Feasibility Study of Unmanned Aerial Vehicle Imagery for Mapping Roadside Milkweeds and Nectaring Plants for the Monarch Butterfly

Details

• Creators:
  Pringle, Elizabeth ; Dilts, Tom ; Xu, Hao ; Zamboni, Gino ; Chen, Zhihui ; King, Victor
• Corporate Creators:
  University of Nevada Reno. Board of Regents, Nevada System of Higher Education (NSHE)
• Corporate Contributors:
  Nevada. Department of Transportation ; United States. Department of Transportation. Federal Transit Administration ; United States. Department of Transportation. Federal Aviation Administration ; National Highway Safety Agency ; National Academies of Science (U.S.). Transportation Research Board. Information Services Committee (B0002)
• Subject/TRT Terms:
  Conservation Convolutional Neural Networks Data Analysis Drones Ecosystems Feasibility Analysis Insects Machine Learning Mapping Milkweed Monarch Butterfly Neural Networks Plant Identification Plants Roadside Flora And Fauna Unmanned Aircraft Systems Vegetation Video Imaging Detectors

  More +
• Publication/ Report Number:
  564-22-803
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States ; Nevada
• Edition:
  Final Report: 11/28/2022 to 8/31/2025
• Corporate Publisher:
  Nevada Department of Transportation
• Abstract:
  In this project, we sought to couple two imaging technologies, drones and vehicle-mounted cameras, with models using artificial intelligence to predict the location of milkweed plants in Northern Nevada. Milkweeds are weedy plants that frequently occur along roadsides, and they are the only food used by caterpillars of the monarch butterfly, whose populations have sharply declined in recent years. Conservation actions aimed at stabilizing populations of the monarch must thus identify the locations of milkweed plants. We conducted a pilot study using drones flown over three milkweed populations and determined that an artificial intelligence model aimed at identifying objects successfully identified individuals of two species of Western milkweed, although different methods were needed to identify flowering plants, whose nectar is needed by adult monarchs. This study also revealed that the abundance of narrowleaf milkweed predicts monarch presence. We conducted another study using images from the NDoT Roadviewer database and again successfully used object-based artificial intelligence to identify milkweed occurrence along Nevada roadways. It was, in general, easier to identify showy milkweed, which has a more distinctive architecture, than narrowleaf milkweed. Both studies revealed the importance of study timing relative to the life cycle of the milkweeds and monarchs.
• Format:
  PDF
• Funding:
  P564-22-803
• Collection(s):
  US Transportation Collection
• Main Document Checksum:
  urn:sha-512:c59f2a9e1ed9d54e0361f88afaed8f1748b5731622a43fdb6f428b5b1e216f14483d4ac2103f7fdaff17195da19e72d5c00a3e12ddcf83a11fb4cd2f3652fc1d
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/87798/dot_87798_DS1.pdf
• File Type:
  [PDF - 8.40 MB ]