Publications des agents du Cirad

Cirad

Optimization of source-sink dynamics in plant growth for ideotype breeding: A case study on maize

Qi R., Ma Y., Hu B.G., De Reffye P., Cournède P.H.. 2010. Computers and Electronics in Agriculture, 71 (1) : p. 96-105.

DOI: 10.1016/j.compag.2009.12.008

The objective of thisworkis to illustratehowa mathematical model of plant growth could be possibly used to design ideotypes and thus leads to new breeding strategies based on the guidance from optimization techniques. As a test case, maize (Zea mays L., cv. DEA), which is one of the most widely cultivated cereals all over the world, is selected for this study. The experimental data reported in a previous study are used to estimate parameters of a functional-structural plant growth model, namely, "GreenLab". As the corn cob and its leaves and stem can be benefited from economically, a single objective optimization problem (maximization of cob weight) and a multi-objective optimization problem (maximization of cob weight, maximization of leaf and stem weight) are formulated, respectively. The Particle Swarm Optimization approach is applied to solve these two kinds of optimization problems based on the GreenLab model. The optimized variables are specific parameters of the GreenLab model, which are the cob sink strength and the coefficients of the cob sink variation function. The optimization results revealed that to achieve breeding objectives, the optimal trade-offs of source-sink dynamics should be considered. Moreover, the optimization results of the multi-objective optimization problem revealed that the harvest index may not be the evaluation factor for yield improvement. The work described in this paper showed that such optimization approaches relying on plant growth models may help improve breeding strategies and design ideotypes of high-yield maize, especially in the current agricultural context with the increasing importance of co-products when designing cultivation practices.

Mots-clés : zea mays

Documents associés

Article (a-revue à facteur d'impact)