The effect of modifying plant structure on the yield and maturity of the white lupin Lupinus albus

Experiments at Rothamsted in the UK and Lusignan in France examined the effect of artificially modifying plant structure on the yield and date of harvest of indeterminate autumn-sown cultivars of Lupinus albus. Experiments in 1989/90 were done on the cv. Lugel at Rothamsted and those in 1990/91 on the cvs Lugel and Lunoble in both the UK and France. At Rothamsted in 1990, when the summer was warm and dry, the indeterminate cv. Lugel ripened at the end of August and yielded 5.2 t grain/ha. At Lusignan in 1991, where summer weather was also warm and dry, the two unpruned indeterminate cultivars ripened in late July and yielded c. 4.9 t grain/ha. However, at Rothamsted in 1991, in a season which was predominantly cool and wet, the two indeterminate cultivars were not harvested until late October and yielded only 1.5 and 2.4 t/ha. Plants pruned to a semi-determinate form consisting of the mainstem plus the 1st-order branches yielded approximately the same as the unpruned indeterminates at Rothamsted and slightly less than the indeterminates at Lusignan. The semi-determinates ripened 10 days earlier than the indeterminates at Rothamsted in 1990 and Lusignan in 1991, and 6 weeks earlier in the UK in 1991. Epigonal plants with only a mainstem inflorescence, ripened even earlier than the semi-determinates but, without branch inflorescences, produced less yield and had poorer yield stability. Where disease was not a factor, as at Rothamsted in 1990 and Lusignan in 1991, epigonal plants ripened 3 weeks earlier than the indeterminates and yielded between 1.6 and 2.6 t/ha. When the pruned epigonal plants were badly infested with Pleiochaeta, as at Rothamsted in 1991, they yielded less than 0.4 t/ha. It is concluded that indeterminate genotypes of L. albus are well adapted to the climatic conditions of Lusignan but lack yield stability further north. For the cooler northern regions of Europe, an autumn-sown, florally-determinate genotype with one order of branching would produce adequate yield and have an acceptably early harvest date and yield stability. Such a genotype would have agricultural and economic value as a new protein break crop with a low requirement for nitrogen fertiliser.