大菱鲆选育家系体重性状的分形特征

分析了2007～2010年度18月龄体重性状的信息维数和关联维数。分析结果显示，2007～2010年度所构建的大菱鲆选育家系群体18月龄体重性状的信息维数介于0．9422～0．9586之间，不同年度之间的信息雏数变异不显著，这说明不同年度各选育家系群体18月龄体重性状的信息维数均为较大值，所统计群体的遗传变异都十分丰富，统计性状的变异范围也较大，显示出良好的育种潜力。相应数据的关联维数分析结果表明，除2008年度小尺度范围的关联维数是0．4284外，其余各小尺度范围的关联维数均高于0．73，这表明，从分析结果的总体来看，大菱鲆各选育家系群体内个体间的遗传结构的相关性较高；信息维数和关联维数作为研究分形特征的指标，从两个角度阐明了选育性状遗传结构的分形特征，在实际的良种选育过程中，应考虑这两方面的因素，确定适当的留种率。

Fractal properties of body weight of turbot selective breeding families

Fractal theory of nonlinear theory is an ideal method for studying the genetic structure and trait genetic variation of animal population. For animal breeding, it is very impor tant to clarify genetic structure of the selected quantitative traits. In this study, the body weight of 18-month-old turbot in four selective breeding families during 2007-2010 was analyzed using fractal theory. The information dimension and correlation dimension were obtained. The results revealed that the information dimension of 18-month-old turbot body weight ranged in 0. 942 2- 0. 958 6 from 2007 to 2010, which showed that all the four populations had high dimension, high variation, and high breeding potential. The correlation dimensions were higher than 0.73except for that of 2008, which showed that the correlations ot the traits in the lour populations were all relatively high; the information and correlation dimensions have illuminated the fractal properties of the traits in the four breeding populations from different points of view, which can provide a theoretic basis for determining appropriate selection proportion during selective breed ing of turbot.