The study area was located in Natshail village of Mohadebpur upazila of Naogaon district in the north-west region of Bangladesh. Mohadebpur upazila occupies an area of 397.67 sq km, located between 24˚48' and 25˚01' north latitudes and between 88˚38' and 88˚53' east longitudes, at a height of 25 meters above sea level and belongs to the High Barind Tract (AEZ 26). Another study area was located in Jonakipara village of Nachole upazila of Chapainawabganj district in the north-west region of Bangladesh. Nachole upazila occupies an area of 283.67 square kilometers, located between 24˚38' and 24˚51' north latitudes and between 88˚15' and 88˚21' east longitudes at 25 meters above sea level and belongs to the High Barind Tract (AEZ 26). During the experimental period, the average maximum temperatures were recorded as 34.8˚C and 36.14˚C during 2017-2018 and 2018-2019, and the minimum temperatures were recorded as 8.35 and 10.43˚C, respectively. The average monthly humidity during the rice growing season was 65%. However, the monthly average distribution of rainfall from January to April was very low and uneven; the range of effective rainfall was 6.00 mm to 43.2 mm, which indicates the necessity of irrigation water for rice cultivation during dry rabi season.

The soil was clay loam in texture, having a low status of organic matter, including phosphorus and potassium. The pH of the soil recorded 6.1 - 6.4, organic matter 0.56% - 0.65%, total nitrogen 0.09% - 0.10%, phosphorus (P) 2.62 ppm, exchangeable K 0.16 - 0.20 meq.100 g⁻¹ soil and Sulphur 12.9 - 13.6 ppm.

The experiment was laid out in a randomized complete block design (RCBD) with three replications. The unit plot size was 10 m² (5 m × 2 m). Each 10 m² plot contained 210 seedlings. To facilitate cultural operations, proper spacing was kept among the plots and blocks. Field experiments were carried out during the periods from December 2017-May 2018 (dry irrigated Boro season) and December 2018-May 2019. In this study, three irrigation treatments were followed: Conventional irrigation (Ic), Alternate wetting and drying (AWD), and aerobic irrigation. Under each irrigation practice, four soil amendments were selected: T1: No NPKS, No amendments, T2: NPKS 100% recommended fertilizer (RFD) + Vermicompost (VC) 2.5 t∙ha⁻¹ + Phospho-gypsum (PG) 1.0 t∙ha⁻¹, T3: N (50% RFD) with recommended PKS + VC 7.5 t∙ha⁻¹ + PG 2.0 t∙ha⁻¹, T4: N (25% RFD) with recommended PKS + VC 10.0 t∙ha⁻¹ + PG 2.5 t∙ha⁻¹.

In conventional irrigation, 5 cm of standing water in the experimental field was maintained from rice seedling transplanting to establishment period and irrigation stopped before two weeks of harvesting. In case of AWD irrigation, perforated PVC pipes were installed in the experimental plots 10 days after transplanting (DAT) according to treatments for measuring soil water depletion to follow AWD techniques. The diameter of the pipe was 8 cm and the length was 25 cm. After irrigation water application, water entered through perforations and water level inside the pipe was at the same level as that of outside. With the progress of time when water level depleted in AWD plots, 5 cm irrigation was done when the depleting water table inside the pipe fell 15 cm below ground level.

Water requirement for rice cultivation was computed by adding applied irrigation water, effective rainfall during growing season and water for land preparation ( Rashid, 1997 ). Water use expressed as m³/ha is the amount of irrigation water used at each experimental site that was estimated using a flow meter connected with irrigation pump plus the total amount of rainfall recorded during the rice growing period.

In the study, water saving percentage was calculated as follows:

$\begin{array}{l}\text{Water Savings}\left(\%\right)\\ =\frac{\text{Water supplied in flooded plot}-\text{Water supplied in AWDI}/\text{Aerobic plot}}{\text{Water supplied in flooded plot}}\times 100\end{array}$

Water productivity is expressed as the ratio of grain yield (kg/ha) per unit water (m³/ha) supplied, including rainfall ( Jaafar et al., 2000 ) and calculated as follows:

