Effects of nitrogen and radiation on dry matter and nitrogen accumulation in the spike of winter wheat

In wheat, spike growth prior to anthesis is a key period influencing kernel set. Incident radiation and crop N status affect the accumulation of dry matter (DM) and nitrogen (N) in the spike, as observed at anthesis. However, the influence of these two factors on the dynamics of spike growth throughout the pre-anthesis period is not fully understood. Our objective was to investigate the effects of incident radiation and crop N nutrition on the dynamics of spike dry matter, N concentration and N content throughout the pre-anthesis period and to propose models describing spike DM and N content during spike growth, per unit area. We also investigated whether kernel number was related to spike N content at anthesis independently of spike DM, a controversial issue. Field experiments were conducted over two seasons with 'Tremie' winter wheat and over an additional season with 'Arche' and 'Soissons'. Treatments involved the application of N fertiliser at different stages and rates to achieve various levels of N deficit and the use of shading nets at various periods during spike growth to reduce incident radiation. Shading reduced spike DM but increased spike N concentration; it reduced spike N content whenever there was a large decrease in incident photosynthetically active radiation (PAR). All these effects occurred only during the second half of the spike growth period, even if shading was applied from the onset of spike growth. Crop N deficit reduced spike DM, N concentration and N accumulation. Spike DM and N content were affected throughout the spike growth period if the deficit started before the beginning of spike growth, and was severe. Continuous and temporary N deficits induced similar responses. The interaction between shading and N fertilisation was not significant for spike DM and N content, but was significant for spike N concentration. Two models, one predicting spike DM and the other predicting spike N content at any time during spike growth were developed. They were validated with an independent data set and by cross-validation with the initial database. Both models fitted the observed data well. The input variables, cumulative PAR after sowing, nitrogen nutrition index (NNI) of the crop and cumulative thermal time after sowing are easy to measure or to simulate. Kernel number was correlated with both spike DM and spike N content at anthesis, demonstrating that N content played no role in determining kernel number independently of DM. (C) 2004 Elsevier B.V All rights reserved.