VAMLIS GeoCon 2024

Natural disturbances, such as wildfires and burns, play a crucial role in promoting plant diversity through the creation of open canopy spaces and the encouragement of understory growth. Over time, many plant species evolved to thrive in routinely fire-disturbed environments. However, the suppression of natural wildfires by humans has led to the decline of ecological niches created by fire disturbance, particularly across the eastern United States. Today, many species relying on disturbance are found in periodically mowed or maintained corridors such as transmission lines, road verges, and gas pipelines. These areas are managed intermittently, allowing for plant regrowth while preventing forest encroachment, making them ideal habitat for various threatened and endangered (T&E) species.

The North Carolina Department of Transportation (NCDOT) has taken on the challenge of identifying optimal habitats for various T&E species by developing a habitat prediction model using geospatial data. With the knowledge of how disturbed areas are critical for some T&E occurrences, ecologists are interested in better identifying disturbed habitats where T&E species often thrive. To enhance this habitat prediction model, Stantec is developing a machine learning approach to systematically identify linear disturbances in forested landscapes. By utilizing publicly available orthoimagery, Stantec’s model can accurately detect these disturbed areas across North Carolina. In this presentation, we will explore the importance of identifying T&E habitats, the role of disturbed areas in habitat prediction, and how machine learning can generate extensive, systematic datasets to support large ecological goals.