Efficiency of the Experiments to Investigate Temperature Effects on Seed Germination

• Main Authors: Huei-You, Jheng, 鄭惠友
• Other Authors: Jiunn-Lung Lin
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/75992413536107203251

碩士 === 國立中興大學 === 農藝學系 === 92 === Many crop physiologists consider temperature the most important environmental factor influencing seed germination. There has been a vast amount of experiments to investigate the influence of temperature on seed germination of various species in the literature. This study attempted to evaluate the efficiency of the designs, particularly the treatment levels of temperature, taken by the researchers in these experiments. Firstly, we proposed a model to simulate the result from an experiment to investigate the temperature effects on seed germination of a seed lot. The model assumes that (1) a proportion (p) of the seeds are dead, (2) the relationship between germination rate and temperature is positive and negative linear at sub-optimal and supra-optimal temperatures respectively, characterized by a set of cardinal temperatures specific to the seed lot, i.e., the base temperature (Tb), optimal temperate (To), and ceiling temperature (Tc), (3) the thermal time requirement to germination of each alive seed, when appropriately transformed, is normally distributed, characterized by the mean thermal time to germination () and the standard deviation of the thermal time to germination (). The estimates of these parameters for soybean, lentil, faba bean, and konagi collected from literature were adopted as true values to create ten hypothetical models for the simulation experiments. From each of the ten hypothetical models, 5000 sets of data were generated to simulate each of the experiments with different designs of which the efficiency would be compared. The mean squared estimation error for each of the parameters was used to assess the efficiency of the experimental designs and the estimation procedures for the parameters. The results show that the efficiency did not change remarkably when the temperature levels were reduced even to a half of the original designs of the experiments reported in the literature.