Transplanting date adjustment in Boro rice cultivation as a strategy to reduce groundwater withdrawal in Bangladesh

Abstract Due to intensive irrigated rice cultivation during dry season, declining trend of groundwater level is observed in many parts of Bangladesh. Field experiments were conducted in 2015 and 2016 at four experimental stations of Bangladesh Institute of Nuclear Agriculture, Bangladesh to evaluate the performance of some Boro rice cultivars in different planting dates with respect to yield and irrigation water requirement, and to optimize between yield and irrigation water requirement, with a view to reduce groundwater withdrawal and hence reducing mining of groundwater. The transplanting dates were 21th January (T1), 15th February (T2), 7th March and 30th March. Four rice cultivars were used. The results showed that the rice growth period was shortened with the later dates of transplanting. Generally, the yield of all cultivars at first and second transplanting (i.e. in T1 & T2) are good, and decrease at the later transplanting dates (T3, T4) and hence could not be recommended among farmers. When we consider the irrigation savings in T2 compare to T1, the irrigation savings varies with location and year, but the general tendency is that the second transplanting can save irrigation water. By considering economic (intrusion of additional rabi crop between two rice crops) and ecological factor (irrigation reduction, and hence reducing groundwater withdrawal), we recommend that the most suitable transplanting time for Boro rice should be 15th February. As such, we can effectively achieve good yield, reduce irrigation requirement, and creating opportunity for possible intrusion of addition Rabi crop between two rice crops. Overall, the findings of the present study can provide effective transplanting time and cultivar to reduce groundwater withdrawal in the present agro-ecosystem of northwestern Bangladesh and other similar areas.

A group researchers investigated the efficacy of different planting dates (PDs) on rainfed maize yield and water-use efficiency in semi-arid region of China [20]. They found that, within a certain time range, dry matter production per plant did not obviously change across PDs. However, the dry matter accumulation in the ear after flowering, the yield, and the WUE in the treatments under appropriate PDs were higher than those in the early or delayed PDs. Other group of researchers observed the yield, water productivity and economic returns of malt barley following different sowing dates and supplemental irrigation [21]. They found that in terms of water use, sowing by mid-November is advantageous since more rainfall is generally available. Some researchers conducted field study under the semi-arid environment of Faisalabad, Pakistan [22]. They observed significant differences among different transplanting dates of wet-season rice, as well as among cultivars. They explored that late transplanting caused yield reduction which could not be recommended among farmers.
For rice, information on different transplanting date and its effects on irrigation requirement and water use efficiency is scarce in Bangladesh. The frequency and total amount of irrigation for the growing season depends on the soil type, cultivar (maturity period, drought resistance capacity), ET demand, and rainfall availability [3,23,24]. Thus, it is not appropriate to make definite recommendation regarding the number and amount of irrigation to be applied for all cultivars under diverse planting dates. Utilization of rainwater without sacrificing yield can reduce valuable groundwater withdrawal, which is desirable for sustainability of the resource in the region [14]. Therefore, the present experiment was designed to evaluate the performance of some Boro rice cultivars in different planting dates with respect to yield and irrigation water requirement, and to optimize between yield and irrigation water requirement, with a view to reduce groundwater withdrawal and hence reducing mining of groundwater.

Materials and methods 2.1. Description of the experimental site
The experiment was conducted at four experimental stations of Bangladesh Institute of Nuclear Agriculture (BINA), Bangladesh, covering different agroecological and hydrological regions ( Figure 1). The latitude and longitude of the areas and main features are provided in Table 1.

Experimental design and field management
The experiment was conducted during Boro season (Jan.

Climate and rainfall distribution
The climate of the regions falls within humid sub-tropic with summer dominant rainfall. The annual rainfall at the study sites varies from 1500 mm to 3300 mm; approximately 70% of this rainfall occurs during the months of May to August, which is noted as monsoon season.

Fertilization, Irrigation and management
The seedlings of rice were transplanted according to the scheduled treatments. Each plot was fertilized uniformly with basal dose of Triple super phosphate, Muriate of Potash, Gypsum, and Zinc by 80, 110, 70, 45, 4.5 kg ha -1 , respectively. Urea was applied in three equal splits by 70 N ha -1 at 10, 30 and 45 days after transplanting. Measured amount of water was applied in each plot through water-meter, attached at the outlet point of rubber hose pipe. All cultural practices were followed as and when necessary. The grain and straw yields were adjusted to 12% moisture content following the equation [25]: where Mi the initial moisture content, Yi the initial yield (at Mi moisture content), Mt the targeted moisture content (say, 12 %), and Yadj the adjusted yield (at Mt % moisture content).

Data analysis
The analysis of variance technique (ANOVA) was carried out on the data. The software "STAR", developed by International Rice Research Institute (IRRI), was used to analyze the data. The significance among the means of the treatments was determined using least significant difference (LSD) technique at 5% probability level.  Means with same letter are not statistically different at 5% probability level by LSD.

Year 2015
The transplanting date showed significant difference in yield for each cultivars in different locations ( Table 2). For the cultivar V1, in general, the yield of second transplanting (T2) is slightly higher than that of first transplanting (T1), and gradually declines thereafter, except at Ishwardi location. For the cultivar V2, in general, the highest yield was obtained in first transplanting (T1), and then gradually declined, with the exception of Ishwardi location, where the second transplanting produced the highest and the last harvesting produced good yield. The yield of cultivar V3 is not consistent, first transplanting produced the highest at one location (Mymensingh), the second transplanting at two locations (Rangpur and Cumilla) and the third transplanting produced the highest yield at one location (Ishwardi).

