An interdisciplinary approach to building data literacy in wildlife survey technologies
June 2019 - August 2021
- University of Maine
Successful conservation planning is reliant on accurate information about species numbers and distributions and a workforce trained to use the best available science and technologies to gather those data. Our project’s overarching goal is to build interdisciplinary data literacy by combining emerging remote sensing and data analysis technologies into user-friendly tools for examining the real-world problem of population assessment of Maine’s difficult-to-survey nesting seabirds that are facing threats in their coastal habitats that are critical to their persistence. We use Gulf of Maine colonial nesting birds as our focal species to develop survey, image collection and processing, and data analysis methods and technologies that are transferable to other taxa and survey goals. We are partnering with the Maine Department of Inland Fisheries and Wildlife and the U.S. Fish and Wildlife Service to compare nesting seabird data collected during plane-, Unmanned Aerial Systems (UAS)-, and ground-based surveys to examine effects of platform, sensors (type, spatial, temporal resolution), timing, and deployment approaches on species’ detectability, counts, and behaviors. Our surveys will generate large datasets that traditionally require significant time for interpretation by observers. Automation of this interpretation and counting process potentially can increase precision, accuracy, and affordability of these surveys; therefore, we will use artificial intelligence and machine learning to develop detection algorithms to process the imagery we collect as well as selected archived imagery. Finally, we will create toolkits that include instructions for best practices of combining UASs, traditional survey methods, and automated analysis for conducting population assessments and a user interface that can be accessed by professionals and citizen scientists in applications requiring coastal wildlife monitoring.