Quantitative assessment of forest fragmentation on genetic components of understory herbs.

为了评估森林片断化对林下草本植物遗传组成的影响，对延龄草属（Trillium）2种植物的生殖、种群统计学和遗传特性进行了研究。在北海道东部的T.camtschatcense种群显示，或被丰富的森林系列包围的连续种群比小的和孤立的种群产生更多的种子。利用6种酶11个位点的等位酶电泳分析了群体的遗传多样性水平、近交系数和群体结构特征。群体大小的对数与多态位点百分率和等位基因丰富度呈显著正相关。为了预测随机变化环境下的种群规模，本研究采用转移矩阵模型对长期调查数据进行分析，并结合田间育种试验，研究了无瓣叶甲的种群动态。矩阵模型表明，处于一叶不育期的幼苗和植株等小个体的死亡率很高，从一叶期过渡到三叶期和从三叶期过渡到开花期的概率很低。虽然无瓣锥蝽具有很强的近亲繁殖和远交繁殖能力，但近亲繁殖仍占主导地位。此外，随后的资源限制只允许一定水平的繁殖力，无论原来的施肥水平。

In order to assess the effects of forest fragmentation on genetic components of understory herbs, reproductive, demographic and genetic characteristics of two Trillium species were studied.Investigation on seed production of, self-incompatible and insect-pollinated, T.camtschatcense populations in eastern Hokkaido revealed that large populations and/or continuous populations surrounded with abundant forest series produced a higher number of seeds than small and isolated populations. The levels of genetic diversity, inbreeding coefficient, and the patterns of local population structure were also assayed using allozyme electrophoresis (6 enzymes and 11 loci). The log of population size had positive and significant relationships with the percentage of polymorphic loci and allelic richness. In addition, all alleles which were not observed in small populations were rare.In order to forecast the future population size under stochastically varying environment, population dynamics of T.apetalon were investigated by application of the transition matrix model to the long-term census data and by breeding experiments in the field. The matrix model demonstrated that mortality rates of small individuals, such as seedlings and plants at the one-leaf sterile stage, were very high and that the probabilities for transition from the one-leaf stage to the three-leaf stage and from the three-leaf stage to flowering were very low. Although T.apetalon had a substantial capacity both for inbreeding and outbreeding, inbreeding was predominant. Furthermore, subsequent resource limitation allowed only acertain level of fecundity, regardless of the original fertilization level.