Cumulative germination percent was calculated for each seed row after 9 d. Results from the experimental trials varied significantly and, consequently, different quadratic and cubic regression models were fit to each experimental trial. Critical temperatures were calculated from the different models.Germination rate was calculated as the inverse of the number of days (d−1) to reach 50% germination for each row of 10 seeds. If the seeds for a given treatment never reached 50% germination, the rate was entered as 0.0 d−1. Germination rate was separated into sub-and supra-optimal ranges by visual inspection of the data. Linear regression models were fit to the sub and supra-optimal data sets.The intersection of the sub- and supra-optimal regressions with in each experimental trial denoted the To for germination rate. Both Tb and Tc were determined from the intercept of the sub- and supra-optimal regressions, respectively.