|
|
|
| Authors: | D. Agostini, R. Habib |
| Keywords: | Actinidia deliciosa, temperature, phenology, model, variability, probability |
Abstract:
A biological based phenology model is described for simulation of flowering in kiwifruit.
The model, with temperature as the principal parameter, describes the distribution of a population of buds, and then of flowers, at different phenological stages, considering the date and delay of flowering as the result of successive random processes.
The model assumes that the development of a flower bud is a stochastic process consisting of accumulated small increments of developmental time, expressed in degree-hours.
The process S(t) is defined as the amount of developmental time a flower bud has accumulated by actual time t. As the amount of development S(t) increases, the flower bud passes through successive stages, with the stages delimited by thresholds.
The heart of the model is a probability distribution for S(t) which is approximated by a logistic distribution.
This model integrates also quantitative elementary components, the estimation of which allows the simulation of the number of flowers.
The model parameters have been estimated for the female variety 'Hayward' and have been fitted to a 1991 data set.
Validation has been done on 1990 and 1992 data sets.
Under the climatic conditions of Corsica, this model seems to give a good representation of the variability of delay and intensity of flowering
|
Download Adobe Acrobat Reader (free software to read PDF files) |
|