Simulation of the effects of genotype and N availability on rice growth and yield response to an elevated atmospheric CO2 concentration

The objective of this study was to identify physiological processes that result in genotypic and N fertilization effects on rice yield response to elevated atmospheric CO2 concentrations ([CO2]). This study conducted growth and yield simulations for 9 rice genotypes grown at 4 climatically different sites in Asia, assuming the current atmospheric [CO2] (360 ppm) and elevated [CO2] (700 ppm) using 5 levels of N fertilizer (4, 8, 12, 16, 20 g m(-2) N fertilizer). A rice growth model that was developed and already validated for 9 different genotypes grown under 7 sites in Asia was used for the simulation, integrating additional components into the model to explain the direct effect of [CO2] on several physiological processes. The model predicted that the relative yield response to elevated [CO2] (RY, the ratio of yield under 700 ppm [CO2] to that under 360 ppm [CO2]) increased with increasing N fertilizer, ranging from 1.12 at 4 g m(-2) N fertilizer to 1.22 at 20 g m(-2) N fertilizer, averaged overall genotypes and locations. The model also predicted a large genotypic variation in RY at the 20 g N treatment, ranging from 1.08 for 'WAB450-I-B-P-38-HB' to 1.41 for 'Takanari' averaged overall locations. Combining all genotypes grown at the 5N fertilization conditions, a close 1:1 relationship was predicted between RY and the relative [CO2] response in spikelet number for crops with a small number of spikelets (less than 30,000 m(-2)) under the current atmospheric [CO2] (n = 18, r = 0.89***). In contrast, crops with a large number of spikelets under the current atmospheric [CO2] showed a significantly larger RY than the relative [CO2] response for spikelet number per unit area. The model predicted that crops with a larger number of spikelets under the current atmospheric [CO2] derived great benefit from elevated [CO2] by directly allocating increased carbohydrate to their large, vacant sink, whereas crops with a smaller number of spikelets primarily required an increased spikelet number to use the increased carbohydrate to fill grains. The simulation analyses suggested that rice with a larger sink capacity relative to source availability under the current atmospheric [CO2] showed a larger yield response to elevated [CO2], irrespective of whether genotype or N availability was the major factor for the large sink capacity under the current [CO2]. The model predicted that the RY response to nitrogen was brought about through the N effects on spikelet number and non-structural carbohydrate accumulation. The genotypic variation in RY was related to differences in spikelet differentiation efficiency per unit plant N content. Further model validation about the effects of [CO2] on growth processes is required to confirm these findings considering data from experimental studies. (C) 2011 Elsevier B.V. All rights reserved.