Alain OURRY, Frédéric Meuriot, Philippe Malagoli, Philippe Lainé, and Erwan Le Deunff. UMR INRA UCBN 950 Ecophysiologie Vegetale, Agronomie et nutritions N, C, S, Esplanade de la Paix, Université de Caen Basse Normandie, 14032 CAEN Cedex, France
Brassica napus is often used as a catch crop to deal with the issue of N leaching, but considering that this culture is also characterized by a low N index, prediction of the crop's N uptake capability and N partitioning is required. Based on the kinetic descriptions of N uptake, partitioning and mobilization in every plant organ, the aim of this work was to build up a compartmental model of N dynamics in oilseed rape. The use of kinetic equations of NO3- transport systems, determined by 15NO3O labeling under controlled conditions, combined with experimental field data from the INRA-Chalons rape database were used to model NO3- uptake during the plant growth cycle. The quantitative effects of different variables such as day/night cycle, ontogenetic stages, root temperature, PAR and soil nitrate availability on different components of the Low- and High-Affinity Transport Systems (LATS and HATS, respectively) of nitrate were taken into account. Simulated uptake correlated well with measured data of N uptake under field conditions for contrasted N fertilization rates tested. A second set of experiments was therefore conducted under field conditions where plants were harvested weekly after sequential 15N labeling in order to quantify in each tissue N coming from uptake or derived from mobilization. Logistic and exponential equations were fitted to the N flows for each compartment. Precise study of N flows from leaves at different nodes revealed the existence of two main groups of leaves in terms of their apparent capacity to mobilize N. Thus, N requirements for seed filling were mainly satisfied by N mobilized from vegetative parts (about 73 % of the total N in pods). A complex pattern of N mobilization from the leaves to vegetative or reproductive tissues was then highlighted but a general model was used to test different relevant parameters.