|Johhson, Cinthia - UNIV. NEBRASKA|
|Eskridge, K - UNIV. NEBRASKA|
|Peterson, G - COLORADO STATE UNIV.|
Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: September 1, 2003
Publication Date: November 2, 2003
Citation: Johhson, C.K., Eskridge, K.M., Weinhold, A.P., Doran, J.W., Peterson, G.A., Buchleiter, G.W., Corwin, D.L. 2003. Designing field-scale experiments using apparent soil electrical conductivity. Soil Science Society of America. Paper No. S08-johnson626997-P. Technical Abstract: Classical experimental design and statistical analysis relies on replication and blocking, requisites impractical for on-farm field-scale research. Alternative approaches are needed to estimate experimental error. When apparent soil electrical conductivity (ECa) is correlated with crop yield, ECa classification (partitioning a field into ranges of ECa) distinguishes yield differences in the absence of treatments, an ideal basis for blocking plot-scale experiments. We hypothesized that, conversely, ECa-classified within-field blocking can be used to evaluate field-scale experiments by approximating plot-scale experimental error. To test this, two disparate field-scale sites were evaluated, a dryland rotational cropping experiment in semiarid northeast Colorado, and an irrigated rotational forage experiment in arid central California. Unsupervised and equal-size response surface ECa classification methods, respectively, were applied as bases for soil sampling. Mean square errors (MSEs) were calculated for multiple soil properties assessed at each experimental site and at nearby randomized complete block plot-scale experiments. Comparison of within-ECa-class and within-block MSEs revealed that most were not significantly different between the two scales. These findings support the use of within-field ECa-classified variance as a surrogate for plot-scale experimental error and a basis for roughly evaluating treatment differences in non-replicated field experiments.