The Ghod River is a tributary of the Bhima River, covering approximately 71,500 hectares in the Ambegaon, Junnar, Shirur, Parner, and Shrigonda blocks of Pune and Ahmednagar districts in Maharashtra ( Figure 1 ). The catchment area of this river receives an annual rainfall ranging between 900 and 1200 mm, while the lower basin, where the surveying was done, receives the annual rainfall of approximately 600–800 mm. Farmers in this area rely on water from the Ghod River and groundwater for agriculture. The selected area is one of the prominent contributors to onion production in Maharashtra (Kebebe, 2019).

The cropping pattern within the Ghod River basin area exhibits rich diversity due to variations in geography, soil type, rainfall, and the availability of irrigation sources. During the monsoon season, a variety of crops, including soybean, bajra, onion, groundnut, and mung bean, are commonly cultivated. Additionally, irrigated crops such as vegetables, flowers, and sugarcane thrive in this region. In areas where irrigation projects are established, sugarcane, onion, and various fruit crops dominate during the post-monsoon season. However, in instances where alternate irrigation sources are lacking, crops like sorghum and gram are grown, and their success is closely tied to rainfall and residual soil moisture levels. This cropping pattern is intricately linked with water availability, prevailing market trends, and proximity to markets (ITC, 2020). Furthermore, the well-developed dairy infrastructure in the area necessitates the cultivation of fodder crops tailored to the specific needs of the livestock industry. The Ghod River plays a crucial role in sustaining agriculture and livelihoods in the region. Its diverse cropping practices reflect the intricate balance between natural resources, market dynamics, and the needs of local communities.

To enhance the water-use efficiency and productivity in onion farming, a range of extension initiatives i.e. training programs, on-field demonstrations (OFDs), and farmer field schools were implemented. A total of 240 farmers (60 nos. from each block) undertook training to enhance their capacity to utilize improved water management technologies. The trained farmers’ fields were selected for conducting demonstrations to evaluate the impact of these improved technologies on onion production. To ensure ongoing support, socially integrated messaging applications were used to form groups, facilitating regular advisory services for participating farmers in the demonstration plots. Additionally, field visits were organized to monitor the progress of the farmers’ plots. Similarly, farmers had exposure visits to the ICAR-Directorate of Onion and Garlic Research in Pune, where they learned from established practices and research. The adoption of these strategies aimed at optimizing onion cultivation while conserving water resources and promoting sustainable agricultural practices.

The raised bed with drip irrigation system has gained prominence in various crop production systems due to its significant advantages in water and soil management (Gadge et al., 2022; Piri and Naserin, 2020; Gorantiwar et al., 1991). This system offers improved drainage for excess rainfall and creates favorable soil conditions. When combined with drip irrigation, one of the most efficient water management methods, studies have shown that it enhances both yield and onion quality (Gethe et al., 2006; Saudy and El–Metwally, 2019). Specifically, for onion cultivation, the recommended raised bed with drip irrigation system (RBWD) ( Figure 2 ) consists of a broad bed furrow (BBF) with dimensions of 15 cm height and 120 cm top width, along with a 45 cm furrow. Each BBF should be equipped with two drip laterals, approximately 16 mm in size, spaced at 60 cm intervals and featuring inbuilt emitters. The distance between these emitters should range from 30 to 50 cm, with a discharge flow rate of 4 liters per hour (Thangasamy and Singh, 2022). This system optimizes water use, promotes efficient onion growth, and contributes to sustainable agriculture practices.

A team of trained enumerators conducted face-to-face interviews with selected farmers using a pre-tested structured interview schedule. During these interviews, the enumerators explained the study’s purpose, obtained informed consent, and assured respondents of the confidentiality of their information. Data collection occurred during the late monsoon and post-monsoon seasons of 2022–23 to capture seasonal variations ( Figure 3 ). The study employed a purposive sampling approach. Four blocks were purposively selected as demonstrations had been carried out there, and seven villages from each block were further chosen purposively, resulting in 28 villages. In total, 480 farmers from four blocks participated in the study. The study focused on 240 demonstrated farmers who had adopted raised beds with drip irrigation from the total of 28 villages. An equal number of farmers served as controls and were selected from the same blocks. Uniformity among the farmers selected as treated and control was assured for their soil type, soil and irrigation conditions etc. to compare them. Further, any field level differences found were eliminated by using more averaging greater population sample and close monitoring for the good agriculture practices. For comparing irrigation conditions, all the farmers in the survey were closely monitored and the actual measurement of each drip and flood irrigated farm has been carried out at each farm level. Among these, 160 farmers were purposively selected for the late monsoon (October to November transplanting locally known as late Kharif planting) season, while 320 farmers were chosen for the post-monsoon season (December to January transplanting locally known as Rabi planting). The interview schedule covered various socioeconomic and demographic aspects of the participant farmers. To assess the field-level impact, the farmers’ onion production, yield and quality-related aspects were monitored over time. Additionally, water use assessment involved monitoring irrigation operations and interviewing related rationales. The study’s general outcomes are represented using simple descriptive statistics, focusing on parameters such as onion productivity, produce quality, and water use. Furthermore, the association between socioeconomic factors and the adoption decision of raised beds with drip irrigation was calculated using a binary regression model.

