Collaborative Research: Novel framework for estimating continuously-varying diversification rates

长期以来，生物学家和古生物学家一直感兴趣的是，确定在生物群体（如昆虫和哺乳动物）之间形成物种数量巨大差异的因素。理解物种多样性差异的数学模型使用两个量，即物种形成率（一个物种变成两个物种）和灭绝率（一个物种消失）。这些数值与用于研究人口、出生率和死亡率的数值相似。尽管古生物学家已经对物种数量的差异提出了解释，但生物学家主要是基于对遗传数据的分析，利用物种家谱来解释这些模式，他们基本上得出的结论是，物种形成比灭绝更重要。这项研究的中心主题是，驱动现代物种多样性的不是物种形成的差异，而是“周转率”的差异，周转率被定义为出生率减去死亡率。周转率衡量的是物种灭绝和物种形成事件在很长一段时间内发生的频率。这项研究将扩展现有的估计物种多样性的数学模型，包括使用数量的新模型，比如周转率，这些模型与生物学的关系更密切。该项目还将在美国东南部（一个重要但代表性不足的地区）提供这些计算技能的培训讲习班。该研讨会将培养学生和其他早期职业研究人员成为具有计算素养的STEM劳动力，这对国家的经济福祉至关重要。该项目还将为两名博士后提供专业培训机会。达尔文的伟大见解是拒绝类型学思维，赞成认识生命中的变异。然而，在多样化分析中，类型学思维不仅仅是为了方法论上的方便而制造的方便拐杖。寻找物种从一种速率“类型”瞬间转变为另一种速率“类型”的点受到了不合理的关注。这项研究将摆脱这种类型学思维，通过假设多样化率持续变化，向更现实的生物学方向发展。这个框架的核心是一个速率“遗传”的主题，在这个主题中，多样化速率会进化，并从祖先遗传给后代，就像树上的特征会进化一样。这项研究将把重点放在物种形成和灭绝，甚至是净多样化上，而不是集中在物种更替和灭绝比例的估计上。最重要的假设是，消光率接近但可能略小于1。换句话说，随着时间的推移，物种形成事件的数量略多于灭绝事件。因此，驱动多样化模式的不是独立于灭绝的物种形成率，甚至不是灭绝率的变化，而是整体的周转率。这项研究将开发一种新的基于可能性的框架，用于评估一大类多样化模型，这些模型更现实地将速率视为树上的连续变化，并将估计重点放在新的、更机械的参数化上。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Determining the factors that have shaped the huge differences in the number of species among groups of organisms, such as insects versus mammals, has long been of interest to both biologists and paleontologists. Mathematical models for understanding differences in species diversity use two quantities, the rate of speciation (when one species becomes two) and the rate of extinction (when a species disappears). These quantities are similar to those used in studying human populations, the birth rate and death rate. Although paleontologists have proposed explanations for differences in the numbers of species, biologists, using mostly family trees of species based on the analysis of genetic data, have largely concluded that speciation is more important than extinction in explaining these patterns. The central theme of this research is that what drives modern species diversity is not differences in speciation, but rather the differences in "turnover rate," which is defined as the birth rate minus the death rate. Turnover rate measures how often extinction and speciation events happen over long periods of time. This research will expand existing mathematical models for estimating species diversity to include new models that use quantities, such as turnover rate, that are much more tied to biology. The project will also provide a training workshop in these computational skills in the southeastern U.S., an important and underrepresented part of the country. This workshop will train students and other early career researchers to become a computationally literate STEM workforce, which is important to the economic well-being of the nation. The project will also provide professional training opportunities for two postdoctoral researchers.Darwin's great insight was a rejection of typological thinking in favor of recognizing the variation across life. Yet, in diversification analyses, typological thinking is not just a convenient crutch made for methodological convenience. Finding the points at which species shift from one rate "type" instantly to another has received unjustified attention. This research will move away from this typological thinking, towards more biological realism by assuming that diversification rates continuously vary. Central to this framework is a theme of rate "inheritance," where diversification rates evolve and are inherited from ancestor to descendant, much in the same way as a trait would evolve on a tree. Rather than focusing on speciation and extinction, or even net diversification, this research will focus biological interpretations on estimates of turnover and extinction fraction. The overarching hypothesis is that extinction fraction is near but probably somewhat less than one. In other words, speciation events slightly outnumber extinction events over time. What drives diversification patterns, therefore, is not speciation rate varying independent of extinction, or even extinction rate varying, but rather the overall turnover rate. This research will develop a new likelihood-based framework that evaluates a large family of diversification models that more realistically treat rates as continuously varying on a tree and focuses estimation on a new and more mechanistic parameterization.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.