$\text{Water productivity}\left(\text{kg}/{\text{m}}^{\text{3}}\right)=\frac{\text{Grain Yield}\left(\frac{\text{kg}}{\text{ha}}\right)}{\text{Total water supplied}\left(\frac{{\text{m}}^{\text{3}}}{\text{ha}}\right)}$

A modified closed-chamber method ( Ali et al., 2008 ; Rolston, 1986 ) was used to estimate CH₄ emission during rice cultivation. Gas samples were collected by 50ml gas-tight syringes at 0, 15 and 30 minutes after chamber placement over flooded plots at different rice growth stages to get average CH₄ emissions. The dimension of closed chamber was 62 cm × 62 cm × 112 cm. Samples were analyzed to determine CH₄ concentration by gas chromatograph (Shimadzu, GC 2014, Japan) with a Flame Ionization Detector. The temperatures of column, injector and detector were adjusted at 100˚C, 200˚C and 200˚C, respectively. A closed-chamber equation ( Rolston, 1986 ) was used to estimate methane fluxes for every treatment.

F = ρ × V/A × Δc/Δt × 273/T

where, F (Flux) = CH₄ emission rate (mg CH₄ m⁻² hr⁻¹), ρ = gas density (0.714 mg cm⁻³), V = volume of chamber (A × h; m³), A = surface area of chamber (length × width; m²), h = height of the chamber (m), Δc/Δt = rate of increase of CH₄ gas concentration (mg∙m⁻³∙hr⁻¹), T (absolute temperature) = 273 + mean temperature (˚C).

Total seasonal methane emission/flux for the entire cropping period was computed by the formula ( Singh, 1999 ): Total CH₄ flux = ${\displaystyle {\sum }_{i=1}^n\left(Ri\times Di\right)}$, where, Ri = rate of methane flux (g∙m⁻²∙d⁻¹) in the ith sampling interval, and n = number of sampling intervals.

CH₄ flux was calculated based on following equation

E = Slope (ppm/min) × VC × MW × 60 × 24 × 22.4 (273 + T/273) × Ac × 1000

The emissions as kg CH₄ (or Kg N₂O)/ha were derived from the slope of the linear regression curve of gas (CH₄ and N₂O) concentrations against the chamber closing time. The slope was referred to as mass per unit area per unit time (mg/m2/h), where VC is the volume of the gas chamber in liters (L), MW is the molecular weight of the respective gas, 60 is minutes per hour and twenty four is hours of the day. The volume of 1 mol of gas in L at standard temperature and pressure is 22.4. T is the temperature inside the chamber (˚C) while 273 is the standard temperature of ˚K. AC is the chamber area (m²) and 1000 is μg/mg.

To estimate the GWP, CO₂ is typically taken as the reference gas, and an increase or reduction in emission of CH₄ is converted into “CO₂-equivalents” by means of their GWPs. In this study, we used the IPCC factors to calculate the combined GWP for 100 years (GWP = 25 × CH₄, kg CO₂-equivalents∙ha⁻¹) from CH₄ under various agricultural irrigation practices. In addition, the greenhouse gas intensity (GHGI) was calculated by dividing GWP by grain yield for rice ( Mosier et al., 2006 ).

Soil redox potential (Eh) was measured during rice cultivation at certain time intervals by a glass electrode Eh meter. At∙harvesting stage, soil bulk density (BD) was analyzed using cores (volume 100 cm³, inner diameter 5 cm), filled with fresh moisture soils. The collected core samples were oven dried at 105˚C for 24 h, and then measured the weight of dried core samples was measured. Soil porosity was calculated using BD and particle density (PD, 2.89 Mg∙m⁻³) according to the equation: porosity (%) = (1 − BD/PD) × 100. At∙harvesting stage, chemical properties of the collected soil samples were analyzed for organic carbon by wet oxidation method ( Allison, 1965 ), organic matter content by multiplying the percent organic carbon with Van Bemmlen factor of 1.73, total nitrogen by Micro-kjeldhal method ( Nelson et al., 1980 ), available phosphorus by Olsen method ( Olsen & Sommers, 1982 ), exchangeable potassium by Flame photometer ( Brown & Lilleland, 1946 ).

At first, experimental data were entered into Microsoft Excel. Then, analysis of variance (ANOVA) was performed using R software (R-4.3.3, 2024 version). Duncan’s multiple range test (DMRT) was conducted to identify statistically significant differences between group means at a 5% and 1% significance level.

