The large comparative field demonstration (over an acre per farmer) was conducted in four successive growing rabi (winter) seasons from 2020–2021 to 2023–2024 (4 years) at demonstration sites of farmer fields in Purnea (25⁰ 77’ N, 87⁰ 47′ E), and Katihar (25⁰ 55’ N, 87⁰ 56′ E), districts of Bihar state (25⁰ 96’ N, 85⁰ 27′ E), India. These project sites (PSs) are situated within the Climate-Resilient Agriculture (CRA) program, strategically located in the EIGPs along the Ganga River region in India. A total of 10 PS /villages were selected for both Purnea and Katihar districts (five sites/villages of each) (Figure 1). A total of 60 farmers (6 for each PS) were chosen for the study.

The PSs comprise Dogachhi, Dholbazza, Basantpur, Kullakhas, and Kullasundar in Purnia, and Basgarha, Mahinathpur, Barakhal, Musapur, and Sisiya in Katihar (Supplementary Table S1). The project sites have a humid subtropical climate with well-defined monsoon and dry seasons. Annual rainfall during the study period (2019–2024) ranged from 1,380 to 1,510 mm in Purnia and from 1,305 to 1,450 mm in Katihar, mainly in the range of June to September. The mean annual temperature ranged between 26.2°C and 27.9°C with seasonal variations impacting crop growth phases. In comparison with historical trends (1991–2020), the study period revealed an increase in temperature (by 0.5–1.2°C) and increased variability in rainfall, suggesting potential climatic changes impacting soil moisture and crop yield (Supplementary Figure S1).

The soils in the study area were predominantly alluvial and sandy loam, with variations observed across different sites. Bulk density ranged from 1.12 to 1.19 g/cm³, field capacity from 39.89 to 45.8%, and wilting point from 9.65 to 12.05%. Soil organic carbon varied between 0.21 and 0.32%, and pH from 6.48 to 6.9. Available N, P, and K ranged from 198.65 to 225.45 kg/ha, 13.45 to 16.9 kg/ha, and 116.78 to 140.1 kg/ha, respectively, indicating variations in soil fertility and moisture retention across project sites (Supplementary Table S2).

Statistical tests were conducted on the raised bed planting (RBP) and flatbed (FB) systems in terms of differences between different parameters using analysis of variance (ANOVA) and Duncan’s multiple range test (DMRT). Parameters were productivity, profitability, water use efficiency (WUE), water productivity (WP), and nutrient use efficiency (NUE). Evapotranspiration (ET) was calculated by the pan evaporation method, in which water evaporation was quantified from a standard Class A pan under open-field conditions. The observed pan evaporation (Epan) values were corrected by a pan coefficient (Kp) of 0.7 to 0.85, considering local climatic factors such as wind speed, relative humidity, and solar radiation. To estimate crop evapotranspiration (ETc), Kc values specific to maize were used for the corrected ET0 (reference evapotranspiration). Rather than taking a constant Kc = 0.85, we took stage-specific Kc values according to FAO-56 guidelines for maize (Allen et al., 1998 and Doorenbos and Pruitt, 1977):

Kc_initial = 0.35 (Emergence up to the early vegetative stage).

Kc_development = 0.75 (Vegetative to flowering stage).

Kc_mid-season = 1.15 (Flowering up to the grain-filling stage).

Kc_late-season = 0.60 (Maturity and drying stage).

The formula for water use efficiency (Equations 1–4) was used from Howell (2001), total water productivity from Zwart and Bastiaanssen (2004), and nutrient use efficiency from Fageria et al. (2010) and Dobermann (2007).

Data showed that productivity varied considerably between the raised bed planting (RBP) and flatbed (FB) systems, with RBP showing higher productivity across all 10 project sites (PSs) (Figure 4). The RBP system had a range of productivity from 9.28 to 10.53 t ha⁻¹. The highest productivity was observed at PS-3 with 10.53 ± 0.49 t ha⁻¹, whereas the lowest was recorded at PS-5 with 9.28 ± 0.80 t ha⁻¹, thus showing significant variation. These differences show the consistent benefit of the RBP system over FB in terms of higher productivity at different locations.

The productivity under the flatbed (FB) system varied considerably across the 10 project sites (PSs), ranging from 5.70 to 8.29 t ha⁻¹ (Figure 4). The highest productivity was recorded at PS-1 (8.29 ± 0.31 t ha⁻¹), while the lowest was recorded at PS-9 (5.70 ± 0.42 t ha⁻¹). In-between value situations existed for most others as well, with some considerable performances at PS-2 (8.12 ± 0.24 t ha⁻¹) and PS-3 (7.87 ± 0.72 t ha⁻¹). The RBP system experienced productivity advantages of 4.84 to 17.82% compared to the FB system, with the former being observed at PS-1, PS-3, and PS-8.

