Study site: The experiment was conducted at a farmer’s field near Rahmatpur, Mymensingh, Bangladesh (24°45' N and 90°24' E). The topography of the land is medium high and the soil belongs to the Old Brahmaputra Flood Plain. The major climatic parameters including air temperature, humidity, rainfall, wind speed, sun shine, solar radiation and evaporation data during the experimental period were collected from the weather yard of the Department of Irrigation and Water Management of Bangladesh Agricultural University.Experiment The soil of the experimental land was first opened with a power tiller and kept exposed to the sun. Later, it was prepared by several ploughing followed by laddering. The weeds and debris were removed from the land. After final land preparation, the plots were demarcated and levees were made around the individual plots to retain irrigation water. The buffer zones were made to prevent the effect of seepage of water between the adjacent plots. Doses of nitrogen, phosphorus, potassium, sulphur, zinc and boron fertilizer for mustard cultivation were 66.7, 41.7, 37.5, 33.3, 3.3 and 3.3 kg ha-1, respectively as recommended by BARC (FRG, 2012).
The experimental field was laid out in split-plot with four irrigation treatments in main plots and varieties in sub-plots and each treatment having three replications. The 6 m × 4 m layout experimental units or plots were separated from each other by 70 cm wide buffer zone to prevent seepage between nearby plots. Similarly, the replication blocks were separated from each other by 30 cm wide buffer zone. Irrigation water was made available to the experimental plots by using hose pipe provided in the buffer zone between two replicate blocks.Irrigation scheduling is defined as when to irrigate and how much water to irrigate. There are different approaches of irrigation scheduling on the basis of growth stage, permissible soil water depletion and measured pan evaporation. In this study, irrigation water was applied to the crop on the basis of growth stages. The four irrigation treatments were as follows: control (T1), irrigation at the vegetative stage (25–30 days after sowing (DAS) up to field capacity (T2), irrigation at the flowering stage (45–50 DAS) up to field capacity (T3), irrigation at the vegetative (25–30 DAS) and flowering stage (45–50 DAS) up to field capacity (T4). Three varieties of mustard selected for the study were as follows: Binasarisha-7 (Brassica juncea L) (V1), Binasarisha-8 (Brassica juncea L) (V2) and Binasarisha-4 (Brassica napus L) (V3).
The mustard seeds were sown by line on 03 November, 2012. The seed rate was 7 kg ha-1. After line sowing the seeds, laddering was done to cover the seeds with loose soils. A calendar of various operations done during the experiment and starting time of different growth stages were recorded. Several weeds grew in the experimental plots that were uprooted manually. Weeding was done after 20 DAS. Irrigation water was applied as per schedule of the irrigation treatments. Two important growth stages of mustard were identified for irrigation as predefined by experimental treatments. Crop was harvested on 25 February, 2013 at the time of 70% maturity. Harvesting was done carefully to avoid field losses.
Data collection: Ten sample plants were selected randomly from each treatment for further data collection. Different agronomic data including plant height, number of branch, number of pod/plant, pod length, number of seed/pod, straw weight, and 1000-seed weight were recorded.
Irrigation water needed was calculated at each stage of the crop growth. Soil moisture content in each plotwas measured by gravimetric method (oven dried at 105 °C for 24 h) up to 0–60 cm depth for every 15 cm increment at the time of sowing, before and after irrigation, and at the time of harvest.
Data simulation: HYDRUS-1D model (Šimunek et al., 2013) was used to see the soil water dynamics between the sampling events and estimate overall soil water storage in root zone area during the study period. Soil physical properties, weather parameters, and crop information collected during the experiment were used as input to this model. A deterministic approach, based on the numerical solution of the Richards equation by means of linear finite elements, was used to simulate soil water movement. The description of the simulation process is brief here because it was not the main focus