NSF Postdoctoral Fellowship in Biology FY 2014

美国国家科学基金会授予雷切尔·M·米切尔的生物学博士后奖学金支持一项名为《使用分层贝叶斯模型预测多样性模式》的研究。NSF生物学博士后奖学金将研究和培训部分结合在一起，为年轻科学家在生物学与其他科学学科交叉的新兴领域的职业生涯做准备；这项研究是生物学和数学的交叉学科，特别是统计建模。预计这些研究员将成为国家未来科学劳动力的领导者。该奖学金的主办机构是杜克大学，赞助科学家是贾斯汀·赖特博士和艾伦·盖尔芬德博士，他们可以帮助培训米切尔博士建立分层贝叶斯模型。因此，米切尔博士在应用于生态学的分层贝叶斯建模方面获得了宝贵的技能。她还扩大了外展和教学机会；与生命与科学博物馆合作，使她能够与来自不同背景的孩子分享与研究问题直接相关的动手活动。她还为杜克大学的研究生举办了一个简短的科学推广和交流研讨会。在过去的十年里，人们对植物功能特征的兴趣有了巨大的增长，这些特征是影响植物适应性的可测量特征。据了解，这些特征现在被降雨、温度和土壤条件等非生物过滤器主动选择。这样的选择是为了确定植物群落是如何形成的，并解释为什么一些地区的物种比其他地区多。直到最近，一个物种内个体之间的性状差异(=种内差异)在很大程度上被忽视了，但最近的研究表明，这种差异是实质性的，在生态上具有重要意义。尽管植物性状和种内变异对群落的形成至关重要，但它们尚未完全纳入预测物种存在和多样性模式的模型中。米切尔博士正在建立的模型结合了植物功能特征、非生物过滤器和种内变异，以预测美国东南部大小规模的多样性模式。这个模型是对生态学和统计模型领域的重大贡献。这个项目对地方和全球尺度的保护都有影响。该模型量化了特征如何调节物种在非生物条件下的存在，这在预测的气候变化情景下是一个重要的问题。此外，它还有助于确定物种特性和功能特征方面的高度多样性区域；这些区域在不断变化的世界中具有很高的保护价值。它澄清了物种和功能多样性之间的联系，如果要保护生态系统及其提供的服务，就必须理解这种联系。

This NSF Postdoctoral Fellowships in Biology to Rachel M. Mitchell supports research entitled `Using Hierarchical Bayesian Models to Predict Patterns of Diversity.` NSF Postdoctoral Fellowships in Biology combine research and training components to prepare young scientists for careers in emerging areas where biology intersects with other scientific disciplines; this research is at the intersection of biology and mathematics, specifically statistical modelling. Fellows are expected to be leaders of the nation`s scientific workforce of the future. The host institution for this fellowship is Duke University and the sponsoring scientists are Dr. Justin Wright and Dr. Alan Gelfand, who can help train Dr. Mitchell in building a hierarchical Bayesian model. Thus Dr. Mitchell is gaining valuable skills in hierarchical Bayesian modelling with applications to ecology. She is also expanding outreach and teaching opportunities; partnering with the Museum of Life and Science allows her to share with children from diverse backgrounds a hands-on activity directly related to the research question. She is also implementing an abbreviated science outreach and communication seminar for graduate students at Duke University.The last decade has seen a tremendous growth of interest in plant functional traits, which are measurable traits that impact plant fitness. These traits are now understood to be actively selected on by abiotic filters like rainfall, temperature, and soil conditions. Such selection is hypothesized to determine how plant communities form and explain why some areas have more species than others. Until recently, the trait variation among individuals within a species (= intraspecific variation) has been largely ignored, but recent research has revealed that this variation is substantial and ecologically important. Although plant traits and intraspecific variation are known to be critical in shaping communities, they have yet to be fully incorporated into models predicting patterns of species presence and diversity. The model being built by Dr. Mitchell incorporates plant functional traits, abiotic filters, and intraspecific variation to predict patterns of diversity at both large and small scales in the southeastern United States. This model - the first of its kind - is a significant contribution to the fields of both ecology and statistical modeling.This project has implications for conservation on local and global scales. The model quantifies how traits mediate species presence in response to abiotic conditions, an issue of importance under projected climate change scenarios. Furthermore, it serves to identify areas of high diversity in terms of both species identity and functional traits; such areas are of high conservation value in a changing world. It clarifies the link between species and functional diversity, a link that must be understood if ecosystems and the services they provide are to be preserved.