The experiment was conducted at Bangladesh Agricultural University Farm, Mymensingh during the Boro season of 2006. The study area includes mainly the Old Brahmputra Floodplain soil. The experimental soil was silty loam. The colour of the soil seemed to be grey, whereas, P", bulk density, field capacity and wilting point of the soil were 5.48 to 6.36, 1.40 gm/cm3, 27.63% and 14.06%, respectively. The seasonal precipitation was 56.18 cm and the average daily air temperature of the experimental area was 25.85°C. The experimental field was prepared by using a tractor-drawn plough, puddler and ladder (wooden plank or leveler). The experiment was laid out in a randomized complete block design (RCBD) with three replications. Each plot was provided with a calibrated scale to measure the water level depth each day. The size of the unit plot was 4 m x 3m. Around each unit plot having levee of 25 cm width and 20 cm height was constructed. Each plot was separated by a 1 m buffer zone between the rows and a 1.5 m wide buffer zone between the plots along rows to prevent seepage between the adjoining plots.
Triple superphosphate and muriate of potash fertilizers were applied at the final preparation of land at the rate of 77 kg/ha and 59 kg/ha, respectively and urea fertilizer was applied thrice, namely, at the final land preparation, 20 days after transplanting (DAT) and 35 DAT equally in all plots at the rate of 304 kg/ha. The study consisted of the following treatments- 5 cm standing water was maintained throughout the growing season (T1), 5 cm standing water was applied as soon as the water disappeared from the field (T2) i.e. saturation was reached, 5 cm standing water was applied 3 days after disappearing of standing water (13), 5 cm standing water was applied 5 days after disappearing of standing water (T4) and 5 cm standing water was applied 7 days after disappearing of standing water (T5). BRRI dhan29 variety selected for the study was an inbred modern variety, developed by the Bangladesh Rice Research Institute in 1995. On average, the growth duration of BRRIdhan29 is about 168 days (BRRI, 1995). The seedlings of the variety were collected from the BAU farm on 2 February 2006. The seedlings were 10 to 17 cm in height and they were 40 days of age.
Transplanting was done on 3 February using 25 cm x 15 cm spacing with 2 seedlings/hill.
The amount of water needed for land preparation varies with soil type, degree of prior drying out and cracks of the soil profile, and the time is taken for the clay fraction to swell. For this study, the amount was assumed to be 240 mm for each treatment. Seepage and percolation losses were computed by applying the water balance technique. These losses were determined using the formula:
S & P = WDt_i-WD-ET (1)
and ET = b+a (EVt) (2)
where, S & P = Seepage and percolation loss (cm/day), "t" refers to the day of measurement, WD= Water depth measured in cm, on the meter stick, ET= Daily rate (cm) of evapotranspiration, EV= Daily rate (cm) of evaporation, a= Pan factor, and b = An empirical constant = 0.5
In this study seepage and percolation rate was considered to be 2 mm/day for 100 days and thus accounts the total seasonal loss of 20 cm. The simple equation used to estimate the water requirements for growing a rice crop is given by Walker (1994) as
0 = W+R+C+E+F . (3)
where, Q = Total water requirements (cm), W = Wetting up and land preparation (cm), R= Surface runoff (cm), C= Conveyance loss (cm), E = Evapotranspiration from water and plants (cm), and F= Seepage and percolation (cm)