2.1. Study area and climate Bangladesh is situated between 200 and 270 N latitude and 880 and 930 E longitude. Of the three ecological zones in Bangladesh, the plains are most suitable for crop production. Four districts within the plains zone were selected as study area (Dhaka, Mymensigh, Rajshahi and Rangpur). The districts (i) are situated in a drought prone region, (ii) use groundwater as the major source for irrigation and drinking water and (iii) have sufficient climatic and hydrological data available. The area is characterized by a subtropical climate with mean annual temperature of 24.8–26.8 0C and average yearly rainfall of 2095–2246 mm which is concentrated between April and October. Although average annual rainfall is high, rainfall during the wheat growing period (December–March) is very low. Mean monthly rainfall during the wheat growing period varies between 44 and 75 mm among the study locations, while mean monthly evapotranspiration varies between 119 and 132 mm.
2.2. Meteorological drought assessment Standardized precipitation index (SPI) (McKee et al., 1993) and reconnaissance drought index (RDI) (Tsakiris et al., 2007) methods have been used to identify meteorological drought and to assess anomalies of precipitation and evapotranspiration. Thirtytwo years (1979–2011) of daily precipitation, daily temperature (maximum, minimum), wind speed and other climatic data were collected from the Bangladesh Meteorological Department (BMD) for four selected stations in the study area. Reference evapotranspiration (ET0) was calculated using the FAO-Penman-Monteith equation (Allen et al., 1998) using temperature, sunshine hours, wind speed, and humidity data (Eq. (1)).
2.3. Field experiments Field experiments were conducted at the experimental farm of the Bangladesh Agricultural University (BAU) at Mymensingh (24.75 0N latitude, 90.50 0E longitude and 18 m above mean sea level) during the wheat growing seasons (December–March) of 2007–2008 and 2008–2009. Wheat was sown respectively on 20 and 6 December for growing season I and II. During the wheat growing season (December–March), rainfall was respectively 0 mm, 36.60 mm, 5.40 mm and 95.30 mm per month in 2007–2008 while in 2008–2009 rainfall occurred only in March: 18.1 mm. Mean daily evaporation was between 1.89 mm day−1 and 3.84 mm day−1. Mean monthly maximum temperature is between 25 and 27 0C while mean monthly minimum temperature was 15 0C.
2.3.1. Field preparation and soil type The study area was divided into three blocks with a distance of 2 m between the adjacent blocks. Each block was subdivided into 5 unit plots each of size 1 m2. The distance between two adjacent unit plots was 2 m. There were fifteen plots in total. A pit of 1 m2 area with a depth of 0.6 m was dug in each plot. The field experiments were performed for five soils with different soil textures: loamy sand (exogenous natural material; T1), sandy loam (T2), loam 1 (T3), loam 2 (T4), and silty loam (control = natural soil in the experimental field; T5). The natural soil (silt loam) dug out from the plots was thoroughly mixed manually with loamy sand as amendment. Loamy sand (T1) was collected from the shore of a local perinnial river (old Brahmaputra), sundried and sieved with a 2-mm mesh to remove any debris from it. It was added as source of coarse textured materialto the finer textured endogenous soil (silty loam, T5) to prepare the soils for treatments T2, T3 and T4(by mixing 75, 50 and 25 percent of loamy sand with silty loam). A polyethylene sheet was placed on the four vertical sides of the pits to prevent seepage of water. Three pits were filled with the airdried loamy sand (T1), three others with dug out silt loam (natural soil, T5) and the rest of the pits were filled with the mixture of silt loam and loamy sand at different percentages. The soil in the field below the 0.6 m is silt loam. The treatments were replicated three times and the experiments were set up in a Randomized Complete Block Design (RCBD).
Soil moisture content at the field capacity of each soil treatment was measured according to the TDR method using a Trime FM moisture meter. The physical characteristics of the different soil treatments were derived using pedotransfer functions using information on particle size distribution (determined by the hydrometer method), and organic carbon content (Saxton and Rawls, 2006) (Table 1). The physical properties obtained from pedotransfer functions are used in further simulation. Water content at field capacity and organic carbon increased with increasing clay and silt content of the treatments. This is also confirmed by soil-water retention curves of different treatments from the same experiments (Mojid et al., 2012; Mustafa, 2009). wheat was cultivated using nutricoat fertilizer containing Urea, TSP, MP and Gypsum @ 200, 160, 50 and 120 kg ha−1 respectively as recommended by Hussain et al. (2006). Cow dung @ 8.5 t ha−1 was also used during land preparation. A wheat variety called Shatabdi, developed by the Bangladesh Agricultural Research Institute was cultivated with the standard intercultural operation.
2.3.2. Irrigation management Wheat was irrigated using the check basin method also known as border irrigation. In this method, the irrigated field is surrounded by soil embankments that limit surface runoff. Irrigation was scheduled based on crop water requirements (quantity of water required to minimize water stress for the crop) and observations of soil moisture in the first 20 cm depth. Soil water content was recorded at sowing, before each irrigation, immediately after the irrigation, and 48 h after irrigation or rainfall. It was also measured at the time of harvesting to know the final soil-water content at the end of the crop growth. Four irrigations per season were applied during the first and the second year experiment. Available soil moisture was also measured with the Trime FM moisture meter in all treatments before irrigation. Irrigation was applied based on the quantity of water required to reach field capacity for each plot. The 1st, 2nd, 3rd, and 4th irrigation were applied respectively at 23, 53, 61, and 72 DAS in the first year and 19, 45, 62, and 77 DAS in the second year. The required quantity of water for a plot in each irrigation was calculated by using the formula.
2.3.3. Data recording Several crop, irrigation and soil measurements were made during the field experiments. Leaf area index (LAI) defined by Redford (1967) was measured for each plot in the second year. The leaves often representative plants were collected from each plot at 75 DAS and their total area was measured using a leaf area meter (give specifications of the meter here). The total area covered by the ten plants was calculated from the density of plant population. Finally, the LAI was calculated as a ratio of leaf area to the ground area. Then, this calculated LAI was converted to canopy cover (CC) using the relationship between LAI and CC defined by Hsiao et al. (2009) and Heng et al. (2009).