After rice was transplanted in the field, CH₄ emission rate within the first two weeks was low ( Figures 1(a)-(c) , which increased significantly at active tillering stage (35 DAT) and peaked at flowering to heading stage (77 - 84 DAT). In general, higher CH₄ emission rates were recorded in conventional irrigation practices compared to AWD and aerobic irrigation practices ( Figures 1(a)-(c) ). Among the amendments, vermicompost (VC) at 7.5 - 10.0 t/ha with phosphogypsum (PG) 2.0 - 2.5 t/ha and half of the recommended N-fertilizers showed lower CH₄ emission rates compared to recommended (100%) N-fertilizers with VC 2.5 t/ha and PG 1.0 t/ha. At 77 DAT, CH₄ peak 30 - 35 mg/m²/hr was observed in recommended chemicals with VC 2.5 t/ha and PG 1.0 t/ha (T2). After that, CH₄ emission rate sharply dropped with rice grain maturation. The least CH₄ emission was observed at 119 DAT before rice harvest.

In this study, maximum cumulative CH₄ flux was observed under conventional irrigation; however, AWD and aerobic irrigation practices significantly reduced the cumulative CH₄ flux for irrigated boro rice cultivation ( Table 1 ). At Mohadebpur, Naogaon experimental field, the highest cumulative CH₄ emission was calculated 221 kg∙ha⁻¹∙season⁻¹ in vermicompost (VC 2.5 t∙ha⁻¹) amended field plot (NPKS 100% RFD + phospho-gypsum 1.0 t∙ha⁻¹) under conventional irrigation; this decreased to 178.0 kg∙ha⁻¹ and 170.0 kg∙ha⁻¹ under AWD and aerobic irrigation practices, respectively. The increasing levels of vermicompost amendments 7.5 - 10.0 t∙ha⁻¹ and PG 2.0 - 2.5 t∙ha⁻¹ decreased cumulative CH₄ emissions by 11% - 17%, 13% - 23% and 13% - 21.7% under conventional irrigation, AWD and aerobic irrigation practices compared to Vermicompost (2.5 t∙ha⁻¹) amendments with NPKS 100% plus phospho-gypsum (1.0 t/ha). At Nachole, Chapainawabganj district experimental site, the highest cumulative CH₄ was found 195.80 kg∙ha⁻¹∙season⁻¹ (T2: NPKS 100% + Vermicompost 2.5 t∙ha⁻¹ + phospho-gypsum 2.5 t∙ha⁻¹), followed by 172.20 (VC 7.5 t∙ha⁻¹ with N (50% of recommended dose) and 152.40 kg∙ha⁻¹∙season⁻¹ with VC 10.0 t∙ha⁻¹ with N (25% of recommended dose) under conventional irrigated field plots. For AWD and aerobic irrigation practices, the seasonal cumulative CH₄ emissions were decreased by 16% - 23.0% and 13% - 21.0% with VC 7.5 - 10.0 t/ha amendments compared to VC 2.5 t/ha application along with chemical fertilizers ( Table 1 ).

Rice grain yield was significantly influenced by irrigation practices and soil amendments. Rice grain yield was significantly increased by AWD irrigation practices compared to conventional and aerobic irrigation practices. At Mohadebpur, the highest grain yield 6530 kg∙ha⁻¹ was recorded with 10.0 t/ha vermicompost, followed by 6450 kg∙ha⁻¹ and 6070 kg∙ha⁻¹ with 7.5 t/ha and 2.5 t/ha vermicompost amendments along with chemical fertilizers under conventional irrigation ( Table 1 ). Alternate wetting and drying (AWD) irrigation revealed the highest grain yield 6800 kg∙ha⁻¹ with 10.0 t/ha vermicompost amendment, followed by 6680 kg∙ha⁻¹ and 6200 kg∙ha⁻¹ with 7.5 t/ha and 2.5 t/ha vermicompost amendments. In aerobic irrigation practice, the highest grain yield was found 5650 kg∙ha⁻¹ at 10.0 t/ha VC application, followed by 5570 kg∙ha⁻¹ and 5180 kg∙ha⁻¹ with 7.5 t/ha and 2.5 t/ha VC amendments, respectively.

