Comparison of Sampling Methods for Estimating Pit and Mound Microtopography

Abstract

AbstractWe developed a method for making rapid estimates of pit and mound microtopography for characterizing research sites, soil map units, and ecological land types. We conducted this study in the drumlinized uplands of northeastern Wisconsin, where pit and mound microtopography is extensive. We established one 10 by 30 m grid and one 300‐m transect on each of four drumlins and identified a pit, a mound, or neither at nodes 1 m apart (300 observations per grid or transect). Pit and mound microtopography was estimated by calculating the proportion of pits, mounds, and points lacking evidence of pit and mound microtopography at nodes within each grid or transect. The optimal node spacing for grid and transect sampling methods was determined by repeatedly subsampling data derived from each method and identifying the lowest mean square error. The precision of grid sampling vs. transect sampling was determined by comparing the standard errors estimated with each method. Transects were more precise and time‐efficient than grids for estimating pit and mound microtopography percentages. Satisfactory results were obtained with 300‐m‐long transects with 3‐m node spacing. Transect methods appear applicable to the quantification of other types of microtopography.