Quantifying and Reducing Uncertainty in Plant Demographic models

• Applicant: Aldo Compagnoni, Ph.D.
• Subject Area: Ecology and Biodiversity of Plants and Ecosystems
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 533825323

Understanding and predicting population dynamics is one of the central goals of ecological science which justifies expensive efforts to obtain data on individual-level demographic processes such as survival, growth, and reproduction. Despite these efforts, however, statistical estimates on demographic processes present substantial uncertainties. These uncertainties have large implications for inference and forecasting, and it is still unclear to what extent such uncertainties can be reduced. We hypothesize that uncertainty and its implications depend on species life history (e.g. species generation time). Moreover, we hypothesize that uncertainty can be reduced by improving measurement methods. RUPDemo is a project that will act in synergy with the CLIMVAR Eigene Stelle project. Within RUPDemo, a PhD student will test these hypotheses in the first project to systematically address the implications of uncertainty across the plant kingdom. First, the PhD student will use a 19-year long dataset on a short-lived perennial plant as a case study to perform the first uncertainty analysis on an integral projection model (IPM). Uncertainty analyses disentangle the combined importance of absolute uncertainty, and the sensitivity of population dynamics to said uncertainty. This project will identify the sources and implications of uncertainty in one of the best replicated plant demographic datasets. Second, the PhD student will explore the correlation between life history and IPM uncertainty. We will perform analyses on the PADRINO IPM database, which reproduces hundreds of IPMs published in the literature. In RUPDemo, the PhD student will double the number of IPMs contained in PADRINO, and correlate uncertainty analyses to species life history (namely, the position along the fast-slow continuum). The results will indicate opportunities to improve our understanding of plant population dynamics. Third, the PhD student will perform a field test on how improvements in size growth measurement affect our inference and forecasting of population dynamics. The student will update size measurement methods of our 19-year long dataset. Then, the PhD student will quantify how this new method decreases parameter uncertainty, decreases uncertainty in population-level estimates (e.g. population growth rate), and improves forecast skill. This work package has the potential to influence demographic data collection protocols worldwide.

• DFG Programme: Research Grants
• Co-Investigator: Professorin Dr. Tiffany Marie Knight