Buffalo husbandry is fundamental to India’s agricultural economy, considerably enhancing national food security, rural livelihoods, and agricultural GDP. India possesses 56.8% of the global buffalo population, with 109.85 million buffaloes recorded in the 20th Livestock Census. These animals produce approximately 100 million tonnes of milk annually, which accounts for 55% of the nation’s total milk output. This production contributes ₹1.5 trillion, or 7.4% of the agricultural GDP, to the Indian economy (DAHD, 2024; 20^th Livestock Census). The riverine buffalo (Bubalus bubalis) is vital in tropical climates, exhibiting enhanced susceptibility to heat, disease resistance, and feed conversion efficiency relative to exotic cattle breeds (Martínez-Burnes et al., 2024; Rout and Behera, 2021). In addition to milk production, buffaloes provide draught power and manure and function as living assets that boost financial resilience for 70 million smallholder households (Escarcha et al., 2020). The socioeconomic value of the sector is enhanced by its disproportionate influence on women’s empowerment; rural women account for 85% of buffalo management operations (Vijayalakshmy et al., 2023). This dual contribution—strengthening household income and enabling women’s economic participation—positions buffalo husbandry as a critical pathway for advancing Sustainable Development Goals, particularly SDG 1 (No Poverty), SDG 2 (Zero Hunger), and SDG 5 (Gender Equality), by combining economic viability with gender-inclusive and climate resilient livestock systems.

As of June 2025, the ICAR–National Bureau of Animal Genetic Resources (NBAGR) has officially recognized 20 registered buffalo breeds in India. These are broadly classified into five geographical groups: North (Murrah, Nili-Ravi, Gojri), West (Jaffarabadi, Mehsana, Surti, Banni), Central (Bhadawari, Chhattisgarhi, Nagpuri, Marathwadi, Pandharpuri, Dharwadi), East (Chilika, Kalahandi, Manda, Luit), and South (Toda, Bargur). Among them, the Murrah breed is considered the most prominent, predominantly reared in Haryana and Punjab. It is extensively utilized across the country due to its high milk yield, rich fat content, and remarkable adaptability to varied agro-climatic conditions. The adoption of scientific buffalo husbandry techniques offers a tested means to improve production, sustainability, and climate resilience. Studies show that using balanced feeding (Abebe et al., 2024), mineral supplementation (Gernand et al., 2016), and artificial insemination (AI) will boost milk yield by 42–58% while lowering methane emissions per litre by 18–22%. (Cardoso Consentini et al., 2021) While preventative healthcare measures can lower calf mortality from 23% to 9% (Wong et al., 2022), hygienic milking practices have been shown to reduce somatic cell counts by 65% and improve milk quality premiums by ₹3–5/litre (Vieira et al., 2021). Notwithstanding these apparent advantages, adoption of scientific management practices is shockingly low, with partial adoption for mineral supplementation and AI (Sharma, 2021; Dattatraya, 2024). Complicated behavioural barriers, including cognitive overload (Langer et al., 2024), habitual resistance (Ouellette and Wood, 1998), and mistrust in institutional advisories (Šūmane et al., 2018; Taylor and Van Grieken, 2015) that traditional extension approaches fail to address lead to this adoption paradox.

