Methodologies to utilize fishing information have only a few results, though the development of them are emphatically needed. Hence, authors considered modeling of the productivity index of a fishing ground
M to be available for detection of f.c. (fishing conditions) or selection of fishing ground.
Using data of radio information on tuna longline f.c. among ships,
M was constructed as equations (1), (2) and (3), where
l1 and
l2: coefficients, μ, σ: mean and standard deviation of total catch in fishing ground;
H: information reliability index;
p: fales information probability;
N: number of ships.
H was adopted in the model for the purpose of correcting the probable false information in f.c. by radio. 1/σ signifies the stability of f.c. and was introduced into the model. μ can be an index of f.c. but changes in a manner of second degree curve. Hence, μ
2 as well as μ must be variables in the model.
Thus the model was constructed theoretically. The practicable equation (6) was obtained through the determination of coefficients by multivariates analysis, wherein
R (the mean of catch by selected training ships) was used as objective variable instead of
M. The results were shown really and satisfactorily in Fig. 4, which verifies the effectiveness of the model.
抄録全体を表示