中文摘要：

【目的】从分子标记等位变异水平上探讨中国野生大豆群体的遗传特征、连锁不平衡特点和地理生态分化的遗传机制，并以重要生态性状全生育期为代表解析性状地理分化的遗传基础。【方法】从全国24个省区不同地理生态型的野生大豆材料中抽选174份组成代表性样本，选用204个SSR标记，利用TASSEL及STRUCTURE2．2软件进行群体连锁不平衡（1inkagedisequilibril．1m，LD）和群体遗传结构分析。在此基础上对群体的地理分化、亚群体特异性及全生育期位点等位变异的地理分化进行遗传分析。【结果】中国野生大豆群体蕴含丰富的遗传变异，20条连锁群中，I和c2连锁群有相对较多的位点平均等位变异和遗传分化。不论是共线性组合，还是非共线性组合，都有一定程度的LD存在，说明历史上发生过连锁群间的大量重组；野生群体口’平均值为0．34，高值多，比栽培大豆高，说明野生群体发生过更多的重组，保留下较高的LD。采用H-W平衡模型将野生群体聚成4类，模型聚类亚群划分与地理生态分类相关、有交叉，推测各地理亚群体发生过材料的迁移。各地理亚群经长期自然选择，各位点等位基因的频率发生变化，还有新生的与绝灭的，因此，各地理亚群问产生明显的等位基因分化。与地理生态性状全生育期关联的有15个位点160个等位变异，其中，亚群特有等位变异共58个，来自14个位点，同一位点可以在多个亚群体产生不同的特有等位变异，其效应从南至北逐渐下降，这说明生态区间不仅有强烈的遗传分化，而且是有规律的遗传分化。【结论】中国野生大豆群体遗传多样性高，共线、非共线位点问连锁不平衡程度高，4个生态亚群体问位点高度遗传分化，产生大量地区特有等位变异，全生育期还表现由南向北降效的规律性遗传分化。

英文摘要：

[Objective] The objective of the present study was to explore the genetic characteristics, linkage disequilibrium （LD） and genetic mechanism of geographic differentiation of the Chinese wild soybean population, and based on it, the geographic differentiation mechanism of days to maturity, the key geographic-ecological trait, is to be revealed. [Method] A sample composed of 174 accessions of wild soybean （Glycine soja Sieb. et Zucc,） from all growing areas （24 provinces） in China was established as the representative population of the Chinese wild soybean. The genotyping data of 204 simple-sequence repeat （SSR） markers on 174 accessions were obtained and analyzed for LD among marker pairs and the genetic structure of the population by using the TASSELand STRUCTURE 2.2 program. Based on it, the genetic mechanism of the geographic differentiation of the population, the characteristics of the geographic subpopulations, and geographic differentiation of alleles of loci related to days to maturity were studied. [ Result ] There showed a plenty of genetic diversity in the Chinese wild soybean population and showed relatively larger average number of alleles and genetic differentiation on Linkage group （LG） I and C2 than others among the 20 LGs. Different LD degrees were detected not only among syntenic marker pairs but also among nonsyntenic pairs, suggesting there had been historical recombination among linkage groups. The average D＇ was 0.34 for the population, larger than the previously reported value for cultivated soybean （G max （L.） Merr.）, implying more recombination occurred in the history and therefore higher LD values remained at present. The population was classified into four subpopulations according to H-W model-clustering analysis which was not completely but somewhat consistent with the geographic classification, indicating multiple migrations among eco-regions happened in the long history. There showed a great deal of allelic differentiation among the geographic subpopulatio