权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

DISSERTATION RESEARCH: Testing the effects of priors on prediction error in Bayesian demographic models

论文研究：测试贝叶斯人口统计模型中先验对预测误差的影响

基本信息

批准号：
0909604
负责人：
Aaron Ellison
金额：
$ 1.27万
依托单位：
University of Massachusetts Amherst
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2009
资助国家：
美国
起止时间：
2009-06-01 至 2011-05-31
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=0909604&HistoricalAwards=false
关键词：
DISSERTATION RESEARCH Testing effects priors

项目摘要

It is critical that forecasts of populations sizes and extinction risks applied to rare species have high accuracy and low uncertainty. One possible solution to reducing uncertainty in these models is to include information on closely related species in a Bayesian population viability analysis (PVA). This study tests the ability of Bayesian PVA to predict real future extinction risks and population sizes. To do this, the researchers will construct Bayesian models for two rare plant species, Calochortus tiburonensis and Pedicularis furbishiae, Calochortus tiburonensis and Pedicularis furbishiae, using external data from closely related congeners. These two species were originally studied in the 1980s, and the original populations will be re-censused. Estimates of the extinction risks and 2009 population sizes derived from different models will be compared with information on the populations obtained from contemporary censuses. This research provides one of the first tests to determine the appropriateness of using external data in Bayesian PVA. The results will be applied to improving conservation and recovery plans for the rare species studied. The researchers will make the modeling software available online and teach others how to use it at international and regional conservation conferences. Further this research promotes cross-disciplinary interactions and will train both a graduate student and an undergraduate in the use of Bayesian statistical modeling.

对稀有物种的种群规模和灭绝风险的预测必须具有高准确性和低不确定性。减少这些模型中不确定性的一个可能的解决方案是在贝叶斯种群生存力分析（PVA）中包括密切相关物种的信息。本研究测试贝叶斯PVA预测真实的未来灭绝风险和种群规模的能力。为此，研究人员将利用密切相关的同源物的外部数据，为两种稀有植物物种Calochortus tiburonensis和Pedicularis furbishiae构建贝叶斯模型。这两个物种最初是在20世纪80年代进行研究的，原始种群将被重新普查。根据不同模型得出的灭绝风险和2009年种群规模的估计值将与当代人口普查获得的种群信息进行比较。这项研究提供了第一个测试，以确定在贝叶斯PVA使用外部数据的适当性之一。研究结果将用于改善所研究的稀有物种的保护和恢复计划。研究人员将在网上提供建模软件，并在国际和地区保护会议上教其他人如何使用它。此外，这项研究促进了跨学科的互动，并将培训研究生和本科生在使用贝叶斯统计建模。