The effect of seasonal anomalies of seawater temperature and salinity on the fluctuation in yields of small yellow croaker, Pseudosciaena polyactis, in the Yellow Sea

To include the effects of environmental factors on the production of small yellow croaker, Pseudosciaena polyactis Bleeker, in the Yellow Sea, we applied time series analysis to the commercial catch and salinity and temperature data for the period 1970 to 1988. Residuals from a weighted least-squares regression of log-transformed catches against year and month were calculated to remove not only seasonal factors but also long-term trends in catches. The residuals of mean and standard deviation (SD) of temperature and salinity were calculated and used for autocorrelation, crosscorrelation and first-order autoregression analysis (AR(1)) using maximum likelihood. The landings showed a decreasing pattern across years with a conspicuous seasonal cycle within years. Catch residuals showed a strong positive autocorrelation and a conspicuous time-lagged cross-correlation with the residuals of mean and SD of seawater temperature at 75 m. AR(1) revealed that positive anomalies of mean temperature were associated with positive anomalies in the production of small yellow croaker with a one year time lag. The decrease in the residual of SD of temperature appears to be related to the high production 0.5-1.0 year later. The effect of salinity was negligible compared with that of temperature. Therefore, the warm spawning period and homogeneous temperature condition of previous years for young fish may cause the increase in the following year's yield of this fish species. When used to predict catches in 1989 and 1990, the AR(1) model explained 40% of the variances of the observed landings.