Testing a nonlinear model for simulating the time of seedling emergence of wheat.

Previous research found that a model that used a series of Beta functions (Beta model) was better than linear models for simulating the days from seeding to seedling emergence (DSE) of wheat (Triticum aestivum L.). To further validate the Beta model we used three sets of field studies conducted in North America: (1) a seeding date study at Moro, OR (38 seeding dates); (2) a multi-year, multi-site study in Canada (98 site-years); and (3) a series of experiments at Swift Current, SK (20 treatment-years). Results demonstrated that the linear emergence module in the Cropping System Model of The Decision Support System for Agrotechnology Transfer (DSSAT- CSM) tended to underestimate DSE. The Beta model using either daily or hourly air temperature markedly improved the simulation of DSE for all three studies. The Beta model using simulated soil temperature slightly improved simulation, but was not as good as that using air temperature, which was caused by the inaccuracy of soil temperature simulation. It seems that the Beta model is an appropriate model to predict seedling emergence of wheat grown in North America. There were no significant differences in DSE among wheat genotypes used in studies at Moro or Swift Current. [ABSTRACT FROM AUTHOR]