Fine and Coarse Scale Sampling of Spatial Variability within a Switchgrass Field in Oklahoma

Abstract

Aim: The objective of this study was to describe the spatial patterns of selected soil properties and biomass yield at fine and coarse scale in a switchgrass field to determine the appropriate sampling approach to enable the calculation of means with minimum variance. Methodology: Spatial variability of biomass yield and soil properties at fine (2.5 m sampling interval) and coarse (10 m sampling interval) scales were assessed through semivariogram analysis. The site located in Chickasha, Oklahoma, consisted of two soil types a Dale silt loam (fine-silty, mixed, superactive, thermic Pachic Haplustolls) and McLain silty clay loam (fine, mixed, superactive, thermic Pachic Argiustolls). Eighty soil samples were collected along two 100 m transects at 2.5 and 10 m intervals established across each soil type in both 2012 and 2013. Results: The semivariograms revealed coarse scale organic carbon (OC) to be strongly correlated with range values from 56–78 m for both soils. Normalized difference vegetative index (NDVI) was consistently moderately correlated with a distance less than 30 m at the fine scale for both years. Switchgrass yield was strongly correlated at the fine scale for McLain silty clay for both years, while a weak spatial dependence over a range of 36 m in 2012 and a moderate dependence at 5 m in 2013 was observed for the Dale silt loam. Conversely, a reliable spatial dependence could not be identified for total nitrogen (TN). Conclusion: These results indicate that spatial correlation of coarse scale OC might have been imposed by the cropping system, while spatial correlation of switchgrass yield was influenced by the soil texture, particularly clay content. The use of the NDVI measurement was useful to describe the spatial dependence of switchgrass yield with good precision at the fine scale.