The average profitability across project sites under the RBP system varied from $1943 to $1998 ha⁻¹, with significant differences. The highest profitability was noted at PS-10 ($1998.83 ± 48 ha⁻¹), whereas the lowest was reported at PS-7 ($1943.25 ± 54.94 ha⁻¹). Other notable performances were reported at PS-8 ($1976.50 ± 40.66 ha⁻¹) and PS-5 ($1975.83 ± 44.79 ha⁻¹), indicating consistent economic advantages of the RBP system. On the other hand, the FB system had a profitability range of $1627–1857 ha⁻¹ across the project sites (Figure 5). The highest profitability was recorded at PS-10 ($1857.25 ± 60.84 ha⁻¹), representing a 14.11% increase compared to other sites, while PS-5 recorded the lowest value at $1627.50 ± 27.87 ha⁻¹. Intermediate profit gains were recorded at PS-3 (+12.20%, $1826.00 ± 83.46 ha⁻¹) and PS-6 (+8.04%, $1758.50 ± 95.56 ha⁻¹). These results highlight the significant economic advantages of using the RBP system over the FB system. The results showed that the RBP systems exhibited higher profitability compared to the FB system across all PS. Profitability increased in the RBP system from 4.64 to 20.43%, with significant improvements noted in PS-1, PS-3, and PS-8 (Figure 5).

Data revealed that the WUE ranged widely among the PSs and planting methods RBP and FB (Figure 6). WUE on the RBP system varied from 2.72% (PS-4) to 30.52% (PS-9). The highest (+30.52%, 4.79 ± 0.64 kg ha⁻¹ mm⁻¹) WUE was reported at PS-9, followed by PS-3 (+23.71%, 4.54 ± 0.32 kg ha⁻¹ mm⁻¹), PS-2 (+22.89%, 4.51 ± 0.13 kg ha⁻¹ mm⁻¹), PS-1 (+15.26%, 4.23 ± 0.48 kg ha⁻¹ mm⁻¹), PS-8 (+13.89%, 4.18 ± 0.19 kg ha⁻¹ mm⁻¹), PS-5 (+11.99%, 4.11 ± 0.53 kg ha⁻¹ mm⁻¹), PS-6 (+7.91%, 3.96 ± 0.59 kg ha⁻¹ mm⁻¹), PS-7 (+5.72%, 3.88 ± 0.21 kg ha⁻¹ mm⁻¹), and PS-10 (+2.72%, 3.77 ± 0.89 kg ha⁻¹ mm⁻¹), and the lowest (3.67 ± 0.42 kg ha⁻¹ mm⁻¹) was reported at PS-4. Meanwhile, in the case of the FB system, it ranges from 15.24 to 49.39% among the different PSs. The highest (+49.39%, 2.45 ± 0.10 kg ha⁻¹ mm⁻¹) WUE was observed at PS-1 (Figure 6), followed by PS-2 (+43.29%, 2.35 ± 0.18 kg ha⁻¹ mm⁻¹), PS-7 (+29.88%, 2.13 ± 0.30 kg ha⁻¹ mm⁻¹), PS-6 (+25.61%, 2.06 ± 0.32 kg ha⁻¹ mm⁻¹), PS-3 (+25.61%, 2.06 ± 0.25 kg ha⁻¹ mm⁻¹), PS-8 (+24.39%, 2.04 ± 0.11 kg ha⁻¹ mm⁻¹), PS-5 (+21.95%, 2.00 ± 0.36 kg ha⁻¹ mm⁻¹), PS-10 (+20.12%, 1.97 ± 0.07 kg ha⁻¹ mm⁻¹), and PS-4 (+15.24%, 1.89 ± 0.18 kg ha⁻¹ mm⁻¹), and the lowest was 1.64 ± 0.18 kg ha⁻¹ mm⁻¹ at PS-9. This trend continued for successive years, and in all PSs and years, the RBP system exhibited better WUE.

Water productivity (WP) varied significantly between the project sites (PSs) and planting techniques (Table 1). Under the RBP system, WP ranged from 3.88 to 4.58 kg ha⁻¹ mm⁻¹, with the highest value observed at PS-9 (+17.85%, 4.58 ± 0.94 kg ha⁻¹ mm⁻¹). This was closely followed by PS-3 (+16.84%, 4.54 ± 0.27 kg ha⁻¹ mm⁻¹) and PS-2 (+15.63%, 4.50 ± 0.23 kg ha⁻¹ mm⁻¹). Conversely, the lowest WP under RBP was reported at PS-5 (3.88 ± 0.64 kg ha⁻¹ mm⁻¹). These findings indicated substantial gains in water productivity with RBP, particularly at high-performing sites such as PS-9 and PS-3.

