How Well Do Phenology Models Established In Michigan And Ohio Predict The Timing Of Activity In Plants And Insects Across Their Ranges?
Presenter: ICREWW Team P111
Co-Author(s): Kyle Skoda, Daniel Williams, Dylan Girone, Ava Lasater, Brandolyn Baeza, Alex Winter, Theresa Crimmins, Martha Whitaker
Advisor(s): Drs. Martha Whitaker & Theresa Crimmins
1SNRE, EEB, HAS, SGDE
Plant and pest activity are intrinsically connected so understanding the timing of these phenomena can better help managers plan for and mitigate risks. This subject has been the topic of recent scientific inquiry; Herms (2004) constructed degree-day models for dozens of species in the Ohio-Michigan region. The geographic extent of the plants studied by Herms extends well beyond the boundaries of the study area, so we aim to assess the ability of these models to accurately forecast the timing of plant and insect activity across the United States. Our team used observations contributed by volunteers to the USA National Phenology Network’s Nature’s Notebook (USA-NPN) to test these models for 21 plant and one insect species. The USA-NPN data were collected from a diverse array of sites, ranging from Florida to Maine, providing an ideal testbed. We encoded the pre-established models into the Python programming language to automate the prediction of expected dates for the different growth stages of the target species. We utilized daily gridded data for maximum, minimum, and mean temperature from the PRISM Climate database as inputs to the Herms’ models. We found that, on average, the phenological events occurred 30-50 days after Herms’ predictions. The results have the potential to create a framework for future model development, starting from more precise local measurements (e.g. Herms, 2004) and extrapolating to a larger scale. The results also demonstrate the effectiveness of citizen science data in contributing to research.