SEED-GERMINATION MODELS

Several authors have developed mathematical models using a parametric approach to the problem of estimating germination in various seeds. Instead, we have used a non-parametric approach which allows us to gain an insight into the actual behaviour without imposing any preconceived probability distribution on the growth. We develop a model for the discrete and continuous cases and then estimate the probability of germination in unit time for several observed data sets. The results show that no simple assumptions can be made in modelling these situations.