Pollination ecology and genetic population structure of three species of cacao (Theobroma spp.) in Bolivia

热带雨林树种的特点是种群密度低，同种个体之间距离大。这种空间分布对基因流动有影响，特别是对传粉媒介的树间运动。在森林破碎化的背景下，了解基因流是理解热带树种繁殖成功的基础。本文以玻利维亚北部的三种可可树（Theobroma，可可树）为模型，利用不同树密度的样地，比较野生种群和栽培种群的传粉者组合、交配系统和基因流模式。我们认为，不同物种间、野生种群与栽培种群间的传粉生态存在差异，传粉者的时空分布主要由气候因素、同时开放花数和种群密度决定。我们进一步推测，在不同密度的种群中，自交与异交的比例不同，较高的种群密度、人工授粉和较高的遗传特异性提高了授粉成功率和坐果率。种群遗传结构可能是中等的，因为授粉和种子传播都依赖动物。除了实地研究包括传粉者捕获研究和坐果实验，我们将使用直接和间接的分子方法来评估基因流动，如空间遗传结构评估和亲本分析，以遵循这些假设。我们的研究结果对于理解热带森林生态系统中交配系统的演变具有普遍的意义。

Tropical rainforest tree species are characterized by low population densities, with great distances between conspecific individuals. This spatial distribution has consequences for gene flow, especially for the intertree movement of pollinators. In view of ongoing forest fragmentation, knowledge on gene flow is elementary for comprehending the reproductive success in tropical tree species. Here, we propose to use three understorey species of Theobroma (cacao) in northern Bolivia as a model system to compare pollinator assemblages, mating systems, and patterns of gene flow in wild and cultivated populations, using plots of different tree density. We assume that the pollination ecology of Theobroma species differs between species as well as between wild and cultivated populations, and that the spatial and temporal distribution of pollinators is mainly determined by climatic factors, by the number of simultaneously open flowers, and by tree population density. We further hypothesize that the ratio between selfing and outcrossing varies in differently dense populations, and that higher population density, artificial pollination and higher genetic distinctness enhance pollination success and fruit set. The population genetic structure will probably be moderate, due to the dependence on animals for both, pollination and seed dispersal. Besides field studies encompassing pollinator capture studies and fruit set experiments, we will use direct and indirect molecular methods for assessing gene flow such as assessment of spatial genetic structure and parentage analysis to follow these assumptions. Our results are of general interest for understanding the evolution of mating systems in tropical forest ecosystems.