Grain yield analysis of maize probe genotypes used to determine environnemental limiting factors in multi-environment trials
Résumé
Genotype x environment interaction was investigated in multi-environment trials for grain yield of maize hybrids (Zea mays) using a crop diagnosis of yield-limiting factors. Factors that limit yield in multi-environment trials of maize, can be identified a posteriori by a crop diagnosis of physical, meteorologial and biologial conditions. Crop diagnosis requires measurements of yield and yield components on a specific set of genotypes, called probe genotypes (Cooper and Fox, 1996). Yield can be split into two main components formed at two distinct periods: kernel number per square meter (KN) which is determined before flowering, and thousand kernel weight (TKW) which is determined after flowering. Maximal TKW is related to KN by a boundary line (Webb, 1972) and KN threshold. Maximal TKW is only reached when KN is low, KN threshold is the value beyond which one the boundary TKW decreases due to competition between the kernels during their filling. Such a model was calibrated by Siband et al. (1999) on maize with data coming from an on-farm survey. When yield or yield components of these genotypes, compared with preestablished optimal values are reduced, one or several factors that limited yield or yield components can be diagnosed. In our study, the model is calibrated with experimental data corresponding to 6 commercial hybrids tested in 37 environments in France (combinations of site and year): Antares (very early flint dent), Banguy (early flint dent), Anjou285 (semi-early flint dent), DK312 (semi-early dent), Naudi (semi-late dent), and Cerbere (late dent). The maximum values for TKW and grain yield were defined for each probe genotype and from the data, assuming that in some sites, the full potential of the cultivar is expressed. Means of optimal values for grain yield and its main components were determined and confidence intervals were estimated by a simple random sampling with replacement. Deviations of yield as well as yield components were determined for each probe genotype in each environment. A great variability was observed among the hybrids for the optimal values and for deviations of yield components in the tested environments. Optimal thousand kernel weight exceeded 300 g for all hybrids, except Anjou285. The two semi-early genotypes have formed their yield in a different manner: maximum TKW seemed to be higher for DK312 than for Anjou285 while their KN thresholds seemed to be similar. The semi-late and late genotypes, Naudi and Cerbere, showed the highest reductions for TKW. Diversity was also observed among environments. Most environments were affected during the whole formation of yield. Few environments were free from stress (Mauguio in 2000, or Ondes in 2001) or seemed to be affected only before flowering (Lusignan in 2000) or only after flowering (Lusignan in 2002). The yield components could be affected for all probe genotypes (Le Pin in 2000 for TKW), or for some genotypes (Mons in 2000 for TKW). Crop diagnosis allows to determine the main yield-limiting factors of the environment, and their period of occurrence. These factors will be integrated in a further detection of QTL for cold tolerance.
Domaines
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |