Modelling the timing of the early development of winter wheat

The early apical development of Avalon winter wheat was modelled using data from a multisite survey of crops grown in the U.K. in one year. Parameter values defining the duration of the emergence to double-ridges and double-ridges to terminal spikelet phases were optimized with respect to the weather data. Different models were compared in terms of the scaled residual variation between crops in the accumulated value of the development index.The development sub-model of the whole crop model ARCWHEAT was taken as a starting point, though daily mean temperature was used instead of sinusoidal variation of diurnal temperature. Compared with the original parameter values, the optimized values indicated greater sensitivity to photoperiod, and less sensitivity to temperature and vernalization. Non-linear responses to temperature and photoperiod were examined, but none represented a consistent improvement over linear responses for these data. The sub-model of ARCWHEAT, which expresses development rate as the product of separate responses to temperature, photoperiod and vernalization, was compared with various alternative models. Whilst none of these fitted better than the ARCWHEAT formulation, some were not distinguishably worse with respect to these data. When tested against independent data, the optimized parameter values predicted double-ridges more accurately, and terminal spikelet as accurately as the original parameter values.