Description and model of the architecture of four genotypes of determinate autumn-sown white lupin (Lupinus albus L.) as influenced by location, sowing date and density

The structure of four new autumn-sown determinate genotypes of white lupin, grown with a combination of sowing dates and plant densities at four locations in France was studied in 1990/91 and 1991/92. Plant architecture was characterized in terms of the number of first-order branches produced and the distribution of leaves on these branches and how they varied with the number of mainstem leaves and the position of the branch on the mainstem. The variation in the number of mainstem leaves on determinate genotypes was satisfactorily described by an earlier model developed for indeterminate genotypes using the intensity of vernalization relative to accumulated temperature above 3 degrees C as input. However the architecture of the first-order branches, defined by their number and the total number of leaves per branch at each subtending position on the mainstem, differed, and was influenced by genotype, location and sowing date. In four determinate genotypes, these two characters were positively correlated, and correlated with the number of leaves on the mainstem. Increasing density from 10 to 40 plants m-2 reduced the number of branches and their total number of leaves. Pruning 15 leaves from the mainstem at the beginning of stem elongation did not influence the development of the first-order branches. There appeared to be a functional relationship between the number of leaves on the mainstem and the number of branches established very early in the growing season. The profiles of number of leaves on the first-order branches on the mainstem with branch position were modelled using two sub-models, one describing the mean number of leaves on the branches actually present, and the second describing the frequency of occurrence of a branch at a particular position on the mainstem. Of the five model parameters, one, representing the minimal number of branches present on the plant, was correlated with the number of first-order branches. Genotype, location, sowing date and density affected the shape of the profiles.