Abstract: The difference in response patterns of development rate to temperature and daylength, the phase division and parameter setting, and the relationship among accumulated growth degree day(GDD), development rate(DVR) and development physiological day number(DPD) were discussed by comparing the phasic development models from CERESrice, ORYZA2000, RCSODS and RiceGrow. The four models showed a similar response pattern of development rate to temperature. The parameter setting in CERESrice, ORYZA2000 and RiceGrow was same in basic vegetative phase and grain filling phase, although different parameters were used in these models, such as GDD, DVR, and DPD. During photoperiodsensitive phase, three parameters, such as development rate, critical daylength and daylength effect factor were used in these models, but one parameter was fixed in CERESrice and RiceGrow. One new parameter was applied in ORYZA2000 to describe the development process in reproductive phase, but no new parameter was introduced to development models in CERESrice and RiceGrow. The response patterns of development rate to daylength in photoperiodsensitive phase in CERESrice, RCSODS and ORYZA2000 was similar, which was different to that in RiceGrow. The simple phase division in RCSODS led to fewer genetic parameters. The phasic development model in ORYZA2000 was the most common, but with maximum parameters, and which could not be obtained merely from cultivar data.

Key words: rice, phasic development, model, comparison

摘要： 通过对CERESrice、ORYZA2000、RCSODS、RiceGrow等模型中发育期模型的比较与分析，揭示了这些模型在发育速率对温度的响应模式、发育期阶段划分与参数设置、日长效应模式的差别，分析了累积生长度日、发育生理日及发育速率的关系。研究表明，4个模型的发育速率对温度的响应模式基本一致；CERESrice、ORYZA2000、RiceGrow在基本营养生长期和籽粒灌浆期参数设置一致，而在光敏感期虽然都有3个参数（发育速率、临界日长、日长效应系数），但CERESrice和RiceGrow中各有一个参数固定，穗发育期CERESrice和RiceGrow均没有引入新参数，而ORYZA2000则引入新参数；CERESrice、RCSODS和ORYZA2000的日长效应模式基本一致，而与RiceGrow差异较大。RCSODS中的发育期模型阶段划分较为简单，参数较少；ORYZA2000模型最为通用，但参数较多，且仅根据发育期资料无法唯一确定。

关键词: 水稻, 发育期, 模型, 比较