Résumé
To simulate rice growth and yield accurately, models must integrate growth and morphogenetic processes. Widely used models such as ORYZA1 give much consideration to environmental effects on growth (e.g. carbon assimilation) but assume that assimilate partitioning among organs is independent of the environment, thereby allowing for no or little phenotypic plasticity. The present study evaluated the effects of nitrogen (N) nutrition on dry matter and N partitioning, using data from previous studies in the Philippines on irrigated transplanted and direct-seeded IR64 rice fertilized with six levels of N in one, and two levels in another study. Nitrogen application reduced dry matter partitioning to roots, particularly in direct-seeded rice. High N resources significantly increased dry matter partitioning to leaf blades at the expense of stems, but did not affect partitioning between panicles and the rest of the plant. Partitioning of dry matter to leaves decreased as the N concentration in the leaves (LNC) decreased, regardless of the cause of low LNC: low N rate, high population in direct-seeded rice, or phenological stage of the crop. Leaf partitioning of absorbed N, compared to dry matter, was high and varied little during early vegetative growth, but varied strongly from panicle initiation onwards, probably due to competition for N between leaves and the stem and the developing panicle. The possible structure of models that would simulate plasticity of assimilate and N partitioning is discussed on the basis of organ-specific phenological time axes and interacting supply and demand functions. The underlying assumption is that the observed effects of N nutrition on partitioning are mediated by the size of incremental assimilate pools.