Figure 3: Rainfall distribution during rice growing period in 2015
The mean yields over locations and average crop duration are summarized in Table 2.2. In general, crop duration decreases with late transplanting dates, mainly due to higher heat units ( Figure 2). The irrigation needed under different treatments (transplanting time) and cultivars is given in Table 2.3. The irrigation amount varied among locations, mainly due to variation of rainfall ( Figure 3). The highest irrigation amount was needed at Rangpur location.
Here, the first and second transplanting date (T1 and T2) needed almost similar irrigation, but the 3 rd and 4 th transplanting saved about 14% and 35% irrigation water compared to 1 st and 2 nd transplanting, coupled with declined yield (Table 2).

Year 2016
The mean grain yields under different treatments (transplanting time) are summarized in Table 3.1. The interaction effects of treatments and varieties showed significant effects at two locations, Rangpur and Cumilla. For the cultivar V1, in general, the yields of 1 st and 2 nd transplanting (T1 and T2) are statistically similar, and gradually declines thereafter, except at Ishwardi location. For the cultivar V2, for all locations, the highest yield was obtained in first transplanting (T1), and then gradually declined.
For the cultivar V3, similar trend of V2, that is, the highest yield at 1 st transplanting and then gradually declined. For the cultivar V4, the yields of 1 st and 2 nd transplanting are statistically similar at three locations (Mymensingh, Rangpur and Ishwardi) and differed at one location. The irrigation needed under different treatments (transplanting time) and cultivars are given in Table 3. The irrigation amount varied among locations, mainly due to variation of rainfall ( Figure 4). Average irrigation savings by T2V1 compared to T1V4 is 37%, while the grain yield decreased by about 14% (Table 3).   Figure 4: Rainfall distribution during rice growing period in 2016 The mean yields over locations and years are summarized in Table 4. Generally, the yield of all cultivars at first and second transplanting (i.e. in T1 & T2) are good and decrease at the later transplanting dates (T3, T4). So, optimization should be made between these two dates and the cultivars considering yield, irrigation water savings, and possible intrusion of addition Rabi crop between two rice crops.   In many areas of Bangladesh, two crops are grown in a year, and thus the land becomes free at first transplanting time. But the major land-type of Bangladesh is medium land, and in such areas, one Rabi crop (pulses, oil-seeds, potato or wheat) can be grown between two rice crops in a yearly cycle, and hence possibilities of increasing cropping intensity of the country. In such case, the land will be free at second transplanting time. In this case, the high performing cultivars at first transplanting (V1 and V4, Table 4) have to be sacrificed to gain additional Rabi crop, and irrigation savings in second transplanting (T2) compare to first one (T1). In second transplanting, the cultivar V1 and V2 produced almost similar yield. So, any of this cultivars can be grown. When we consider the irrigation savings in T2 compared to T1 (Table 4), the irrigation savings varies with location and year, mainly due to variation of rainfall amount and distribution. The crop duration of the cultivar also plays a role. The general tendency is that, the second transplanting can save irrigation water. When we consider the crop duration of cultivar V1 and V2 (Table 2), the V1 is shorter duration, and hence should be used (as the yield is similar).

Considering the average yield and possible intrusion of addition Rabi crop between two rice crops
By considering economic factor (intrusion of additional rabi crop between two rice crops) and ecological factor (irrigation reduction, and hence reducing groundwater withdrawal), we recommend that the most suitable transplanting time for Boro rice should be 15 th February. As such, we can effectively achieve good yield, reduce irrigation requirement, and creating opportunity for possible intrusion of addition Rabi crop between two rice crops. Overall, the findings of the present study can provide effective transplanting time and cultivar to reduce groundwater withdrawal in the present agro-ecosystem of northwestern Bangladesh and other similar areas.

Conclusion
As surface water supply is decreasing day by day, irrigation pressure is going towards groundwater resource. Judicious withdrawal of groundwater in intensive irrigated area is a crucial need for maintaining sustainability of the resource. Field experiments were conducted of Boro rice cultivars in different planting dates with respect to yield and irrigation water requirement, and to optimize between yield and irrigation water requirement, with a view to reduce groundwater withdrawal and hence reducing mining of groundwater. The results showed that in general, the yield of all cultivars at first and second transplanting are good and decreased at the later transplanting dates, which could not be recommended among farmers. When we consider the irrigation savings in second transplanting compared to first one, the irrigation savings varies with location and year, but the general tendency is that the second transplanting can save irrigation water. The cultivar Binadhan-14 (V1) can be selected for its good yield at second transplanting and short duration. By considering economic actor (intrusion of additional rabi crop between two rice crops) and ecological factor (irrigation reduction, and hence lower groundwater withdrawal), we recommend that the most suitable transplanting time for Boro rice should be 15th February. In this case, we can achieve good yield, reduce irrigation requirement, and create opportunity for possible intrusion of addition Rabi crop between two rice crops. The findings of the present study can help in selecting transplanting time and cultivar to reduce groundwater withdrawal in the present agro-ecosystem of northwestern Bangladesh and other similar areas.