At Nachole experimental site, rice grain yield was significantly increased by AWD and aerobic irrigation practices. AWD irrigation revealed the highest grain yield, 6850 kg∙ha⁻¹ followed by 6700 kg∙ha⁻¹ and 6300 kg∙ha⁻¹ were recorded with 10 t/ha VC, 7.5 t/ha VC, and 2.5 t/ha VC applications, respectively. In aerobic irrigation, the highest grain yield 5460 kg∙ha⁻¹ was found with 10.0 t/ha VC application, followed by 5380 kg∙ha⁻¹ with 7.5 t/ha VC amendment and 5060 kg∙ha⁻¹ with 2.5 t/ha VC application. In conventional irrigation, the highest grain yield, 6430 kg∙ha⁻¹ was found with 10.0 t/ha VC application, followed by 6350 kg∙ha⁻¹ with 7.5 t/ha VC and 5960 kg∙ha⁻¹ with 2.5 t/ha VC application. The lower grain yield in conventional irrigated field compared to AWD irrigation may be due to waterlogged conditions throughout the growing season, which affected rice yield components and grain yield ( Table 1 ).

At Mohadebpur experimental site, the yield scaled CH₄ emission (GHGI) was significantly decreased with AWDI and aerobic irrigation practices compared to conventional irrigation practices. The highest yield scaled CH₄ emission (GHGI) from the conventional irrigated field plot was found 0.037 kg CH₄∙kg⁻¹ (in T2), followed by 0.029 (T3) and 0.027 kg CH₄∙ha⁻¹ (T4). In regards to AWD irrigation, the yield scaled CH₄ emission (GHGI) 0.028 kg CH₄∙kg⁻¹ yield was found with 2.5 t∙ha⁻¹ VC amendments (NPKS 100% + Phospho-gypsum 2.5 t∙ha⁻¹), followed by 0.023 kg CH₄∙kg⁻¹ yield and 0.020 kg CH₄∙kg⁻¹ yield) with VC amendments 7.5 - 10.0 t/ha. In case of aerobic irrigation, the highest yield scaled CH₄ emission was recorded 0.033 kg CH₄∙kg⁻¹ yield followed by 0.027 and 0.026 kg CH₄∙kg⁻¹ yield. At Nachole experimental site, the highest yield scaled CH₄ emission (GHGI) from the continuous irrigated plot was found in 0.032 kg CH₄∙kg⁻¹ (T2), followed by 0.027 and 0.025 kg∙ha⁻¹. Conversely, under AWD treatment, the highest scaled CH₄ emission (GHGI) was recorded 0.025 kg CH₄∙kg⁻¹) with VC amendment 2.5 t/ha followed by 0.020 and 0.019 kg CH₄∙kg⁻¹ yield. In general, AWD and aerobic irrigation practices reduced yield scaled CH₄ emissions (GHGI) significantly than that of conventional irrigated field plots ( Table 1 ).

The GWPs decreased significantly with AWDI and aerobic irrigation practice treatments compared to conventional irrigation practices. At Mohadebpur experimental site, the highest GWPs value 5425 kg CO₂-eq∙ha⁻¹ was found in conventional irrigated field with 2.5 t∙ha⁻¹ VC amendments (NPKS 100% RFD + Phospho-gypsum 2.5 t∙ha⁻¹), which decreased significantly under AWD and aerobic irrigations. The maximum decrease in GWPs were obtained by 20% - 25% and 22% - 27% with 10.0 t∙ha⁻¹ Vermicompost amendments (plus phospho-gypsum 2.5 t∙ha⁻¹ + 25% recommended N/ha + recommended PKS/ha) for AWD and aerobic irrigations, respectively.

At Nachole experimental site, the highest GWPs 4875 kg CO₂-eq∙ha⁻¹ was observed in conventional irrigated field plot with VC amendment 2.5 t/ha (T₂) followed by 4305 and 3810 kg CO₂-eq∙ha⁻¹ in VC amendment 7.5 t/ha (T3) and VC 10.0 t/ha amendment (T4). In case of AWD irrigation, the highest GWPs was found 4000 kg CO₂-eq∙ha⁻¹ in VC amendment 2.5 t/ha (T₂) followed by 3365 kg CO₂-eq∙ha⁻¹ (T3) and 3085 kg CO₂-eq∙ha⁻¹ T₄. Under aerobic irrigation treatments, the lowest GWP 3025 kg CO₂-eq∙ha⁻¹ was noticed in VC amended (10 t/ha) field plot (T₄) followed 3695 and 3315 kg CO₂-eq∙ha⁻¹. In general, higher GWPs were found in conventional irrigated field plots compared to AWD and aerobic irrigations.

