|JANSEN, VINCENT - University Of Idaho
|KOLDEN, CRYSTAL - University Of Idaho
|TAYLOR, ROBERT - Nature Conservancy
Submitted to: International Journal of Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/28/2015
Publication Date: 12/18/2015
Citation: Jansen, V., Kolden, C.A., Taylor, R.V., Newingham, B.A. 2015. Quantifying livestock effects on bunchgrass vegetation with Landsat ETM+ data across a single growing season. International Journal of Remote Sensing. 37(1):150-175. https://doi.org/10.1080/01431161.2015.1117681.
Interpretive Summary: Grasslands provide important ecosystem services, including forage for the livestock industry and vital habitat for native wildlife. Rangeland managers need timely and accurate landscape scale estimates of vegetation amounts to determine the effects of their land management decisions. Prior remote sensing studies of grazing metrics have largely sought to estimate vegetation amounts during peak greenness only. We tested a range of spectral indices in an Oregon grassland to identify the best spectral predictors of grazing impacts and their associated field measurements (vegetation cover, biomass, and vertical structure) across the entire growing season. Here we show that the remotely-sensed measures of vegetation are sensitive to the varying amounts of grazing across the study period. We also demonstrate that models built only on sampling during peak greenness are not applicable across the entire grazing season, and that models should change across the season to capture senescence. This research provides insight into the feasibility of using multiple, within season, remotely sensed data to produce grazing management decision support tools.
Technical Abstract: Grassland systems provide important habitat for native biodiversity and forage for livestock, with livestock grazing playing an important role influencing sustainable ecosystem function. Traditional field techniques to monitor the effects of grazing on vegetation are costly and limited to small spatial scales. Remote sensing has the potential to provide quantitative and repeatable monitoring data across large spatial and temporal scales for more informed grazing management. To investigate the ability of vegetation metrics derived from remotely sensed imagery to detect the effect of cattle grazing on bunchgrass grassland vegetation across a growing season, we sampled 32 sites across four prescribed stocking rates on a section of Pacific Northwest bunchgrass prairie in northeastern Oregon. We collected vegetation data on vertical structure, biomass, and cover at three different time periods: June, August, and October 2012 to understand the potential to measure vegetation at different phenological stages across a growing season. We acquired remotely sensed Landsat Enhanced Thematic Mapper Plus (ETM+) data closest in date to three field sampling bouts. We correlated the field vegetation metrics to Landsat spectral bands, 14 commonly used vegetation indices, and the tasselled cap wetness, brightness, and greenness transformations. To increase the explanatory value of the satellite-derived data, full, stepwise, and best-subset multiple regression models were fit to each of the vegetation metrics at the three different times of the year. Predicted vegetation metrics were then mapped across the study area. Field-based results indicated that as the stocking rate increased, the mean vegetation amounts of vertical structure, cover, and biomass decreased. The multiple regression models using common vegetation indices had the ability to discern different levels of grazing across the study area, but different spectral indices proved to be the best predictors of vegetation metrics for differing phenological windows. Field measures of vegetation cover yielded the highest correlations to remotely sensed data across all sampling periods. Our results from this analysis can be used to improve grassland monitoring by providing multiple measures of vegetation amounts across a growing season that better align with land management decision making.