Reliability of early and final counts in cold and cool germination tests for predicting maize seed vigour

Cold test (CT) is widely used in maize for predicting field performance in early sowing, even though it is rather labour intensive and requires almost two weeks to complete. The objective of this work was to examine alternative procedures to CT suitable for routine testing. Standard germination (SG), CT and cool germination (CG) were carried out on 24 maize lots chosen with high (> 85%) SG. For the three tests, besides the final count of normal seedlings, early counts of germinated seeds were made. Lots were evaluated in two field trials (six replicates each) sowed at a very early and early date. For all the laboratory traits, the differences among lots were significant and attained higher values in the early counts compared to the final ones. Moreover, the correlations were high between early counts performed during the three tests but were in general lower between early and final counts. Early counts showed lower correlations with field performance than the final ones, except for the count at 144 h during CG that attained a correlation level with field emergence (r = 0.81) similar to the final counts of CG and CT. Nonetheless, the evaluation of seedling vigour during the final count of CT showed the highest predictive value of field performance. All the laboratory measures were in close agreement with field performance when the latter was evaluated under less unfavourable conditions.