Before starting a breeding program for a specific crop or variety, it can be helpful to know how traits behave in determining yield under different conditions and environments. Crop growth models can be used to generate valuable information on the relevance of specific traits for an environment of interest. In this paper, the simulation model CMS-Cropsim-CERES- Wheat was used to test the performance of input parameters which describe cultivar differences concerning plant development and grain yield. In so-called ideotyping sequences, the specific cultivar parameters were varied and the model was run with the same management information in four different scenarios. The scenarios consisted of two locations, Wuqiao (37.3°N, 116.3°E) and Quzhou (36.5°N, 115°E) in Hebei Province (North China Plain), and a dry and a wet growing season for each location. The input parameter G1 (corresponding trait: kernel number per spike) followed by G2 (corresponding trait: kernel weight) had the biggest influence on yield over all scenarios. The input parameters P1V (corresponding trait: vernalization requirement) and P1D (corresponding trait: photoperiod response) also played an important role in determining yield. In the dry scenarios a low response in vernalization and photoperiod generated a higher yield compared to a high response. The lower responses caused earliness and the period of late water stress was avoided. The last relevant parameter that affected yield was PHINT (corresponding trait: leaf area of first leaf). The simulation showed that with an increasing PHINT, yield was enhanced over all scenarios. Based on the results obtained in this study, plant breeders could carefully select the relevant traits and integrate them in their breeding program for a specific region.
展开▼
