太湖主要鱼类渔获量间关系的回归模型分析

根据太湖1983～2002年连续20年鱼类渔获量统计资料,应用偏相关分析及逐步回归分析等方法,研究了太湖鱼类组成、数量变化的连续变化规律,并结合鱼类食物组成调查对主要鱼类间的相互影响进行了分析。结果表明:除青、草鱼和其余主要鱼类间不存在线性关系外,建立的6个多元线性最优回归模型其复相关系数均达到极显著水平,说明统计学意义上的可靠性很高,较好地反映了近20年太湖主要鱼类组成和数量变化的比例关系。分析认为,其中鲚鱼模型W1=12803-5.5246W3+2.3736W4-3.8909W5、银鱼模型(W2=186.7576+1.5174W3+0.0919W7)、鲢、鳙鱼模型(W5=1633.5789-0.0837W1+0.2553W4)和其他鱼类模型(W7=7652.7351-0.4232W1-5.3439W3+2.2624W4)可能还反映了太湖主要鱼类间相互影响的关系。

Regression Model Analysis of the Relationship among Catches of Main Fishes in Lake Taihu

Based on the statistics of fish catches from 1983 to 2002 in Lake Taihu,with the methods of correlative and stepwise regression analysis,the relationship among catches of main fishes had been studied. Results showed that except for no correlation between the catch of black carp and grass carp and the catch of other main fishes. Six stepwise regression models were educed and their correlative coefficients reached to very significant at the level of <0.001.According to the investigation of main fishes' food composition,mechanisms of the relationship among catches of main fishes were analyzed. The conclusions were that the six stepwise regression models described the change rules of main fishes' catches in the lake; however the four stepwise regression models W1=12 803-5.524 6 W3 2.373 6 W4-3.8909 W5,W2=186.757 6 1.517 4 W3 0.091 9 W7,W5=1 633.578 9-0.083 7 W1 0.255 3 W4 and W7=7 652.735 1-0.423 2 W1-5.343 9W3 2.262 4 W4 maybe express the mutual relationship among main fishes in the lake.