Résumé
Specific leaf area (SLA), defined as the ratio between leaf blade area and its dry weight, determines the physiological cost of producing leaf area. Theoretically, at a given pattern of assimilate partitioning among plant organs, varieties having high SLA should produce a leaf area more rapidly than those having low SLA, or thick leaves. This is because the reduction in leaf photosynthesis caused by high SLA is more than offset by the gain in light interception during exponential growth. The same reasoning would call for higher relative growth rates (RGR) during exponential growth for varieties having high SLA. According to studies in 1986-90 at the International Rice Research Institute, RGR is correlated with tillering. One can therefore argue that SLA is physiologically linked with tillering ability. Studies at the West Africa Rice Development Association in 1995-98 indicated a positive correlation across genotypes between SLA and tillering ability, leaf area index (LAI), and weed competitiveness. The study used diverse genotypes, including traditional and improved Oryza sativa indica (lowland-adapted) and japonica (upland) types, as well as O. glaberrima landraces and interspecific crosses. A subset of these materials showed that varietal differences in SLA were already expressed before the plant became autotrophic and were carried through to the flowering stage. Recent studies at the Centre de coopération internationale en recherche agronomique pour le développement disaggregated the components of leaf area development, such as tiller and leaf appearance rate and genealogy, as well as leaf extension rates and final dimensions for an O. sativa indica (high tillering, small leaves), an O. sativa japonica (low tillering, large leaves), and an O. glaberrima cultivar (high tillering, intermediate leaves). For all cultivars, a common linear relationship was observed between relative tillering rate (RTR) and RGR. This relationship was quite robust across development stages, except the reproductive stage, and across different levels of temperature and drought. Differential tillering rates and leaf area development among varieties and environments were therefore related to dry matter acquisition, based on a common relationship for morphologically and phenologically different materials.We conclude that there is a strong case for SLA being responsible to a large extent for varietal differences in tillering ability and LAI in rice. Varietal differences in tillering ability may in fact be due to inherent growth ability. High SLA (thin leaves) is a major factor enabling early ground cover and light interception; therefore, it determines potential growth in many situations. Once ground cover is complete, however, high SLA becomes disadvantageous because leaf photosynthetic rates then limit growth. Consequently, the most efficient plant type in terms of growth would have high SLA during early growth stages and low SLA for leaves produced later on. Such a plant type would have high and early tillering. This reasoning, if accurate, has far-reaching consequences for plant-type concepts and selection criteria because selection for low-tillering types would most likely favor low-SLA types, associated with poor vegetative growth vigor and weed competitiveness. The new plant type is discussed in the light of these results. (Résumé d'auteur)