Factors of Catch Fluctuations of Skipjack Tuna in the North-eastern Waters of Japan and Its Forecasting

Abstract

In this paper, we attempt to develop a reasonable catch forecasting method for skipjack tuna caught in the northeastern waters of Japan. In developing such a method, it is very important to find the main factors which control the catch fluctuations. The conventional method is based on the data of mean length and mean fatness of the fish harvested just before the fishing season, and the reproductive relationship estimated using the catch history. This approach, however, failed to forecast the catch in the years when the catch was extremely high. In this study, we therefore focused on the relationship be tween catches in the northeastern waters of Japan and the following three variables; (1) catch caught in the waters off Chiba Prefecture, (2) the Kuroshio current pattern, (3) the first date of capture of skip jack tuna in the waters off Chiba Prefecture, and then forecasted the catch using regression analysis The results showed that the accuracy of forecasting was very high almost every year including year when the catch was very high. It is considered that the method proposed here overcomes the problems of the conventional method.

Catch fluctuations of the diamond squid Thysanoteuthis rhombus in the Sea of Japan and models to forecast CPUE based on analysis of environmental factors

“…The goodness of fit and parsimony of the models were examined based on three values: (1) the significance level of the simple or multiple correlation coefficient r or R (F-tested for R), (2) the sum of squares (SS) for the difference between the observed and predicted CPUEs and (3) the Akaike information criterion (AIC) value, which provides a combined measure of statistical fit and model parsimony (Akaike, 1974). We also tested the extrapolation for forecasting using the method of Sakuramoto et al (1995). First, we estimated the parameters of regression models using data collected during 1989-1998 (n = 10) and forecasted the CPUE for 1999 using the independent variables for 1999.…”