珪藻赤潮の重回帰式と海域区分について〔広島湾奥部の赤潮発生予察に関する研究-3-〕

概要

論文の詳細を見る
The simultaneous observations were carried out at 37 stations using six boats in June and september of 1979. In this paper, the multiple regression equations of diatom red tides were calculated with stepwise forward method of multiple regression analysis (MRA) and the classification of the northern part of Hiroshima Bay were done with cluster analysis. The results obtained are as follows. 1) With five predictor variables of MRA, the partial regression coefficients found to be significant at the 1% level were salinity (Sal.), total dissolved phosphorus (TDP), dissolved inorganic nitrogen (DIN) in June and water temperature (WT), salinity, TDP, DIN in September respectively. 2) In the best multiple regression equation by MRA with nine predictor variables, the coefficients of determination were found to exceed the 80% level in June and september. 3) In all cases, the multiple correlation coefficients showed a high value of 0.9 and the multiple regression equations were significant at the 1% level. 4) From the cluster analysis (D=1.7) based on six environmental variables, the 37 stations of the northern part of Hiroshima Bay were divided into twelve sea areas.
日本水産學會の論文