Soil Variables and Interactions Affecting Prediction of Crop Yield Pattern

The pattern and magnitude of crop yield variability in a field reflect soil conditions that influence yield to varying degrees during critical portions of the crop season. Variables accounting for the majority of soybean [ Glycine max (L.) Merr.] yield variation, measured on 40 farm fields on either Cecil or Pacolet soil (clayey, kaolinitic, thermic Typic Hapludults) landscapes during 1982 and 1983, were previously identified by using factor analyses and components regression. The objective of this study was to determine the nature of interactions among clay at 0 to 0.1 m, C at 0 to 0.1 m, rainfall (Day 214 to 280), pH at 0 to 0.1 m and 0.45 to 0.55 m, and integrated soil water tension at 0.1 and 0.5 m. Analyses were done in relation to soybean grain yield using two‐point and four‐point, two‐dimensional sliding polynomials. As clay content at 0 to 0.1 m increased from 3 to about 18%, crop yield was greatly decreased, modulus of rupture was increased, and the yield response to pH, rainfall, and soil water tension was reduced. Soybean yield on soils with clay contents greater than 20% at 0 to 0.1 m was significantly modified by soil C content. At 39.2% clay, crop yield was affected very little by soil water tension at 0.1 m, by rainfall, or by pH at 0 to 0.1 m. More than 255 mm of rainfall between Day 214 and 280 increased yield only on sites with high C levels, e.g., 11.93 g kg −1 . In spite of significant variate interactions, linear and noninteractive 5 and 10‐term component regression models demonstrated good potential for predicting soybean grain yield.