|Camp Jr, Carl|
Submitted to: Soil & Tillage Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 14, 1997
Publication Date: N/A
Interpretive Summary: Understanding the relationship between soil strength and soil water content is important for everything from road construction to lawn care. In the Coastal Plain of the southeastern US, hard, dry soils retard plant root growth and reduce crop yields. Generally, soils are softer when wet and harder when dry. As a result, natural variation of water content across a field can mask differences in soil strength. If strength values can be corrected to what they would be at a single water content, scientists can compare strength data collected from different parts of a field. We tested several equations, several mathematical relationships between soil strength and soil water content, that could be used to correct the soil strengths. Usually, corrections were dependent on experimental treatments. This was both good news and bad news. The good news was that a relationship could be found and used to correct soil strength for differences in water content. The bad news was that most corrections were dependent on the specific experimental treatment. In this case, different treatments from the same experiment would have different corrections. This could bias treatment effect. A method still needs to be developed that can calculate unbiased or uniform corrections among treatments.
Technical Abstract: Because soil strength varies with soil water content, field variation of water content can obscure comparisons of measured strengths. Correction of cone index data to a single water content would help interpret differences among treatments. We used equations from TableCurve software and the literature to correct cone indices for differences in soil water contents. Data were taken from two field experiments where cotton was grown using conventional and conservation tillage and beans were grown using microirrigation. Equations fit the data with coefficients of determination ranging from 0.55 to 0.92 and error mean squares from 1.37 to 6.35. Data fit the equations best when separate parameters were developed for individual treatments. Before correction, cone indices varied significantly with water content in analysis of variance. After correction, water content error mean squares and F values were reduced in the ANOVA. Corrections using more than one equation based on treatment differences reduced significance of water content on cone index more than a single correction equation. However, differences after multiple-equation correction may be a manifestation of the different corrections for the individual treatments. Finding a common correction that is valid and consistent across treatments may depend on coordinating a set of equations.