The wheat breeding program has generated an expressive number of cultivars which differ in sprouting rate of seed still attached to the spike. In order to assess these differences, an experiment was carried out with seeds harvested from two locations, Palotina and Cascavel, in Paraná State, Brazil, consisting of twelve cultivars in a randomized block design with four replications. Seeds were harvested at physiological maturity, in which one sample was submitted to a pre-chilling period to overcome seed dormancy, while the other was untreated (control). Seeds which did not germinate were submitted to the tetrazolium test to detect dormancy occurrence. Results showed that in both locations the germination percentage of 'Frontana' and 'IAPAR 53' seeds was extremely low, indicating the presence of a significant number of dormant seeds. The method to overcome dormancy from the Palotina samples was not efficient to result in a complete seed germination. On the other hand, pre-chilling of samples harvested in Cascavel resulted in complete seed germination. The dormancy phenomenon detected in both 'Frontana' and 'IAPAR 53' seeds seemed to be efficient in preventing germination of seeds still attached to the spike at physiological maturity.
Triticum aestivum; seed physiology; germination; tetrazolium
