On Thursday, May 15 @ 11am ET, join us for a virtual scientific seminar by Dr. Laura Dee, University of Colorado, Boulder.
A fundamental aim in basic and applied ecology is to understand and quantify causal relationships between ecological variables and processes. Traditionally, experiments have been the predominant approach to causal inference in ecology. While experiments are critical—and often the gold standard—they face important challenges, including logistical, ethical, and spatial scale constraints, that may limit their feasibility at large scales and ability to generalize inferences. This underscores the need for complementary approaches that leverage observational data. Causal inference in observational data is important for 1) testing ecological theory in natural and human-dominated ecosystems; 2) scaling up ecological understanding to landscapes; and 3) evaluating the effectiveness of conservation interventions or assessing the impacts of large-scale disturbances such as extreme events. However, causal inference in observational data faces many challenges and sources of bias.
In this talk, Dr. Dee will present her research group’s ongoing efforts to demonstrate how interdisciplinary approaches to causal inference can help advance each of these aims for ecological research. Dr. Dee will show how advances from other fields, including economics, epidemiology, and computer science, are used to tackle some of these thorny challenges inherent in causal inference from observational data.
She will first share research that tests ecological theory by quantifying the effects of biodiversity on productivity across natural systems worldwide. Second, she will share research combining remote sensing and causal inference to scale up estimates of invasive species impacts to biodiversity and ecosystem functioning in South Africa, drawing on new data from the NASA BioSCape campaign. Finally, she will present research using quasi-experimental approaches from other disciplines to evaluate the impacts of fire—and the effectiveness of hazardous fuels treatments—on multiple ecosystem services, particularly recreation in Colorado.
Dr. Dee will end by reflecting on the opportunities and remaining challenges of using causal inference in observational ecological research, and how cross-disciplinary approaches can help us better understand and manage ecosystems in an era of rapid global change.
Free and open to all. Registration required via Eventbrite.
