Submitted to: Environmental Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: February 23, 2012
Publication Date: September 12, 2012
Citation: Phillips, B.L., Ngugi, M., Hendrickson, J.R., Smith, A., West, M.S. 2012. Mixed-grass prairie canopy structure and spectral reflectance vary with topographic position. Environmental Management. 50:914-928. DOI 10.1007/s00267-012-9931-5. Interpretive Summary: Vegetation canopy characteristics at summits are lower than at toeslopes, with more short-grass and less tall-grass species, lower amounts of standing crop biomass, lower leaf area, canopy nitrogen, and canopy water contents. We hypothesized spectral reflectance collected from Landsat and ASTER satellite image data would also be significantly different at summits. If so, then spectral reflectance information collected at summits may be used to indicate low structure. Application of topographic position information (summit, midslope or toeslope) and imagery across a 36,000-ha landscape could be used to target toeslope areas with reflectance’s similar to summits. We found the Normalized Difference Vegetation Index (NDVI) and the simple ratio of Landsat Band 7/Band 1 (SR71) were useful in discriminating summit vegetation from surrounding vegetation. Both indices were significantly different at summits, and this was found at 72 field sites and over the entire landscape. Results spatially highlighted toeslopes with low structure. This approach would facilitate low-cost, proactive management and comprehensive ecosystem assessment for large, mixed-grass prairie landscapes during peak season.
Technical Abstract: Managers of the nearly 0.5 million ha of public lands in North and South Dakota rely heavily on manual measurements of canopy height in autumn to ensure conservation of grassland structure for wildlife and forage for livestock. However, more comprehensive assessment tools could be developed for these mixed-grass prairie ecosystems using imagery collected during the growing season. We approached this problem by first determining if vegetation attributes measured in the field (canopy height, nitrogen and water contents, standing crop biomass, and leaf area) and satellite-borne spectral data were different at summits as compared to mid and toeslope positions. We used a 36,000-ha landscape-of-interest (LOI), where three topographic positions (summit, midslope, toeslope) were geo-located within 24 large (1-km) randomly selected plots. We tested for effects of topographic position on vegetation and spectral reflectance at field plot and landscape scales. We found summit vegetation was characteristically low in structure, with lower leaf area, standing crop biomass, canopy height, nitrogen, and water contents than at toeslopes. We also found spectral reflectance was significantly different at summits. At field and landscape scales, the simple ratio of Landsat Band 7/Band 1 (SR71) was greater at summits than at toeslopes. Low vegetation structure at summits correlated with higher SR71 and was used to assess structure at other positions in the landscape. Results suggest application of topographic and Landsat spectral data facilitate a proactive management approach for comprehensive ecosystem assessment aimed at geo-locating low structure during peak season.