Evaluation of Selected Nonlinear Regression Models in Quantifying Seedling Emergence Rate of Spring Wheat

Abstract

Fast and uniform seedling emergence increases yield potential of spring wheat (Triticum aestivum L.) in short-season areas. An accurate method of quantifying rate of seedling emergence is needed. In this study, we compared the relative effectiveness of the Gompertz, Logistic, and Weibull models in quantifying emergence rate of spring wheat. ‘Roblin’ wheat was grown in a growth room under five soil water potential: − 0.002, − 0.165, − 0.41, − 1.00, and − 1.45 MPa. Daily-recorded emergence data were fitted to each of the models. The analyses of stability and accuracy functions, residual sum of squares, and variance showed that the Weibull model was not appropriate in quantifying rate of emergence.The Gompertaz and Logistic models functioned in a similar way with great stability and accuracy in most cases. The Gompertz predictions most closely fitted the observed set of responses with residual points scattered around zero. For lognormally distributed emergence patterns common under field conditions, the Gompertz model provided the most appropriate characterization of emergence.