2020 ESA Annual Meeting (August 3 - 6)

PS 20 Abstract - Conservation of showy stickseed (Hackelia venusta) using drones to monitor plant populations and survey for potential habitat

Andrew Foster, Environmental & Forest Sciences, University of Washington, Seattle, WA, Wendy Gibble, UW Botanic Gardens Center for Urban Horticulture, University of Washington and L. Monika Moskal, School of Environmental and Forest Sciences, University of Washington, Seattle, WA
Background/Question/Methods

The rare endemic plant Showy Stickseed (Hackelia venusta) is one of Washington’s most rare plants, restricted to very steep, unstable, coarse soils, with difficult access for regular monitoring. The broad focus of this study is to determine if Unmanned Aerial Systems (UAS), known as drones, and remote sensing techniques can be used as a low-cost accessible tool for land managers to detect and survey plant populations - rare, invasive, or of other interest – as well as survey potential habitat for conservation efforts. In order to address the first question, monitoring plant populations via UAS, spectral signatures were derived for Showy Stickseed, other plants and natural objects within the habitat of the extant population. Spectral reflectance of plants was acquired in the field with an ASD FieldSpec Handheld spectroradiometer, as well as a MicaSense RedEdge multispectral sensor, both detect wavelengths in the visible and near infrared range. Spectral signatures were compared between the two sensors to identify key regions within the electromagnetic spectrum to guide aerial identification and determine if the RedEdge will be adequate for UAS monitoring. This analysis establishes a baseline spectral library for calibrating UAS monitoring of Showy Stickseed and supports future efforts of habitat modeling and prediction.

Results/Conclusions

Initial results show spectral signatures from flowering Hackelia venusta seedlings were significantly different from other field objects, supporting the approach to survey for plant populations using a drone system capable of collecting multispectral data, currently planned for Spring 2020. Monitoring these plants during peak flowering period will be important, as the showy white flowers stand out with high reflectance values. Comparisons between the RedEdge and the FieldSpec showed similar reflectance values in correlated bandwidths, although the difference in spectral resolution could lead to key regions in the spectral signature being lost by the RedEdge. A robust, repeatable, spectral field sampling protocol was developed for the FieldSpec and applied to the RedEdge. These results show that the FieldSpec or similar field spectroradiometer could be used as a tool to field calibrate UAS multispectral data. Due to the complex terrain and previously identified habitat requirements for Showy Stickseed, this multispectral data will help to contribute to locating suitable habitat for future outplantings. This work is the first step toward accurately monitoring plants via drone and predicting spatially explicit habitat, demonstrating a low-cost method for land managers, researchers and stakeholders to monitor plants remotely and efficiently, furthering conservation efforts.