Tempo de germinação de sementes dormentes por análise de sobrevivência

Abstract

Dormancy ensures that seeds germination occurs at the right time and place. In statistical analyzes on germination, time is usually estimated by measurements (initial, medium and median). However, this method does not consider seeds that died in the time interval or those that remained at the end of the experiment. Thus, the objectives of this research refer to apply survival models to determine the germination period of dormant seeds of four forest species of Fabaceae family, considering dead and unembedded seeds in the models; to compare these times with those obtained by moments; and to quantify the risk of seed death due to dormancy overcoming methods. As they present distinct dormancy pattern, the species Bowdichia virgilioides, Mimosa caesapiniifolia, Parkia pendula and Schizolobium parahyba were selected to compose samples from different origins and years to guarantee quality. The germination experiments were conducted in a completely randomized delineation with four repetitions in plots consisting of 25 seeds in a factorial scheme, with variable structure depending on the number of samples per species and the number of methods to overcome dormancy. The germination probabilities for Bowdichia virgilioides, Mimosa caesapiniifolia, Parkia pendula and Schizolobium parahyba by survival are equal to or greater than the accumulated germination percentages. Moreover, the dormant seeds germination periods of B. virgilioides, M. caesapiniifolia, P. pendula and S. parahyba are underestimated when they are obtained by the method of moments, as they do not consider seeds that died or did not germinate until the end of the experiment. The only inconvenience of Survival Analysis is that it does not distinguish death or hard seed censorship at the end of the research. From the physiological point of view, dead seeds have low quality, and hard ones still have high potential of germination. The risk of death of P. pendula and S. parahyba seeds is high when submitted to dormancy overcoming methods, especially emerging. On the other hand, sulfuric acid and hot water were less lethal to B. virgilioides and M. caesalpiniifolia seeds.