NSFDEB-BSF: Breaking Barriers to the Study of Traits that Affect Speciation and Extinction

随着新物种的出现和老物种的灭绝，地球上的生命在数百万年的时间里被塑造和重塑。生物学家试图了解有机体的特定特征--例如动物的大小或植物的花色--如何影响产生新物种或遭受灭绝的机会。该项目开发了计算工具，以帮助识别持续影响物种形成和灭绝的特征。它进一步应用这些工具来了解特定植物群的历史，包括那些具有农业和经济价值的物种。本科生的科学机会和指导专门针对军队退伍军人。研讨会和个人咨询将使年轻的研究人员掌握对未来各种职业有价值的技术技能。美国和以色列实验室之间的国际合作将培训学生和博士后研究人员，并举办研讨会，建立年轻科学家社区，使用系统发育比较方法进行工作。从系统发育树推断依赖于特征的谱系多样性是困难的。目前需要既强大又可靠的方法，并需要工具来区分方法执行的好坏。在这个项目中，两种新的建模技术针对了当前方法最近暴露的弱点。第一种是将序列进化结合到性状进化模型中，以允许在这两个水平上用过程来解释多样化比率的变化。第二个重点是多样化和性状变化之间的联系，而不是它们的价值，以避免假复制的统计问题。该项目提供的两个计算工具改进了现有和未来方法的使用。一个是由社区驱动的基准测试数据库，用于比较各种方法在不同场景中的执行情况。另一种是评估模型充分性的方法，以引导人们对经验结论和未来模型发展方向的信心。最后，本项目研究了对植物表型和遗传型有重要影响的两个性状的宏观进化动态：育种系统和全基因组复制。方法学上的进步和扩大的数据集为物种选择在塑造生物多样性方面的重要性提供了迄今最有力的证据。

As new species arise and older species become extinct, life on Earth is shaped and re-shaped over many millions of years. Biologists seek to understand how particular traits of organisms---the size of an animal or the flower color of a plant, for example---affect the chances of giving rise to a new species or suffering extinction. This project develops computational tools to help identify traits that consistently affect speciation and extinction. It further applies the tools to understand the history of particular plant groups, including ones with agriculturally and economically valuable species. Undergraduate science opportunities and mentoring specifically target veterans of the armed services. A workshop and individual advising will equip young researchers with technical skills that are valuable for a variety of future careers. The international collaboration between American and Israeli labs will train students and postdoctoral researchers and hold workshops to build communities of young scientists working with phylogenetic comparative methods.Inferring trait-dependent lineage diversification from phylogenetic trees is difficult. There is currently a need for methods that are both powerful and reliable, and for tools to distinguish when methods perform well or poorly. In this project, two novel modeling techniques target weaknesses recently revealed in current methods. The first incorporates sequence evolution into models of trait evolution, to allow shifts in diversification rate to be explained by processes at both those levels. The second focuses on connections between diversification and trait shifts, rather than their values, to avoid the statistical problem of pseudoreplication. Two computational tools provided by this project improve the usage of existing and future methods. One is a community-driven database of benchmark tests to compare how methods perform across diverse scenarios. The other is a means of assessing model adequacy, to guide confidence in empirical conclusions and directions for future model development. Finally, this project investigates the macroevolutionary dynamics of two traits that have major effects on plant phenotypes and genotypes: breeding system and whole genome duplication. Methodological advances and expanded datasets provide the potential to make the strongest cases yet for the importance of species selection in shaping biodiversity.