中文摘要：

研究残存居群的数量性状分化及其地方适应性有助于了解不同居群对环境的适应性并制定相应的保育策略。本研究采用同园实验，利用随机区组设计，对中华水韭（lsoete ssinensis）3个居群的9种优势基因型的数量性状进行巢式方差分析，并采用Bayesian方法计算等位酶（FST）和数量性状（QST）的居群分化值。结果表明，在测量的14个数量性状中，有10个性状在居群间有显著性差异，仅有3个性状在居群内有显著性差异；多重比较表明其中5个性状的平均值均以松阳居群最高，休宁居群最低；建德居群的大孢子囊的性状和植株高度的平均值最高，休宁居群最低；推测这可能是与奠基者效应、物种间竞争及中华水韭在自然生境下形成的营养生长和生殖生长之间的权衡有关。对居群数量性状分化的QST值和等位酶分化的FST值分析比较发现，在假设为自交系时有8个性状的QST值均大于FST表明存在明显的地方适应性，居群间个体的混合存在潜在的远交衰退的风险，所以不宜采用居群间相互移植个体的方法来进行中华水韭居群的遗传复壮。对休宁和松阳两个居群的保育应采用提高居群内基因流和改善生境环境的方法，促进居群内的遗传复壮；对建德居群应尽量减少人为干扰，宜建立特定生境的保护小区。

英文摘要：

Aims lsoetes sinensis （Isoeteaceae） is a critically endangered aquatic fern ally in China. Understanding the genetic differentiation of quantitative traits and local adaptation of the remnant populations should provide insight into the adaptation of different populations to their local environment and useful information for formulating appropriate conservation strategy. Methods The quantitative traits of nine isozyme multilocus genotypes from three remnant populations of I. sinensis were investigated using nested analysis based on random block design in common-garden experiments. Population differentiation of allozyme （FST） and quantitative traits （QST） were estimated using the Bayesian approach. Important findings Of 14 quantitative traits examined, ten were found to be significantly different among populations and three were significantly different within populations. Multiple comparison by Turkey tests indicated that the mean values of five traits were the highest for the Songyang population and lowest for the Xiuning population, whereas the mean values of the macrosporangium characteristics and plant height were highest for the Jiande population and lowest for the Xiuning population. This result probably reflected different scenarios of founder effect, habitat competition ability and trade-off between vegetative and reproductive growth. Local adaptation was evaluated as a difference of QST vs. FST value, and eight important reproductive fitness traits were found to be significant （ QST 〉 FST, p ≤ 0.05 ） under the selling assumption, suggesting local adaptability in the remnant populations. It is more likely that there is a potential risk of outhreeding depression if genetic enhancement is implemented by translocation of individuals from different populations. Therefore, translocation among these remnant populations is not recommended. Instead, we recommend increasing gene flow within each population and improving habitat management by reducing competition with companion species in in s