Rice seedling emergence prediction method for timing of weed control in direct-sowing rice on well-drained paddy fields in Hokkaido

We aimed to determine the timing of weed control in direct sowing of rice on well-drained paddy fields in Hokkaido region. To this end, we developed calculation models to estimate the date of beginning of seedling emergence from the sowing date, based on an effective accumulated temperature model. The model had different base temperatures set for the earlier and later periods, from sowing to emergence, to calculate daily effective temperature. As non-selective herbicides to remove weeds must be applied before the emergence of rice, new methods to ensure their appropriate application prior to the seedling emergence date are required. These methods must be accurate, as an estimated seedling emergence date that occurs after seedling emergence will be detrimental to the crop. Our results indicate that the use of daily mean soil temperature is more appropriate to estimate the emergence date, and that it is possible to use the daily mean soil temperature which is estimated from daily mean air temperature. The root mean square error of the estimated seedling emergence date is 2.0 days, with only one case of the estimated date occurring after the actual date.