The logistic regression model was used to analyze the relationship between socio-economic variables and a binary outcome. Specifically, the outcome variable takes the value of ‘0’ when farmers practice flat bed with flood irrigation (FBWF) method, and ‘1’ when they use the improved method, raised bed with drip irrigation (RBWD). A comprehensive review of previous studies by Bravo-Ureta and Evenson (1994); Iráizoz et al. (2003) and Wilson et al., 2001) suggest that several factors influence farmers’ decisions regarding technology adoption. After conducting a contextual analysis of the study area, socioeconomic factors assumed to impact technical efficiency were selected. These factors, along with their expected signs, are described in Table 1 .

To predict the dependent variable i.e. the adoption decision of the RBWD system, the Logit model can be described as.

Where P n is the probability that a farmer adopted RBWD.

Where 1 − P n is the probability that a farmer did not adopt RBWD.

Taking logarithms on both sides it becomes-

Where,

α is the intercept,

β is the vector of the response coefficient.

e is the vector of random disturbance.

and

X_n is the set of explanatory variables.

So, the equation to be estimated can be written as.

Adoption=β₀+β₁ age+β₂ Gender+β₃ Education+β₄ Land holding+β₅ Experience+β₆ Income++β₇ Extension contact+β₈ Socialpart+β₉ Information source+β₁₀ Irrigation river+β₁₁ Irrigation tubewell.

Where,

Adoption is a binary variable for adoption and non-adoption of RBWD irrigation technology by onion growers,

β 1   β 11 are coefficients associated with each explanatory variable and e n is the error term.

The maximum likelihood estimation method was employed to obtain the parameter estimates. The model was assessed for goodness-of-fit using appropriate diagnostic tests (Hosmer-Lemeaqshow test, ROC curve). Additionally, multicollinearity among independent variables was also checked.

Irrigation water productivity in kg.m³ (IWP) serves as a valuable metric for assessing water use efficiency in agricultural systems. Unlike climatic factors, which have a lesser impact on IWP, agronomic practice factors significantly influence it (Lv et al., 2019). For the flat bed with flood irrigation (FBWF) method, the irrigation water applied during the entire onion production season is calculated based on the average number of irrigations, hours per irrigation, and the water discharge efficiency of the pump used by individual farmers. Whereas, for raised bed with drip irrigation (RBWD) adoption, the per-hour water discharge from a dripper and the average drip system use time throughout the season are considered (Guo et al., 2023; Yang et al., 2023b). The resulting water consumption, in terms of irrigation applied to the field, is then used to calculate irrigation water productivity as.

To compare the economics of RBWD and FBWF systems, detailed quantitative data on input quantities, prices, and production cost variables were systematically collected through a semi-structured survey, rigorously monitored by field representatives to ensure accuracy. The economic metrics, including the cost of cultivation and the Benefit-Cost ratio, were calculated in accordance with the Alagh Committee Report (https://cacp.da.gov.in/), one of the standard methodologies employed by the Commission for Agricultural Costs and Prices (CACP) for estimating crop production costs in India. Comprehensive details on the cost of cultivation are provided in Supplementary File 1 . The cost analysis was further disaggregated across multiple levels of costs such as—A1, A2, B1, B2, C1, and C2—encompassing key variables such as family labor and imputed rental values of land. To refine the comparison, per-hectare costs were converted to per-ton production costs, enabling a precise evaluation of the differential economic outcomes between the two irrigation systems.

In the pilot survey, adopters of Raised Bed with Drip Irrigation (RBWD) were examined to identify the key positive factors influencing their decision to adopt the technology. Additionally, challenges faced post-adoption were investigated to understand the dynamics following implementation. The major factors identified in the pilot survey were further evaluated in the main survey using a Likert scale ranging from 1 to 5, where 1 indicated ‘not relevant’ and 5 indicated ‘very relevant’. The collected scores were then normalized into a unity index by dividing the observed score for each factor by the maximum attainable score.

