FISHERY BULLETIN: VOL. 79, NO. 4 



„ 150- 



>- 



< 



CO 



z 

 o 



100 — 



5 5 0- 



I- 

 < 



—\ — ^ — I — I — \ — ' \ ' I ^ \ ' r~ 



200 400 600 800 1000 1200 1400 



AVERAGE EFFORT (DAYS AT SEA) 



1600 



Figure 5. — Linear and exponential regressions of annual 

 1969-79 catch per unit effort versus observed effort averaged 

 over a 2-yr period for three species of thread herring captured 

 on the Pacific coast of Costa Rica. 



200 400 600 800 1000 1200 1400 1600 1800 

 AVERAGE EFFORT (STANDARD DAYS AT SEA) 



Figure 6. — Linear and exponential regressions of annual 

 1969-79 catch per unit effort versus standardized effort averaged 

 over a 2-yr period for three species of thread herring captured on 

 the Pacific coast of Costa Rica. Effort was standardized to 

 account for improved capture efficiency and reduced resource 

 availability after 1973 (see text). 



effects on effort at Ys, the exponential model pro- 

 duced higher fs estimates, but only significantly 

 with observed effort data. 



Using "goodness of fit" as the only criterion, the 

 most acceptable Ys estimate was 6,430 t. This Ys 

 was surpassed by 760 t in 1974 and by 1,160 t in 



1975. Catch remained at this maximum equilib- 

 rium level in 1976 and 1977, but declined to the 

 pre-1974 level in 1978 and 1979. In fact, all the 

 yield analyses except the linear model as applied 

 to standardized CPUE and effort data indicated 

 overfishing in 1974 and 1975. 



According to the "best" unadjusted fs estimate 

 (1,036 d), fishing effort in 1975 abruptly 

 exceeded — by 20*%^ — the level which produced the 

 maximum equilibrium catch and remained exces- 

 sive in 1976 and 1977. Observed effort declined 

 dramatically after 1977. Using standardized effort 

 as a guide, the most acceptable /^s estimate (1,096 

 standard days) was exceeded by an average an- 

 nual amount of 57% during 1975-77. Assuming 

 that the standardization procedure was accurate, 

 "real" effort remained slightly above 1,100 stan- 

 dard days in 1978 and 1979 even though catches 

 were considerably below the estimated Ys of 6,430 

 t. The exponential model predicted that the 

 maximum equilibrium yield would be maintained 

 when CPUE was 6.1 t/d or 5.9-6.0 t/standard day. 

 The linear model predicted considerably higher 

 Ys/fs values. 



A comparison of the predicted yield curves for 

 both models — as applied only to average observed 

 and standardized effort — with actual 1969-79 

 catches (Figures 7, 8) revealed a closer overall 

 agreement between catch and predicted yields for 

 the exponential model even though the linear 

 model produced a closer agreement for certain 



8.0-1 



. 1 — I — \ — \ — \ — \ — \ — rn — \ — \ — rn i r 



200 400 600 800 1000 1200 I4(X) 1600 1800 

 AVERAGE EFFORT (DAYS AT SEA) 



Figure 7. — Costa Rican thread herring yield curve predicted by 

 the exponential and linear surplus production models and 

 annual 1968-79 catch and average observed effort data. 



698 