At Mohadebpur experimental site, during boro rice cultivation, the required irrigation frequency was 16, average total volume of water was 30,807 m³∙ha⁻¹, average cost for irrigation was 15,100 Tk∙ha⁻¹ under conventional irrigation. In case of alternate wetting and drying (AWD), the irrigation frequency was 12, average total water volume was 22,800 m³∙ha⁻¹, average cost for irrigation was 11,850 Tk∙ha⁻¹, average water savings was 26.0% and the average irrigation cost saving was calculated as 3250 Tk∙ha⁻¹. For aerobic irrigation practice, the required irrigation frequency was 8, average total water volume was 18,838 m³∙ha⁻¹, average cost for irrigation was 9825 Tk∙ha⁻¹, average water saving was 39.0% and the average irrigation cost saving was 5275 Tk∙ha⁻¹ ( Table 1 ). At Nachole experimental site, the required no. of irrigation was recorded as 16, average water volume was 29,500 m³∙ha⁻¹, average cost for irrigation was 14,168 Tk∙ha⁻¹ under conventional irrigation. For AWD irrigation, the required no. of irrigation was 12, the average water volume was 22,275 m³∙ha⁻¹, the average cost for irrigation was 11,775 Tk∙ha⁻¹; average irrigation water savings was recorded 24.5%; and the average cost savings for irrigation was 3050 Tk∙ha⁻¹. In case of aerobic irrigation, the required number of irrigation was recorded 8, average water volume was 18,600 m³∙ha⁻¹, average cost for irrigation was 9750 Tk∙ha⁻¹, average irrigation water savings was 37.0% and the average cost savings for irrigation was 5075 Tk∙ha⁻¹ ( Table 1 ).

Water productivity value (WP) was increased with AWD and aerobic irrigation practices compared to conventional irrigated field plots at both locations. In addition, increasing levels of vermicompost amendments 7.5 - 10 t/ha also influenced towards higher grain yield and ultimately maximized water productivity. Alternate wetting and drying (AWD) irrigation revealed maximum water productivity value 0.270 - 0.297 kg∙m⁻³ and 0.295 - 0.297 kg∙m⁻³ with vermicompost amendments 7.5 - 10 t/ha at Mohadebpur and Nachole experimental field rice cultivation. Conversely, under aerobic irrigation practice, higher WPI value were recorded 0.285 - 0.295 kg∙m⁻³ and 0.288 - 0.291 kg∙m⁻³ with vermicompost 7.5 - 10 t/ha amendments compared to other field plots at both locations. In general, AWD and aerobic irrigation practices significantly increased the WPI with vermicompost 7.5 - 10 t/ha amendments ( Table 1 ).

Soil amendments with vermicompost and phosphogypsum increased soil porosity, SOC, T-N, soil pH, available phosphate, available sulfate, available silicate (SiO₂), and soluble iron oxides in the post-harvest soils ( Table 2 ). Comparatively higher soil porosity was observed in AWD and aerobic irrigated field plots amended with higher levels of vermicompost compared to conventional irrigation. Maximum soil porosity 54.5% - 56.7% and 54% - 55% were recorded with VC 10 t/ha plus PG 2.5 t/ha amendments under AWD and aerobic irrigation practices, respectively.

