MODELING CAULIFLOWER DEVELOPMENT FROM TRANSPLANTING TO CURD INITIATION

A model of the development of cauliflower during the juvenile phase and the curd induction phase is set up and described. The model is based on plant growth and climate data from our own experiments and data from two previously published Dutch experiments. The duration of the juvenile phase is described by a temperature sum starting at transplanting and using a base temperature of 0-degrees-C. The temperature sum requirements (thermal time in Kelvin days) of six cultivars in our experiments ranged from 26 to 83 K d compared with about 250 K d of two other cultivars in the Dutch experiments. Possible reasons for the difference are discussed. Leaf initiation was found to be well described by temperatures above 0-degrees-C. The end of the juvenile phase can thus also be expressed as a certain number of leaves initiated. This is of practical value because leaves can easily be counted. A good description of leaf initiation data (R2 values of 0.9) was achieved when a gradual increase in the leaf initiation rate with increasing leaf number and increasing temperature was used. The end of the juvenile phase was estimated to occur on average when the plant had 12 leaves initiated in our own experiments compared with 17 to 19 leaves initiated in the Dutch experiments. The duration of the curd induction phase in the model is described by using linear responses to temperature that are symmetrical below and above an optimum temperature. A common optimum temperature for curd induction was estimated to be 12.8-degrees-C for two cultivars in the Dutch experiment, the base temperature was estimated to be 0-degrees-C, and the maximum temperature, therefore, is taken to be 25.6-degrees-C. The best fits of data in the combined model of juvenile and curd induction phase show R2 values from 0.4 to 0.6. The model needs refinement and validation especially in modelling the duration of the curd induction phase for different cultivars. Still, the descriptive development model of cauliflower divided into distinct juvenile and curd induction phases gave a better fit than a model with no phase separation.