Statistical evaluation of the germination of Cleome gynandra L. seeds

Abstract

Radicle protrusion and subsequent seedling establishment are two ways of defining seed germination. However, international rules for testing seeds adopted the first and final counts through consensus as defining stages of seed germination. In this study statistical guidelines have been used to determine these stages. Cauliflower and broccoli for which the first and final counts have been published in the international rules for testing seeds and Cleome gynandra L. for which there is no data on the first and final counts, were used as test crops to determine the statistical significances of the first and final counts. Broken-stick modelling and survival analyses showed the fourth day as appropriate for the first count. There was no evidence of further seed germination after the seventh day as shown by probability density and hazard rate. It is suggested that for C. gynandra, the first count should be performed on the fourth day and the final count on the seventh day of the germination test. This could be an objective and fast method of determining first and final counts in seed germination testing of new cultivars or wild species.