After rice harvest, the soil organic carbon contents were found to be 1.15% - 1.43%, 1.27% - 1.65% and 1.30% - 1.68% in conventional irrigation, AWD and aerobic irrigation practices, respectively ( Table 2 ). Comparatively, higher soil organic carbon contents were found in vermicompost (7.5 t∙ha⁻¹ - 10 t∙ha⁻¹) amended field plots under AWD and aerobic irrigated conditions. Total N contents also varied in post-harvest soil. The highest N content (0.19%) was found in Vermicompost amended soil (10 t∙ha⁻¹ with phospho-gypsum 2.5 t∙ha⁻¹) under AWD and aerobic irrigation practices, while the lowest N was detected (0.11%) in recommended fertilizer (NPKS 100%) applied field soil under conventional irrigation. The available P (23.5 - 31.5 ppm) and SiO₂ (67.5 - 71.3 ppm) contents increased significantly with increasing levels of vermicompost applications under AWD and aerobic irrigations. In addition, higher water soluble sulfate (27.5 - 35.5 ppm) and total dissolved iron contents (10.5 - 13.6 ppm) were detected in the VC 10.0 t/ha and PG 2.5t/ha amended field plots under AWD and aerobic irrigations

Bangladesh is one of the most climate change risk-vulnerable country. The country needs to produce more rice to meet up the food demand of the expanded population, where irrigated rice farming will play a major role. Although farmers generally prefer irrigated rice cultivation during dry boro season due to high production and yield performance of HYV rice genotypes, however, irrigated rice farming is a major source of CH₄ emission and high energy consumption, which eventually causes higher GWPs and cost of production. In fact, 20% - 30% of the rice production cost is incurred for irrigation only in case of irrigated rice, depending on soil type and mode of payment ( Alam et al., 2009 ). It has been reported that about 79% of the total cultivated area in Bangladesh is irrigated by groundwater, whereas the remaining is irrigated by surface water ( Qureshi et al., 2014 ). Unfortunately, the groundwater table became unstable and is declining due to climate change, thereby may affect badly on irrigated rice production badly. Considering the climate change and water crisis issues, efficient water-saving irrigation practices such as AWD and aerobic irrigation practices hold a vital role for sustainable water management and rice productivity. AWD is basically a water management system, which does not involve any extra investment, just using a perforated plastic pipe (PVC) or bamboo pipe, to measure the soil surface water layers. In this experiment, the total amount of irrigation water applied for boro season rice cultivation was estimated 29,500 - 30,800 m³∙ha⁻¹ (two seasons’ average) under conventional irrigation system, which was reduced by 24.5% - 26.0% and 37% - 39.0% through the AWD and aerobic irrigation practices, respectively. The higher water requirement in conventional irrigation methods may be due to higher irrigation frequencies (16), seepage and runoff, and higher evapotranspiration compared to AWD (12) and aerobic irrigation (8) practices. Hiya et al. (2020) reported that the total volume of irrigation water applied for Boro rice cultivation at BAU Farm was 19,430 m³∙ha^−1, which was decreased by 18.0, 16.0 and 13.0% through AWD irrigation at 20, 15 and 10 cm, respectively. It has been shown that one (01) ton of rice production requires approximately 2500 L of water with AWD; and 5000 L of water for conventional irrigation ( Bouman, 2009 ). Furthermore, the maximum saving of irrigation water was found to be 50% ( Peng et al., 2006 ) and 47% ( Oo et al., 2018 ) with aerobic and AWD irrigation, respectively. It has also been reported that AWD irrigation practices increased rice production by 5% - 15%, while decreasing CH₄ emission by 40% - 70%, saving irrigation water 30% - 60% ( J. Yang et al., 2016 ; Linquist, 2015 ), which may be due to periodic aerobic process and field dryness. It has also been reported that AWD irrigation reduces CH₄ emissions by 34% - 42% in paddy soils during the drying period ( Yang et al., 2025 ), while stimulating N₂O emissions via nitrification (aerobic) ( Tran et al., 2018 ) and denitrification (anaerobic) processes.

The irrigation water application cost was calculated 14,825 - 15,100 Tk∙ha⁻¹ for conventional irrigation systems, which was decreased to 11,775 - 11,850 Tk∙ha⁻¹, and 9750 - 9825 Tk∙ha⁻¹ for AWD and aerobic irrigation systems, respectively. Rahaman et al. (2022) reported that irrigation cost for boro rice (medium high land) cultivation was 11,226 Tk/ha, which was higher than that low lying main haor areas rice cultivation. Kashem and Rahman (2018) reported that irrigation cost for boro rice cultivation in Rajshahi region was Tk. 5178/ha through prepaid irrigation programme, which was Tk. 10,963/ha for private irrigation programme.

