Aspectos germinativos de 25 espécies florestais brasileiras: eficiência do método, anormalidades de plântulas e mortalidade de sementes

Abstract

CHAPTER II: The Rules for Seed Analysis determine mandatory methods for Analysis Laboratories Seeds accredited by the Ministry of Agriculture, Livestock and Supply. However, there are few validated methods available for forest species in this manual. For this reason, doubts about the conduct tests of these seeds are constant, especially, those related to the type of substrate, sterilization, temperature, pre-germination treatments and periods of tests closure. Starting from these considerations and based on literary records the goals pre-germination tests were run in order to allow the development of protocols for 25 Brazilian forest species. It is concluded that sodium hypochlorite is effective for disinfection of non-dormant and dormant seeds as estrellensis, C. speciosa and P. tomentosum. The filter paper in the form of rolls and a constant temperature of 25 º C ensure the germination of all species, except G. ulmifolia, which was used blotting paper in boxes germitest type and P. pendula at 30 º C . Scarification and clipping, are effective treatments for breaking dormancy, however, damage may occur to the seeds and seedlings depending on the size of the rubbed area and depth, as occurred in S. parahyba var. amazonicum. The scores ended up in second reading for most species, determining the starting time for the seeds with higher germination and the final time by lower germination potential. CHAPTER III: In the validation of methodologies for germination, the only variable compared between lots and labs are normal seedlings. However, the percentages of abnormal seedlings and dead seeds are essential for identification of sensitivity of the seeds to the methodology applied. Therefore, the aim of this work was to carry out the statistical analysis of the germination tests with validated methodologies, and analyze the impact of the transformation of the percentage of abnormal seedlings and dead seeds in lots. The data of abnormal seedlings and dead seeds from 25 Brazilian forest species were tested for normality and homogeneity, in the original and transformed scale, and analysis of variance. We conclude that heterogeneous variances for abnormal seedlings and dead seeds, both in the original scale, accounted for 48% and 52% of cases, respectively, and residues with asymmetric distribution, 28% and 24% of cases. Data transformation may prove detrimental to the analysis of forest species being certain that this practice may result in loss of the assumptions of normality and homogeneity. The percentages of abnormal seedlings and dead seeds show that the seeds have lots of sensitivity methods applied, there is influence of handling in the laboratory and fungal contamination.