Evaluation of linear and nonlinear regression models to describe response of emergence to temperature in lentil (Lens culinaris Medik. )

This study was done to evaluate different linear and nonlinear regression models to determination of cardinal temperatures and biological day's requirement for emergence of lentil. Therefore, a split plot experiment was conducted based on three replications. Seeds of lentil (Gacgsaran, Kimia and Bilehsovar) were sown in field at 12 different dates. Beta, dent-like and segmented models were applied to evaluate the relationship between germination rate and temperature. Root mean square deviation (RMSD), coefficient of determination (R2), variation coefficient (CV (and linear regression coefficients (a and b) were used to select the perfect model. Results of models fitting indicated that the response of lentil emergence to temperature is best described by a segmented model. Cardinal temperatures estimated by this model were -1.27 to -1.62°C for base temperature, 23.15 to 25.92°C for optimum temperature and 30°C for ceiling temperature based on air temperature. There was not any significant difference among cultivars in view of in base temperature and optimum temperature but cultivars had significant difference in biological days. The biological day’s requirement was 6.99, 8.56 and 8.78 for Bilehsovar, Gacgsaran, and Kimia, respectively. The quantitative information provided in the present study can be used to predict the emergence of lentil cultivars.The main applicable result of this study was that reaction of seedling of the lentil is describable best by the use of segmented model, so this model and its derived parameters are applicable in predicting emergence in some given lentil cultivars.