|O'DONNELL, THOMAS - University Of Missouri
|GOYNE, KEITH - University Of Missouri
|MILES, RANDALL - University Of Missouri
|ANDERSON, STEPHEN - University Of Missouri
|Sudduth, Kenneth - Ken
Submitted to: Geoderma
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/8/2010
Publication Date: 3/15/2011
Citation: O'Donnell, T.K., Goyne, K.W., Miles, R.J., Baffaut, C., Anderson, S.H., Sudduth, K.A. 2011. Determination of representative elementary areas for soil redoximorphic features by digital image processing. Geoderma. 161:138-146.
Interpretive Summary: Soil redoximorphic features (SRFs) are patches of different color soil material found in soil horizons. They are formed by chemical reactions associated with alternating periods of aeration and water saturation. They provide scientists and land managers with information about past and present soil moisture regimes and are thus useful in land management decisions. The standardization of SRF quantification requires an objective determination of the representative elementary area (REA), the minimum area needed so that measurements are independent of the sample size and account for spatial variations. The objectives of this study were to define, determine, and compare this minimum area for features of low and high color intensity, present in claypan soils of northeastern Missouri, USA, at different soil depths, and for different landscape positions. Results showed that large soil cores (greater than 8 cm [3.2”] diameter) are necessary to accurately quantify features of high color intensity. For low color intensity features, a greater than or less than 5 cm [2”] diameter core is recommended to reach an REA, allowing accurate quantification for soil classification purposes and hydric soil determinations. Federal agencies requiring quantifiable measures of soil pedofeatures for land management decisions may greatly benefit from determining these minimum measurement scales and ensuring appropriate data collection methods.
Technical Abstract: Photography has been a welcome tool in documenting and conveying qualitative soil information. When coupled with image analysis software, the usefulness of digital cameras can be increased to advance the field of micropedology. The determination of a Representative Elementary Area (REA) still remains a critical information need for soil scientists so that field measurements are independent of sample size and account for spatial heterogeneity. An objective of this study was to define and determine an REA for Low Chroma and High Chroma soil redoximorphic features (SRFs) present in claypan soils of northeastern Missouri, USA using a digital camera and image classification techniques. An additional objective was to examine REA differences between these two SRF types, soil depths, and landscape positions to highlight sampling considerations when quantifying SRFs in the field. Three metrics were chosen to quantify SRF heterogeneity, including percent occurrence, mean Euclidean distance, and the Interspersion/Juxtaposition Index. The relative change in these metrics was determined for 16 image sizes ranging from 2.5 cm2 to 40 cm2 and used to identify an REA. Results showed REAs (mean plus or minus SE) for Low Chroma (17.7 cm2 plus or minus 0.4) and High Chroma (25.4 cm2 plus or minus 0.7) were significantly different (Alpha=0.05). Further review of REAs indicated large sampling diameters (greater than 8 cm) are necessary to simultaneously capture REAs of Low Chroma and High Chroma SRFs. When SRFs were considered separately, a greater than or less than 5 cm diameter core is recommended to reach an REA for Low Chroma, allowing accurate quantification for soil classification purposes and hydric soil determinations. Federal agencies requiring quantifiable measures of soil pedofeatures for land management decisions may greatly benefit from determining these minimum measurement scales, ensuring appropriate data collection methods in the future.