For the FB system, WP exhibited a broad range of variability, spanning from 1.57 to 2.41 kg ha⁻¹ mm⁻¹ across the sites (Table 1). PS-1 recorded the maximum WP (+53.50%, 2.41 ± 0.12 kg ha⁻¹ mm⁻¹), followed by PS-2 (+39.49%, 2.19 ± 0.10 kg ha⁻¹ mm⁻¹) and PS-7 (+33.76%, 2.09 ± 0.18 kg ha⁻¹ mm⁻¹). The lowest WP in the FB system was observed at PS-9 (1.57 ± 0.24 kg ha⁻¹ mm⁻¹), highlighting the significant advantage of RBP over FB in optimizing water productivity, especially at low-performing sites.

Nitrogen use efficiency (NUE) varied significantly across the project sites (PSs) and planting techniques (Figure 7). The NUE under the RBP system ranged from 68.45 to 75.83 kg grain kg⁻¹ N. The highest NUE was recorded at PS-9 (+10.74%, 75.83 ± 8.59 kg grain kg⁻¹ N), followed by PS-3 (+8.94%, 74.63 ± 5.36 kg grain kg⁻¹ N) and PS-1 (+7.38%, 73.46 ± 6.12 kg grain kg⁻¹ N). In contrast, the lowest NUE was observed at PS-4 (68.45 ± 5.74 kg grain kg⁻¹ N). These results highlight the efficiency of the RBP system, especially at the better-performing sites such as PS-9 and PS-3. For the FB system, NUE values varied between 42.89 and 60.69 kg grain kg⁻¹ N across the project sites (Figure 7). The highest NUE was documented at PS-1 + 41.48%, with 60.69 ± 1.81 kg grain kg⁻¹ N, followed by PS-2 + 36.72%, with 58.62 ± 1.94 kg grain kg⁻¹ N, and PS-3 + 30.47%, with 55.97 ± 2.13 kg grain kg⁻¹ N, respectively. At the bottom end, the minimum NUE was documented at PS-9 with 42.89 ± 4.63 kg grain kg⁻¹ N; this marks a significant variance in nutritional use efficiency among RBP and FB systems. The consistently higher NUE under RBP demonstrates its effectiveness in optimizing nitrogen utilization.

Phosphorus use efficiency (PUE) varied significantly across project sites (PSs) and planting techniques, with the RBP system showing higher values overall (Figure 8). Under the RBP system, PUE ranged from 164.79 to 188.66 kg grain kg⁻¹ P. The highest efficiency was observed at PS-2 (+14.48%, 188.66 ± 13.43 kg grain kg⁻¹ P), closely followed by PS-9 (+14.43%, 188.60 ± 14.12 kg grain kg⁻¹ P) and PS-10 (+11.90%, 184.44 ± 15.22 kg grain kg⁻¹ P). On the lower end, PS-6 recorded the least PUE (164.79 ± 9.91 kg grain kg⁻¹ P), indicating variability across sites. These results highlight the consistent advantage of RBP, particularly at high-performing locations such as PS-2 and PS-9.

In contrast, the FB system exhibited lower PUE values, ranging from 103.69 to 148.89 kg grain kg⁻¹ P across the sites (Figure 8). PS-1 achieved the highest PUE (+43.63%, 148.89 ± 11.75 kg grain kg⁻¹ P), followed by PS-3 (+38.80%, 143.89 ± 5.89 kg grain kg⁻¹ P) and PS-2 (+37.79%, 142.86 ± 6.88 kg grain kg⁻¹ P). The lowest efficiency was recorded at PS-9 (103.69 ± 3.55 kg grain kg⁻¹ P), demonstrating a significant disparity between the best- and worst-performing sites. Overall, the RBP system outperformed FB, reinforcing its effectiveness in optimizing phosphorus utilization across diverse conditions.

Potassium use efficiency (KUE) exhibited significant variability across project sites (PSs) and planting techniques (Figure 9). In the RBP system, KUE varied from 210.56 to 240.99 kg grain kg⁻¹ K. The highest value was recorded at PS-9 with +14.46% (240.99 ± 13.81 kg grain kg⁻¹ K), followed closely by PS-2 with +12.58% (237.06 ± 12.50 kg grain kg⁻¹ K) and PS-10 with +11.93% (235.68 ± 15.09 kg grain kg⁻¹ K). At the other end, PS-6 produced the lowest KUE values (210.56 ± 15.01 kg grain kg⁻¹ K), suggesting site-specific variability in potassium utilization efficiency. These results indicated that RBP was better at harnessing maximum KUE, especially at high-performing sites such as PS-9 and PS-2.

