Published online by Cambridge University Press: 01 September 1999
Pod and seed growth were studied in two experiments in which the plant's source-sink relationships were modified by (a) manually pruning an autumn-sown, indeterminate white lupin variety, Lunoble, to a determinate form, and (b) by growing a determinate variety, Lucyane, at densities ranging from 7 to 35 plants/m2. The pruning experiments indicated that the faster pod growth rate of determinate genotypes was not an inherent genetic trait but an indirect physiological consequence of the plant's changed architecture. In the density experiment, crop dry matter (DM) and nitrogen (N) were maximum at the end of pod extension in late July and similar across the plant density range at c. 12 t DM and 320 kg N/ha. Therefore, the amount of dry matter per plant decreased proportionately with the increase in plant number. The DM and N contents of the pod walls were also maximum at the end of pod extension, but seeds contained only a third of their final DM and a quarter of their final N. Protein accumulation during the final stages of seed growth, therefore, depended on the remobilization of nitrogen from other plant organs, primarily the leaves and pod walls. Nitrogen withdrawn from the leaves accounted for 44% of the gain in the pods, and N withdrawn from pod walls for 50–60% of the gain in the seed.
Seed number/m2 was the major yield component. Seeds and pods mainly aborted during early development, but seed number per pod was also decreased by some seed abortion after full pod extension, especially in first-order pods of plants grown at high density. The number of late-aborted seeds was negatively correlated with the amount of N remobilized from the pod wall. In determinate lupins, which have highly synchronous flowering and pod development, the large and sudden remobilization of nitrogen from leaves and pod walls for seed growth and protein accumulation triggered crop senescence.