The experimental findings confirmed maximum rice yield under AWD irrigation method irrespective of soil amendments with vermicompost, phosphogypsum and NPKS fertilizers application in both locations. Considering the yield performance at both locations, AWD irrigation revealed higher rice grain yield by 4.0%, 4.5% and 5.3% compared to the grain yield performance in conventional irrigation amended with vermicompost at 2.5 t/ha plus PG 1.0 t/ha, 7.5 t/ha plus PG 2.0 t/ha and 10.0 t/ha plus PG 2.5 t/ha applications. This yield increment with vermicompost and phosphogypsum amendments following AWD irrigation could be due to improved soil redox status (Eh oxidative-reductive condition) and better soil porosity, which enhanced higher nutrients availability to rice plant, thereby contributing to maximum yield performance compared to other irrigation methods. Several studies have shown an increase in rice yield for AWDI and aerobic irrigation practices compared to conventional irrigation ( Zhang et al., 2009 ; Qin et al., 2010 ; Ye et al., 2013 ; Chu et al., 2015 ; Islam, 2021 ). It has also been reported that aerobic rice needs 30% - 51% less total water for land preparation, 32–88% higher crop productivity, 50% saving on labor ( Wang et al., 2002 ) and reduced GHG emission by 50% ( Weller et al., 2016 ) compared to Puddled transplanting rice ( Bouman et al., 2005 ). However, the lack of stable yield performance and dry direct-seeded adapted varieties for aerobic systems is a major limitation in achieving the maximum yield potential under water and resource-limited conditions. Ali et al. (2019) also reported that AWD irrigation showed better rice yield performance and reduced cumulative CH₄ emission, enhanced water savings and water productivity during the dry season boro rice cultivation in Bangladesh. It was also reported that moderate wetting and drying increased rice yield, decreased water use and CH₄ emissions ( Hiya et al. 2020 ; Yang et al., 2016 ). Ahmad et al. (2014) mentioned that boro water productivity varies within 0.95 - 1.35 kg/m³ in the north west region of Bangladesh. In our field experiments, water productivity (WP) value increased with AWD and aerobic irrigation practices compared to conventional irrigated field plots at both locations. Alternate wetting and drying (AWD) irrigation revealed maximum water productivity values of 0.270 - 0.297 kg∙m⁻³ and 0.277 - 0.295 kg∙m⁻³ with vermicompost amendments 2.5-7.5-10 t/ha at both experimental field rice cultivation. Conversely, under aerobic irrigation practice, higher WP values were recorded, 0.276 - 0.295 kg∙m⁻³ and 0.271 - 0.288 kg∙m⁻³ with vermicompost 7.5 - 10 t/ha amendments compared to other field plots at both locations. It has been reported that the water productivity of rice varied within 0.20 - 1.2 kg grain m⁻³ water under safe AWD with the threshold of −15 cm ( Bouman et al., 2007 ; Lampayan et al., 2004 ). Our water productivity value lies within the mentioned range, a bit low moderate productivity probably due to critical soil hydrological properties in the selected locations.

