Back to search

Michigan Technological University

DETECTION AND CLASSIFICATION OF EURASIAN WATERMILFOIL WITH MULTISPECTRAL DRONE-ENABLED SENSING

Abstract

dc:description.abstract

Littoral zones support growth of submerged aquatic vegetation, creating productive areas that provide food and habitat for fish, amphibians, macroinvertebrates, and other parts of the food web. Understanding macrophyte dynamics requires the identification of submerged aquatic vegetation (SAV) taxa, which can be possible if taxa of interest are spectrally distinct with data collected at appropriate scales. Eurasian watermilfoil, Myriophyllum spicatum or EWM, is a non-native SAV species that forms thick, often monotypic beds that reduce benthic species richness, restrict recreation, reduce property values, clog water intakes, and lower dissolved oxygen concentrations. Remote sensing of SAV species has to address the presence of lake color constituents that reduce lake clarity, making identification of species of interest more challenging. To address this challenge, I first investigated how to collect spectral data of SAV from boatside and drone platforms to determine the number and types of bands needed to identify EWM. Hyperspectral numbers of bands such as 65 10-nm wide bands between 350 and 1000nm reliably identified EWM, while use of a modified normalized difference vegetation (NDVI) index provided significant differences among SAV vs. other dominant aquatic vegetation groups. We demonstrated this for classifications at five sites over three years in the littoral areas of the Les Cheneaux Islands in northwestern Lake Huron, Michigan, USA, with 78.7% average producer’s accuracy and 76.7% average user’s accuracy, higher than most previous efforts at remote sensing of SAV. Finally, we applied these mapping capabilities to two areas in the Les Cheneaux Islands and one area in the Keweenaw Peninsula in Michigan’s northwestern Upper Peninsula that received treatments to reduce EWM. One site underwent mechanical harvesting, a second had a native fungus applied as a method of biological control, and a third site had diver-assisted suction harvesting completed. Classifications before and after treatment showed that it was possible to quantify the reductions of 63-89% in EWM extent due to these efforts. These results help demonstrate that UAS-enabled multispectral sensing can produce useful quantitative data on the presence and extent of SAV taxa of interest, providing a tool for monitoring treatment effects and improving understanding of aquatic ecology.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy in Biological Sciences (PhD)
Level thesis:degree_level
Open Access Dissertation
Discipline thesis:degree_discipline
Department of Biological Sciences

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Brooks, Colin
Committee members dc:contributor.committeemember
  • Casey Huckins
  • Evan Kane
  • Robert Shuchman
  • Martin Auer
Contributors dc:contributor
  • Amy Marcarelli

Subjects

dc:subject × 7

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
Michigan Technological University
Base URL
digitalcommons.mtu.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Brooks, Colin. DETECTION AND CLASSIFICATION OF EURASIAN WATERMILFOIL WITH MULTISPECTRAL DRONE-ENABLED SENSING. Open Access Dissertation thesis, https://doi.org/10.37099/mtu.dc.etdr/1038