Mapping spatially explicit vegetation gaps in Florida rosemary scrub using unmanned aerial vehicles

Abstract

Abstract Advances in remote sensing technologies offer new means to monitor habitats of importance on large scales. Florida rosemary scrub is one such threatened habitat, found in patches across the landscape in relatively elevated areas, and is often characterized by shrub‐less areas (gaps) among the dominant shrubs, which provide favorable microhabitats for many endemic and endangered plants and animals. However, gaps are difficult and time‐consuming to characterize, especially across large areas, using traditional ground‐based field methods. We developed and tested a method for rapidly classifying gaps using an unmanned aerial vehicle (UAV or drone). Aerial data were collected by a UAV‐mounted camera in April 2018, and stratified, random ground surveys to verify UAV data were conducted March through April 2018 at Archbold Biological Station in south‐central Florida, USA. We used mosaicked and georeferenced digital surface and terrain models to calculate vegetation height across 33 rosemary scrub sites (~230,000 m2 at 0.064 m2 pixel resolution). Gaps were defined as >1 m2 areas where vegetation height was <10 cm. We found that gap areas from UAV models and field surveys were significantly correlated across varying gap sizes, times‐since‐fire, and relative elevations. We also observed a significant decrease in mean gap area and percent gap space with increasing time‐since‐fire, a pattern consistent with smaller‐scale, ground‐based sampling, and a marginally significant increase in gap area with relative elevation. This remote sensing method lends itself to better exploration of how gap areas, their spatiotemporal patterns, and associated fire history, elevation, soil, and other geographic data affect structural vegetation dynamics across the landscape. This study illustrates the success of UAV modeling of gap space in Florida rosemary scrub, a result of regional consequence for the southeastern United States, but more broadly, it encourages the use of UAV technology as a tool to enhance traditional field‐based methods in systems globally. As habitat fragmentation and loss become increasingly problematic for the conservation of threatened habitats, understanding these complex spatial dynamics is crucial to the conservation and management of vegetation communities and their biodiversity.

Keywords

• aerial survey
• canopy height
• drone
• ecology
• elevation model
• habitat modeling