION OF NORTHERN ANCHOVY 



bongo nets because only bongo or similar nets (1 m 

 ring nets) were used for sampling eggs and larvae 

 prior to 1978. 



Two series of HEP estimates were constructed. 

 Series 1 assumed a constant IMR for the egg stage 

 with tc = ti , whereas series 2 with tc = tys 

 assumed a constant IMR throughout the embry- 

 onic period (Table 3). Both Ptj and Pty, , the daily 

 larval production at hatching and yolk-sac stage, 

 were obtained from the fitted line of Equation (8B) 

 with tc = ti . 



Under series 1, nearly half of the egg IMR (a) 



were negative (11 out of 24 yr). This was because 

 the egg IMR depended on the value of q through 

 Equation (10) where q = mtj/Ptj. However, judg- 

 ing from Equation (10), q = ti for a = 0. Therefore 

 for those years where q < ti , egg IMR would be 

 less than 0. The small ^'s could result from the 

 underestimated mt/ or overestimated Ptj or both. 

 The poor results of IMR (a) were likely due to the 

 underestimation of m</ . As a result, the standing 

 stock of eggs and that of yolk-sac larvae were 

 combined into one group in series 2, to eliminate 

 the negative IMR's. 



Table 3.— Two time series of estimated historical egg production {Pq), and egg mortality (a), larval 

 mortality coefficient (/3), mean egg abundance (mti), mean egg and yolk-sac larval abundance 

 (mtys '- January- April, and mean larval abundance (La ) per 0.05 m^ 1951-82 with standard error 

 in parentheses. 



' Series 1 and 2 are two methods used for estimating daily egg production (Pq) 

 stage whereas series 2 assumed a constant IMR for egg through yolk-sac larval 

 ^Computed from annual larval abundance for the central subpopulation (Table 



Sehes 1 assumed a constant IMR tor egg 



stage. 



2, Stauffer and Charter 1982). 



145 