The socioeconomic profile of the farmer respondents revealed distinct characteristics between adopters of RBWD technology and non-adopters, particularly those employing the traditional FBWF method. Among the adopters, the middle age group constituted the largest portion of adopters, accounting for 59.2% followed by the older age group at 22.1% ( Table 2 ). In contrast, farmers practicing the traditional irrigation method were predominantly above the age of fifty (47.08%). Adopters tend to have higher education levels compared to non-adopters. Additionally, there was a slightly higher proportion of marginal and small landholder farmers in RBWD adopters. Non-adopters exhibited moderately lower levels of extension contact, social participation, and utilization of information sources. The majority of drip adopters (68.75%) resourced water from the dug wells. Whereas, for the non-adopters, the river (56.67%) was the primary source of irrigation water. These disparities in socioeconomic attributes and adoption of specific irrigation systems likely arise from a combination of various factors, including differential access to resources, educational backgrounds, and technological proficiency among the respective farmer groups.

Various socioeconomic characteristics influence farmers’ decision to adopt technology. A logistic regression analysis was conducted to explore the association between these characteristics and the adoption of the RBWD method which revealed several significant effects ( Table 3 ). Higher household income positively influenced the decision to adopt RBWD irrigation, indicating that farmers with greater financial resources are more likely to invest in new technologies. It also implicates that farmers with higher household incomes are better able to deal with the initial investment expenditures involved with establishing drip irrigation systems, which can be prohibitively expensive for low-income households. Findings are corroborating with the results of previous studies by (Nyathi et al., 2019; Narayanamoorthy, 2016). The results revealed a significantly positive effect of extension contacts (8.95, P<0.01) on adoption. Increased interaction with extension services is also associated with a higher RBWD adoption rate, highlighting the importance of access to information and support from extension agents. This also emphasizes the importance of access to information and technical assistance offered by extension agents in supporting technology adoption. Extension services serve to address the knowledge gap by educating farmers about the advantages, proper application and long-term benefits of drip irrigation systems, particularly raised bed methods in onion. Farmers who use these services more frequently are more likely to be informed and confident in implementing such advances, resulting in higher adoption rates. These results are consistent with findings from previous studies (Kebebe, 2019; Jogo et al., 2021; Nyathi et al., 2019; Chinseu et al., 2019). Conversely, larger land holdings had a negative impact on RBWD adoption. In the study area, scattered land holding also imposed restrictions in the management of drip for large land holder farmers. Additionally, this could be attributable to the increased managerial effort and cost of setting up and operating drip systems on such scattered farms. Managing the system across large fields may necessitate additional effort, extra monitoring and higher cost of inputs, discouraging farmers with larger land sizes from using this technology. The difficulties of scaling up the system across large fields could discourage these farmers, leading them to remain with traditional irrigation methods that require less extensive management. Surprisingly, the association between information source use and adoption was negative (-7.95, P<0.05). The negative association may arise from farmers’ reliance on informal sources (e.g., peers, traders, or local networks) that often lack accurate or practical guidance on RBWD, creating confusion and hesitation. Intensive information-seeking may also reflect uncertainty among non-adopters who, despite exploring multiple sources, ultimately avoid investment. In addition, limited adoption could be linked to gaps in the quality, credibility, or accessibility of formal channels such as extension services, training programs, and demonstrations. Further research is needed to comprehend this finding. Overall, the socioeconomic factors, including income, extension contact, land holding, and information sources, collectively shape the adoption of RBWD technology.

Farmers who own dug wells tend to adopt drip irrigation more frequently compared to those relying on tubewells, rivers or canals (-1.83, P<0.05). The abundance of water in rivers and canals often leads to less motivation for adopting water-saving measures among farmers using these sources. In contrast, farmers using dug wells as their irrigation source must carefully manage limited water resources, particularly during the post-monsoon season, which prompts their adoption of water-efficient technologies to optimize water usage. Water quality also plays a significant role in RBWD adoption, with water from rivers and canals being more turbid than water from dug wells. Thus, filtration efforts are necessary before using this water in micro-irrigation systems to prevent clogging and internal damage. The limited water availability in dug wells may modestly impact the adoption of drip irrigation for efficient use. Adoption of sustainable agricultural technologies, specifically irrigation technology, is influenced by various factors, including, technical support availability to the farmers, as the latter predominantly benefit by direct access to technical guidance and assistance. Moreover, technological, social, and institutional environments along with economic considerations affecting affordability and cost-effectiveness are crucial contextual factors that influence adoption decisions. The learning willpower of farmers shapes the continuous learning and adaptation capacity, contributing to sustained adoption (Chinseu et al., 2019; Wilson et al., 2001). Furthermore, technology traits, adaptation to local conditions, and the suitability of the technology delivery system play important roles in modulating long-term adoption (Jogo et al., 2021).

