Water and air temperature impacts on rice (Oryza sativa) phenology

Abstract

Air temperature (Ta) is commonly used for modeling rice phenology. However, since the growing point of rice is under water during the vegetative and the early part of the reproductive period, water temperature (Tw) is likely to have a greater influence on crop developmental rates than Ta during this period. To test this hypothesis, we monitored Tw, Ta, and crop phenology in three commercial irrigated rice fields in California, USA. Sampling locations were set up on along a transect from the water inlet into the field. (Water warms up as it moves into the field.) Ta averaged 22.7 °C across sampling locations within each field, but average seasonal Tw increased from 22 °C near the inlet to 23.4 °C furthest away from the inlet. Relative to Tw furthest from the inlet, low Tw near the inlet delayed time to panicle initiation (PI 5 days) and heading (HD 8 days) and the appearance of one yellow hull on the main stem panicle (R7 9 days). Using Tw instead of Ta when the active growing point is under water until booting (midway between PI and HD) in a thermal time model improved accuracy (root-mean-square error, RMSE) for predicting time to PI by 2.5 days and HD by 1.6 days and R7 by 1.8 days. This model was further validated under more typical field conditions (i.e., not close to cold water inlets) in six locations in California. Under these conditions, average Tw was 2.6 °C higher than Ta between planting and booting, primarily due to higher daily maximum Tw values. Using Tw in the model until booting improved RMSE by 1.2 days in predicting time to HD. Using Tw instead of Ta during this period could improve the accuracy of rice phenology models.