Collaborative Research: MRA: Evaluating hypotheses of long-term woody carbon dynamics with empirical data

埋藏在湖泊、沼泽和沼泽沉积物中的化石花粉记录了从上一个冰河时代末期到19世纪之间一万年来美国中西部北部植被的变化。使用这一化石记录是必要的，因为书面的植被调查直到工业时代才开始。过去对这些花粉组合的研究表明，随着气候变化(如冰川后变暖)和自然干扰(如阻止树木入侵草原的野火)，植被发生了变化。该项目利用新的统计和计算机模拟方法，以对社会有用的方式量化冰川后中西部植被和气候的变化(几个世纪以来景观上树木的数量和生物量的变化)。更好地量化工业化前的中西部植被非常重要，原因有两个。首先，工业化前的景观是以下方面的关键基准：土地管理者(养护、木材管理等的基线)；全球变化研究人员(工业化前植被生物量的变化是全球碳循环中最不确定的组成部分之一)；以及公众(工业化前植被有助于确定土地的特征)。该项目将通过以下方式与每个社区接触：会议、公开提供的数千年植被变化的重建，以及免费的在线教育模块，将向从高中生到专业资源管理者的广泛学生传授。第二，更好地量化长期植被填补了我们对植被变化生态学的基本理解的一个关键空白。研究人员知道，气候变化和火灾等干扰可能会突然不可逆转地改变大面积的植被，但缺乏长期植被变化的良好量化记录，一直是生态学理解这种变化实际发生的方式和时间的障碍。该项目使用一个综合建模框架，以确保新的植被变化重建将补充现有的对植被变化的理解，而现有的理解主要是基于近几十年来收集的数据。除了会议、公众宣传和在线培训外，该项目还将教育和培训4名定量古生态学研究生。该项目建立并改进了最近开发的一个统计模型，该模型描述了过去10,000年来中西部地区地上木本生物量池(AGWB)的变化。这些对AGWB在空间和时间上的重建是对一系列关于中西部植被历史的假设的检验，这些假设的中心是整个问题，即内部反馈(例如，植被和野火之间)是否在稳定或破坏AGWB对全新世气候变化的反应方面发挥了重要作用，以及现代土地利用是否显著改变了这些反馈。该项目将对变化的生物质池、森林人口比率以及受AGWB重建、全新世气候轨迹模型和森林变化生态过程替代模型约束的气候驱动因素进行联合概率推断。植被对气候的非线性阈值响应是这些追溯性“事后预测”的一个特别焦点。该项目开发的量化工具通过练习、模块和课程为学生、土地管理人员和其他研究人员提供教育和外联计划，研究人员将在一系列不同的短期课程中测试这些课程，然后在网上向公众提供。四名研究生的培训也将与研究和推广目标交织在一起。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Changes in the vegetation of the American Upper Midwest over a ten-thousand year period, between the end of the last ice age and the 19th century are recorded in fossil pollen buried in the sediments of lakes, swamps and bogs. It is necessary to use this fossil record because written vegetation surveys did not begin until the industrial period. Past investigation of these pollen assemblages has revealed shifts in vegetation with climate change (such as post-glacial warming) and natural disturbance (such as the wildfires that prevented trees from invading prairies). This project takes advantage of new statistical and computer modeling approaches to quantify post-glacial changes in Midwestern vegetation and climate in terms that are useful to society (changes in the numbers and biomass of trees on the landscape over centuries). Better quantification of pre-industrial Midwestern vegetation is important for two reasons related to this project. First, the pre-industrial landscape is a key benchmark for: land managers (a baseline for conservation, timber management, etc.); global change researchers (changes in pre-industrial vegetation biomass are among the most uncertain components of the global carbon cycle); and the public (pre-industrial vegetation helps define the character of the land). This project will engage with each of these communities through: conferences, publicly-available reconstructions of thousands of years of vegetation change, and free, online education modules that will be taught to a broad array of students, from high-schoolers to professional resource managers. Second, better quantification of long-term vegetation fills a key gap in our basic understanding of the ecology of vegetation change. Researchers know that changing climate and disturbances, like fire, can change vegetation suddenly and irreversibly over large areas, but the absence of well-quantified records of long-term vegetation change has been an obstacle to ecological understanding of how and when such changes actually occur. This project uses an integrated modeling framework to ensure that the new reconstructions of vegetation changes, which span centuries of tree-lifetimes, will complement the existing understanding of vegetation change, which is primarily based on data only collected over recent decades. In addition to the conferences, public outreach and online training, this project will also educate and train 4 graduate students in quantitative paleo-ecology.The project builds on and improves a recently-developed statistical model of changes in aboveground woody biomass pools (AGWB) across the Midwest over the last 10,000 years. These reconstructions of AGWB over space and time are the data that test a set of hypotheses about Midwestern vegetation history centered on the overall question of whether internal feedbacks (for example between vegetation and wildfire) played important roles in stabilizing or destabilizing the response of AGWB to changing climates over the Holocene, and whether modern land-use has altered these feedbacks significantly. The project will make joint probabilistic inference about changing biomass pools, demographic rates in forests, and climate drivers constrained by the AGWB reconstructions, an ensemble of modeled Holocene climate trajectories, and alternative models of the ecological process of forest change. Nonlinear threshold responses of vegetation to climate are a special focus of these retrospective “hindcasts”. The quantitative tools developed in the project underlie education and outreach programs to students, land managers, and other researchers through exercises, modules, and curricula that the researchers will test-drive in a diverse set of short courses, then made available to the public online. Training of four graduate students will also be interwoven with the research and outreach objectives.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.