In this study, the highest cumulative CH₄ emission 221 kg∙ha⁻¹∙season⁻¹ was recorded in vermicompost (VC 2.5 t∙ha⁻¹) amended field plot (NPKS 100% RFD + PG 1.0 t∙ha⁻¹) under conventional irrigation; this decreased to 178.0 kg∙ha⁻¹ and 170.0 kg∙ha⁻¹ under AWD and aerobic irrigation practices, respectively. The increasing levels of vermicompost amendments 7.5 - 10.0 t∙ha⁻¹ and PG 2.0 - 2.5 t∙ha⁻¹ decreased cumulative CH₄ emissions by 11% - 17%, 13% - 23% and 13% - 21.7% under conventional irrigation, AWD and aerobic irrigation practices compared to Vermicompost (2.5 t∙ha⁻¹) amendments with NPKS 100% plus phospho-gypsum (1.0 t∙ha⁻¹). The contents of water soluble ${\text{SO}}_4^{2-}$ and total dissolved Fe increased significantly in the selected experimental plots due to the higher application rates of vermicompost (7.5 - 10.0 t/ha) and phospho-gypsum (2.0 - 2.5 t/ha). The lower CH₄ emissions under AWD irrigations may be due to increased aeration (54.5% - 56.3% soil porosity), stabilization of soil organic carbon, improved soil redox potential status (−18.6 - 29.7 mv), accumulation of free iron oxides (9.3 - 13.7 ppm) and sulfate ions (27.5 - 33.6 ppm), which acted as electron acceptors, thereby, reduced methanogens’ activity and increased methane oxidation. Vermicompost application increased the availability of nitrogen in soil mostly in the form of nitrate relative to ammonium due to better soil aeration ( Chatterjee et al., 2021 ), thereby improving soil redox status and eventually reducing CH₄ emissions. Other studies also revealed that AWD reduced cumulative CH₄ emissions by 32% ( Lakshani et al., 2023 ), 29% ( Pramono et al., 2024 ), 37% ( Matsuda et al., 2023 ), 49% ( Chidthaisong et al., 2018 ), 52% - 55% ( Anapalli et al., 2023 ) and 77% ( Echegaray-Cabrera et al., 2024 ) compared to continuous flooding. Phospho-gypsum is used in rice fields due to its low cost, easy availability and nutrient content, such as high content of Ca, Silicon dioxide (SiO₂) and especially sulfate in gypsum, acting as electron acceptors, thereby increasing the activity of sulfate-reducing bacteria over methanogens for the common substrates ( Hori et al., 1993 ). In our previous research trial in rice field, it has revealed that seasonal cumulative CH₄ flux was reduced by 25% - 27% and 32% - 38% under continuous and intermittent irrigations, respectively ( Ali et al., 2009 ). It has also been reported that intermittent irrigations significantly reduced total seasonal CH₄ emissions by 27% compared to conventional (124 kg CH₄/ha) irrigated rice paddy field ( Ali et al. 2013 ). Soil amendments with vermicompost and phosphogypsum in combination with chemical fertilizers increased soil organic C and total N, enhanced soluble sulfate and total dissolved iron Fe contents, thereby reducing cumulative CH₄ emissions in rice field, which is supported by our previous research findings ( Ali et al., 2009 ). It has also been reported that dry seasonal cumulative CH₄ emissions were decreased by 14.7%, 18.9% and 24.8% with biochar amendments at 15 t/ha, 20 t/ha and 30 t/ha respectively under conventional irrigation; while cumulative CH₄ emissions were reduced by 10.6%, 26% and 41.6% respectively, under AWD irrigation system ( Ali et al., 2021 ).

In our experimental sites, the maximum GWPs value 4875 - 5425 kg CO₂-eq∙ha⁻¹ was found in conventional irrigated field with 2.5 t∙ha⁻¹ VC amendments (NPKS 100% RFD + Phospho-gypsum 2.5 t∙ha⁻¹), which decreased significantly under AWD and aerobic irrigations. The maximum decrease in GWPs were obtained by 21.8% - 25% and 22% - 27% with 10.0 t∙ha⁻¹ Vermicompost amendments (plus phospho-gypsum 2.5 t∙ha⁻¹ + 25% recommended N/ha + recommended PKS/ha) for AWD and aerobic irrigations, respectively. Tran et al. (2018) also reported that global warming potentials (GWPs) of CH₄ and N₂O under AWDs were 26-29% lower than those under continuous flooding in Central Vietnam. AWD irrigation decreased total GWPs by 34% - 64% to those of continuous flooding in paddy fields of Central Taiwan region ( Yang et al., 2025 ).

In this study, seasonal cumulative CH₄ emissions were found positively correlated with GWPs and SOC, while negative correlations were recorded with grain yield, soil porosity, soil Eh, available P, available SiO₂, water soluble sulfate and iron oxides; being supported by Hiya et al. (2020) . Van Der Gon et al. (2002) also reported that rice grain yield was negatively correlated with seasonal CH₄ flux. The increased grain yield, water productivity, irrigation water volume and cost savings, mitigating cumulative CH₄ flux as well as GWPs may positively inspire rice growers to adopt AWD irrigation technique. At the same time, the Govt. Policy makers, Govt. Organizations and NGO should strengthen their coordination and linkage to implement the water savings AWD and aerobic irrigation technique for sustainable rice farming at other agro-ecological zones of the country and disseminate this water-saving AWD technology in other rice-growing countries.