MENDELSSOHN: USING BOX-JENKINS MODELS IN FISHERY DYNAMICS 



Table l.— Autocorrelation functions for 12th differenced efTort series of the Hawaii skipjack tuna fleet, 1964-78. 



Item 



15 



16 



17 



18 



Lag (mo) 

 19 20 



21 



22 



23 



24 



25 



26 



27 



Table 2. — Autocorrelation functions for 12th differenced catch series of the Hawaii skipjack tuna fleet, 1964-78. 



were hypothesized as the appropriate univariate 

 models for both the catch and the effort time 



series. 



ESTIMATION AND CHECKING 



Given a tentative model, such as Model (1), the 



next step is a recursive procedure of estimating 

 the parameters of the model, calculating the 

 autocorrelation and partial autocorrelation func- 

 tions of the residuals from the estimated model, 

 and then testing the residuals for significant 

 departure from the assumption that they are 

 white noise. When a final model has been identi- 



o 



0.7 

 06 

 0.5 

 0.4 

 0.3 

 0.2 

 0.1 

 

 -0.1 

 -0.2 

 -0.3 

 -0.4 

 -0 5 

 -0.6 



FIGURE 



IT 

 O 

 O 

 O 



1- 



< 



p — I !— 



12 



18 24 30 



LAG IN MONTHS 



36 



42 



48 



4. — Estimated effort autocorrelation function for the fishing trips by the Hawaii skipjack tuna fleet near Oahu, 



Hawaii, 1964-78. 



891 



