Measuring erosion in long-term tillage plots using ground-based lidar

Abstract

Erosion remains a serious problem for agricultural soils throughout the world. Tillage significantly affects a soil's susceptibility to erosion. Erosion research is usually conducted in situ by capturing eroded sediment in brief, natural or artificial rainfall events. Methods for measuring long-term erosion are needed to better understand long-term effects of soil management. Landscape change resulting from erosion may be accurately characterized using ground-based lidar. Ground-based lidar data were collected in 2010 at a long-term (28-yr) trial of nine tillage treatments in the North Carolina Piedmont. Tillage effects on plot-surface elevations were examined after removing large-scale variation in elevation (slope) by detrending with first- through fourth-order polynomials. Residuals represented the elevation difference from the trend for each location. Mean plot elevations were calculated for datasets from each detrending model and used to assess erosion. In the subsequent elevation analysis, data derived from the second-order polynomial had the highest R2, attributing 66% of the variation in elevation to block and treatment. Treatment elevations relative to no-till (NT) ranged from +3.20cm in the fall chisel (CHfa) plots to −13.28cm in the fall moldboard plow plus disk treatment. Weeds in lesser-tilled treatments such as CHfa and no-till plus in-row subsoiling resulted in artificially high elevation measurements. In general, the most intensely-tilled treatments had the lowest elevations and the least-tilled treatments had the highest. NT was used as the reference elevation for no change, and soil loss was calculated using these data along with field-collected estimates of bulk density. The relative elevation differences corresponded to a maximum soil loss of 1891Mgha−1, which corresponds to an average annual soil loss of 67.5Mgha−1yr−1. Soil loss estimates were similar to others estimated from soil profile truncation. This research indicates that ground-based lidar data can be used to estimate soil elevation changes and thus soil loss due to tillage-induced erosion.