In the current survey, the fundamental positive factors influencing the adoption of drip irrigation methods were examined. The primary driving force behind the adoption of RBWD was its substantial productivity benefits (scored at 0.94), closely followed by the advantages of water savings (scored at 0.94; Figure 5 ). It is evident that the likelihood of technology adoption diminishes unless farmers perceive tangible advantages or anticipate enhanced utility from its integration (Bonabana-Wabbi, 2002). When there is no significant difference in outcomes between the two options, and the returns from alternative and conventional practices are similar, it becomes less likely that especially small-scale farmers adopt the new practice (Abara and Singh, 1993). The positive factor of water savings is particularly important because farmers must recognize the existence of a compelling water-related issue that demands an alternative solution. Thus, the adoption of RBWD proves the importance given to water-related challenges by farmers in the Ghod River basin.

The adoption of RBWD was driven by several factors, including its versatility for multiple crops (scored at 0.93) and the labor-saving aspect (scored at 0.93), educational status, extension contact, social participation of the grower along with use of information sources (Kebebe, 2019). This labor saving becomes crucially important when the farmers in the basin are facing labor shortages as the availability of alternate non-farm employment in nearer industrial and urban vicinities. Additionally, as the water permits, farmers in the area often cultivate short-seasonal vegetables and fruits like cucumbers and watermelons. In this short period, RBWD saves the additional tillage and bed preparation operations. Farmers use the RBWD beds after harvesting onions, which saves machine and human labor involvement. Interestingly, two unexpected considerations emerged from this survey regarding the adoption of drip irrigation. First, the need for protection from wild animal attacks (scored at 0.92). In the surveyed blocks, where sugarcane is the dominant cropping system, the presence of leopards posed a significant concern for farmers. Implementing drip irrigation helped minimize human exposure in the fields, providing a safety measure against such risks. Second, dealing with frequent electricity cut-offs during daytime operations (scored at 0.91) influenced the adoption of RBWD. The interrupted power supply forced farmers to irrigate their crops at night, and the drip irrigation system, where, automated techniques such as programmed scheduling, auto on-off functionality, and remote operation, proved highly beneficial in this scenario and further encouraged farmers to adopt RBWD technology.

The study also focused on the challenges encountered by farmers while adopting drip irrigation systems. These constraints shed light on the critical factors influencing farmers’ decisions not to adopt this technology. Farmers’ adoption choices are influenced by investment considerations (Caswell et al., 2001). Capital-intensive technologies are more accessible to wealthier farmers, limiting adoption among those with fewer resources (El-Osta and Morehart, 1999; Khanna, 2001). Higher initial investment or the cost of setting up drip irrigation systems is a significant barrier (rated at 0.92). This financial constraint affects adoption decisions, especially for farmers with limited resources. Insufficient credit availability (rated at 0.72) and lack of comprehensive operational knowledge about drip irrigation techniques (rated at 0.90) pose additional obstacles. Policy innovations can play important role to address this constraint. Successful models such as the Andhra Pradesh Micro Irrigation Project (APMIP) and Gujarat Green Revolution Company (GGRC) have streamlined subsidy disbursement, reduced bureaucratic hurdles, and provided technical support, thereby improving adoption rates. In contrast, poorly designed subsidy systems can inflate costs due to bureaucratic delays. Thus, to address these adoption barriers, requires a dual strategy: (i) improving farmers’ access to affordable credit and flexible subsidy mechanisms, such as direct benefit transfers and (ii) strengthening extension services through hands-on training, demonstrations, and farmer-to-farmer learning. Emphasizing the profitability and rapid payback period of drip systems can further encourage uptake ( Figure 6 ).

The constraint of fragmentation of land (0.85) primarily affected those who owned land scattered across different locations. Fragmentation was seen to complicate the implementation and management of drip irrigation systems and mechanized practices like constructing raised beds, thereby consuming more time and resources for these farmers. The logistical hurdles and additional efforts required for such fragmented setups hindered the broader adoption of RBWD among this group of farmers. Interestingly, the lack of nearby dealers did not pose a significant issue for most farmers (Zeng et al., 2018).