Measuring Desert Terrain and Ecosystem Attributes Using Airborne Lidar and Imagery.

The desert southwestern United States hosts habitats for several threatened and endangered species, one of which is the desert tortoise (Gopherus agassizii). These reptiles are found throughout the Mojave and Sonoran Deserts, generally below an elevation of ~1,100 meters. Remote sensing technologies could be used to rapidly predict potential habitats for study or protection, which would be especially useful in the large and remote land areas inhabited by these tortoises and the substantial variability of attributes that are favorable for habitats (shrub canopies of specific volume and area, presence of ledges, etc.). The goal in this study was to develop a fine-scale model that predicts potential habitat locations of G. agassizii in the Boulder City (Nevada) Conversation Easement area (35,500 hectares) and surrounding landscapes. This was done by analyzing airborne Lidar data and color imagery and determining vegetative cover, percent shade cover, shrub height and area, species richness of perennial and ephemeral plant species, NDVI, and the presence/absence of (topographic) ledges greater than 50 degrees. We then overlaid these attributes atop digital soil maps and known locations of tortoise burrows, creating a predictive model of potential tortoise habitats. Arizo soils (sandy-skeletal, mixed, thermic Typic Torriorthents) and creosote bush (Larrea tridentada) and white bursage (Ambrosia dumosa) dominated the soil and vegetation, respectively. Field measurements of ~250 shrubs were taken in different geomorphic terrains in Eldorado Valley to confirm the study estimates. Results showed that the final point cloud (5-7 points/m²) used to create a 0.25 m DEM and an imagery pixel resolution of 0.15 m were sufficient to estimate larger shrub islands and ledges pinned by petrocalcic horizons that are suitable for tortoise habitats. More than 13 million shrubs were revealed over the study area, which were then binned by shrub density, shrub height, and NDVI for subsequent statistics. Overall, creating digital representations of desert ecosystems to supplement digital soil models provided significant baseline data for habitat assessment.