Ultra-High-Resolution Remote Sensing for Species Identification and Restoration Monitoring
Alicia Langton1*, Richard Alward2, Tamera Minnick3, Danielle Johnston4
1EcoloGIS, Grand Junction, CO, USA; alicia.langton@gmail.com
2Aridlands, LLC, Grand Junction, CO, USA; ralward@aridlands-nrc.com
3Colorado Mesa University, Grand Junction, CO, USA; tminnick@coloradomesa.edu
4Colorado Parks and Wildlife, Grand Junction, CO, USA; danielle.bilyeu@state.co.us
Next year initiates the United Nations Decade on Ecosystem Restoration. Monitoring of outcomes is a critical component of ecological and management-directed restoration. Both institutional (financial) and technical (poor design) challenges can limit effective implementation of monitoring programs. In this case study, we compared two methods that can be used for monitoring habitat restoration: ground-collected line-point intercept (LPI) data vs. remote sensing using a small Unmanned Aerial System (sUAS; aka, “drone”). Pinyon-juniper woodland overstory was thinned mechanically in Rio Blanco County, Colorado within twenty-one 0.8 ha plots to promote increasing cover of the shrubs, perennial grasses, and native annual forbs preferred by mule deer and other wildlife. Vegetation cover was estimated on the ground with 300 pts along 13 transects in each plot and by air with a fixed-wing sUAS outfitted with a 5-band multispectral sensor (ground sampling distance ≤ 6 cm). Ground-based data collection required 21 person-days in the field while multispectral image collection required just 2 person-days. We analyzed multispectral imagery using eCognition, an object-based image analysis software. We found strong correlations between LPI and remotely sensed cover estimates for shrubs (R²= 0.89) and two shrub species that are key indicators of habitat quality: serviceberry (Amelanchier alnifolia / A. utahensis; R²= 0.78) and snowberry (Symphoricarpos rotundifolius; R²= 0.88). Restoration monitoring using ultra-high-resolution sensors and sUAVs provides managers with more efficient, effective, and flexible tools. These technologies offer potential benefits of reduced costs and impacts to the environment while obtaining accurate assessments of wildlife habitat and restoration success.