Get reliable rangeland science

Remote sensing provides researchers the ability to study and assess landscape level changes in plant communities over broad spatial and temporal scales.� The use of high resolution imagery often improves capacity to capture patch level changes in structure and community transitions that may be lost at coarser scales. Ability to identify plant community differences and changes in phenology over a growing season can greatly aid in understanding how vegetation dynamics influence livestock grazing behavior.� Few studies have evaluated the use of high resolution satellite imagery to identify and map distinct plant communities within the Northern Great Plains, and track phenological changes associated with plant communities through time.� A study was conducted in north-central South Dakota to remotely sense five plant communities located both on and off prairie dog towns in the Northern Great Plains.� These included forb and grass dominated sites on-town, warm- and cool-season grass dominated sites off-town, and snowberry patches off-town.� During 2015 and 2016, Pleiades satellites were tasked to image the study site for a total of five monthly images each summer from June to October to coincide with livestock grazing at the site.� Imagery was converted into a normalized difference vegetation index (NDVI).� Training sites were mapped for each plant community of interest, and all spectral bands extracted to our training sites were used to construct a random forest algorithm to facilitate classification of plant communities.� Plant community phenology was tracked using NDVI time series values and analyzed for differences between plant communities, months, and years.� Results from this study will help researchers 1) determine optimal timing to collect satellite imagery based on plant community of interest, and 2) build thematic maps that can be overlain with livestock grazing behavior to better understand grazing selection.