The FB system produced comparatively lower KUE values in the range of 109.03 to 181.81 kg grain kg⁻¹ K for different sites (Figure 9). The highest value was found at PS-3 (+66.78%, 181.81 ± 7.96 kg grain kg⁻¹ K), followed by PS-1 (+61.97%, 176.59 ± 6.24 kg grain kg⁻¹ K) and PS-5 (+61.25%, 175.82 ± 7.85 kg grain kg⁻¹ K). The lowest value was reported at PS-9 (109.03 ± 3.62 kg grain kg⁻¹ K), which reflects a drastic difference in the performance of KUE among the sites. Overall, the RBP system outperformed FB at all locations and demonstrated its effectiveness in improving potassium utilization.

The RBP system demonstrated higher productivity than the FB system by enhancing nutrient use efficiency and supporting optimal plant density, leading to improved crop growth and yield potential. RBP demonstrated superior yield performance across all production systems, with the highest and most stable productivity observed at certain sites. In contrast, the FB system showed lower and more variable yields. These findings highlight the effectiveness of RBP in enhancing maize productivity and emphasize the influence of site-specific factors on optimizing yield outcomes. The increase in production can be attributed to various elements that are inherent (+ aeration, + drainage ~ temperature) in the RBP system. The RBP system is particularly beneficial in high and intensive rainfall areas (Ahmad et al., 2020; Jaseem et al., 2024). This system also improves (10–35%) productivity and soil sustainability (Chauhdary et al., 2019). In addition, the RBP system improves germination, plant growth, and development (Singh et al., 2023). RBP system improves the loading resistance and optimizes the yield (Pooniya and Sharma, 2023; Jat et al., 2020; Du et al., 2021b). The RBP system showed its superiority to the FB system.

The results showed that the RBP system significantly improved both productivity and resource use efficiency compared to the FB system. The profitability of RBP was much higher than the FB, with a spread of improvements from 15 to 42% across various project sites. PS-5 exhibited the highest profitability under RBP, and PS-10 had the highest total profitability. Instead, the FB system was characterized by a relatively lower profit margin with a narrower scope of increases at all sites. These results indicate that in addition to increasing crop yields, the RBP system provides even better economic benefits (Bakhsh et al., 2018), especially on selected sites, and therefore deserves more serious consideration for application in sustainable farming (Yigezu et al., 2021). Ahmad et al. (2020) reported that the RBP system has a favorable condition for root growth and development. RBP system also inputs use efficiency (water 1,244% and nutrients 18–34%) at different soil and climatic conditions. Pooniya and Sharma (2023) and Chauhdary et al. (2019) reported that it improved soil sustainability and crop productivity. It also improves the soil biological health of different cropping systems (Sinha and Kumar, 2023).

WUE and WP vary across planting techniques and production systems. The RBP method improved WUE, highlighting its role in water conservation (Supplementary Figure S2). While some systems benefited more, others showed lower efficiency, reflecting management influences (Chauhdary et al., 2019). By means of improved infiltration and lower evaporation, the RBP approach guarantees more efficient water use and less waterlogging (Jaseem et al., 2024).

The RBP system improves water use efficiency, is more effective with limited inputs, and supports soil sustainability in high moisture conditions (Singh et al., 2023). Chauhdary et al. (2019) and Connor et al. (2023) reported that this system saved more water (40–50%) than the FB system.

NUE, PUE, and KUE varied among the tested planting systems (i.e., RBP system and FB system). The RBP system showed significantly higher efficiency compared to the FB system. NUE demonstrated notable differences, with the RBP system achieving higher efficiency. Similarly, PUE and KUE were also higher in the RBP system, indicating improved phosphorus and potassium utilization. These findings emphasize the role of RBP in enhancing NUE, contributing to improved resource utilization and sustainable crop production. The adoption of the RBP system improves NUE (25–30%) than the FB system (Ahmad et al., 2020). Adoption of the RBP system improves NUE and reduces soil loss (Pooniya and Sharma, 2023; Chauhdary et al., 2019). The RBP system also helps to improve biological health, contributes to the nutrient mineralization process, and improves productivity and profitability (Sinha and Kumar, 2023). Climate conditions during the experiment also affected WUE and NUE (Supplementary Figure S3). Greater temperatures in 2023 and 2024 would probably have raised evapotranspiration levels (Lahlali et al., 2024), impacting soil moisture balance and nutrient mobility. Similarly, changes in rainfall patterns impact irrigation requirements. However, surplus rainfall in 2021 would have resulted in nutrient leaching losses, while deficit rainfall in 2022 and 2024 necessitated greater dependence on supplemental irrigation to sustain optimal plant growth. Soil texture contributed significantly to the efficiency of water retention at the study sites (Bhushan et al., 2007). Sandy loams in some of the sites showed quicker drainage and needed more regular irrigation adjustments, whereas loamy alluvial stored water longer and enhanced the water productivity ratio. These emphasize the need for climate-responsive management of nutrients and irrigation to achieve maximum resource use efficiency (Katerji and Mastrorilli, 2009) in dynamic agro-ecological settings.