Location: Soil Management ResearchTitle: Spatiotemporal dynamics of three-dimensional soil maps in a long-term field experiment
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 11/13/2019
Publication Date: N/A
Technical Abstract: A model framework was developed to statistically analyze spatiotemporal interactions between nutrient stocks and their direct and indirect relationships with biomass and grain yield, among other ecosystem services, in a long-term field experiment. Main factors (cropping systems, management options, crop rotations, crops, and physico-chemical properties of soil profiles of major and minor soil series) and their multi-way interactions were expressed in a series of sequential equation models using exploratory and confirmatory factor analyses. Model coefficients were used to build 3D maps of the soil profile at each of three soil depths (shallow, 0-15 cm; medium 15-30 cm; and deep 30-45 cm) for each soil series and at the start, middle and last phase of the field experiment. Each of the 3D maps was expressed as a 2D matrix and correlated with a matrix representing annualized biomass and grain yield. Results of this modeling procedure will be contrasted with results of classical statistical analyses using mixed models and time series procedure. A minimum data set was identified that can capture no less than 68.0% of spatiotemporal variation in soil physico-chemical properties and will be used to validate the model framework at a field scale.