Rangeland Inventory and Monitoring With Unmanned Aerial System Imagery

Abstract

Publicly managed rangelands today are seeing higher demand from society for the goods and services they can provide, including livestock production, wildlife habitat, myriad forms of recreation, and ecosystem services. Adaptively managed multiple-use lands could benefit from more objective and synoptic data to evaluate ecosystem function and to carry out and defend land health assessments that allow or exclude certain land use activities. Field methods to measure critical soil and vegetation indicators are well-established and becoming standardized across jurisdictions. However, field methods have two main limitations: 1) most can only observe small portions of the landscape, which may produce an incomplete picture of the status and trend of rangeland health; and 2) field methods cannot measure some indicators very well or not at all. This research focused on developing methods to measure soil and vegetation characteristics from unmanned aerial system (commonly known as drones) imagery, which can observe significantly more land than their field counterparts. I demonstrated the measurement of one soil (erosion/deposition) and four vegetation (forage utilization, fractional cover, vegetation height, canopy gaps) indicators using drone imagery and compared each with established field methods. The results show that drone imagery methods can serve as a complement to field methods or even a replacement in some cases. I found that drone imagery methods can precisely map topographic change and forage utilization across extents not previously possible. Imagery methods can outperform field methods for vegetation heights and canopy gaps in some vegetation communities. Drone-imagery indicators have matured to the point where they can start being integrated into adaptive land management. An online space dedicated to sharing imagery workflows amongst the range community could quicken the pace of identifying best practices to facilitate the transition toward this technology. Adopting drone-based inventory and monitoring data, however, will not replace field skills in plant identification, knowledge of vegetation phenology and succession, and logical interpretation of the data for land health assessments.