From Rogers’ (1962) Diffusion of Innovations, theoretical frameworks for technology adoption have developed into more complex models, such as the Unified Theory of Acceptance and Use of Technology (UTAUT) (Venkatesh et al., 2003). Although agricultural settings have benefited from the Technology Acceptance Model (TAM) (Davis et al., 1989) and Theory of Planned Behaviour (TPB), UTAUT’s thorough integration of performance expectancy, effort expectancy, social influence, and facilitating conditions offers better explanatory power (Rahi et al., 2019). Recent adaptations have included habit as a fifth construct, especially important for agricultural systems where routine procedures rule (Tamilmani et al., 2019). In agricultural settings, modified UTAUT frameworks have explained acceptance patterns for precision farming technologies (Nguyen et al., 2024; Larasati et al., 2024; Ronaghi and Forouharfar, 2020) and mobile-based advisory services (Kassem et al., 2021). The application of UTAUT remains conspicuously absent from the domain of livestock innovation, particularly in traditional systems such as buffalo husbandry. This omission is surprising given that buffalo farming in India is deeply embedded in socio-cultural routines, gendered labour patterns, and intergenerational practices—conditions that align well with UTAUT’s behavioural constructs. Moreover, buffalo husbandry entails both technological (e.g., scientific feeding, breeding, and healthcare) and behavioural innovations, where adoption is influenced not only by perceived utility but also by effort, habit, and social context. Therefore, this study represents a significant methodological advancement by applying UTAUT to a livestock based setting, offering empirical insights into how smallholder buffalo farmers perceive and adopt scientific husbandry practices. It demonstrates the model’s adaptability and emphasizes the importance of behaviourally informed frameworks in achieving sustainable livestock development. This study aims to solve several limits of previous studies on agricultural technology acceptance in India. First, most research uses binary adoption models—adopters against non-adopters—that miss the gradations and complexity of partial adoption (Nath et al., 2025; Seth et al., 2025; Feyissa et al., 2025). Second, the general emphasis on crop technologies—such as hybrid seeds or irrigation (Gaffney et al., 2016; Fadda et al., 2020)—has left a knowledge vacuum about innovations in livestock systems. Third, behavioural elements are sometimes studied separately without considering their relative importance or interacting effects (O’Donoghue et al., 2024). Fourth, the important function of habitual behaviours - which may be the most important obstacle to change in conventional livestock systems - is still mostly unexplored (Ouellette and Wood, 1998). Finally, few studies use strong theoretical models to examine the psychological processes guiding smallholder adoption choices methodically (Wauters and Mathijs, 2013; Singh et al., 2016). Our study aims to fill some research gaps by using a modified UTAUT framework to thoroughly examine behavioural elements influencing scientific buffalo husbandry adoption among Indian smallholders, We specifically want to investigate the relative relevance of performance expectancy, effort expectancy, social influence, and habitual practices in developing adoption intentions; also, we want to find context-specific barriers and facilitators that mediate these interactions; and finally, we want to create evidence based recommendations for developing more behaviourally-informed extension strategies.

This study offers application of the Unified Theory of Acceptance and Use of Technology (UTAUT) in the context of scientific buffalo husbandry, a field that has received limited attention in agricultural innovation literature. Although UTAUT has been widely used in healthcare and IT environments (Rouidi et al., 2022; Janet and Adetokunbo, 2024; Lee et al., 2025; Al-Momani and Ramayah, 2025; Kelkay et al., 2025), little is known about how well it can be adapted to livestock management, especially in smallholder systems. Our research addresses this gap by empirically examining the constructs of UTAUT (performance expectancy, effort expectancy, social influence, and habit) within a distinctive agricultural environment, where behavioural adoption is shaped by socioeconomic constraints, traditional traditions, and infrastructural constraints. The findings not only support UTAUT’s applicability but also question its universal assumptions, particularly regarding the importance of social influence, while emphasising habit as a significant predictor—a factor less stressed in the original UTAUT conceptions (Tamilmani et al., 2019).

Performance expectancy (PE), characterised as the extent to which farmers perceive that the adoption of scientific techniques will improve productivity, was identified as a substantial yet modest predictor (β=0.156, p=0.027) of behavioural intention. Farmers identified the benefits of innovations such as artificial insemination in enhancing milk production and herd quality, consistent with Rogers (2003) innovation-decision process, wherein relative advantage propels adoption. Nonetheless, the weaker effect size relative to habit indicates that perceived advantages alone are inadequate to surmount adoption obstacles. This reflects findings in African smallholder systems, where farmers recognised the advantages of hybrid seeds but were reluctant due to elevated prices and requisite skills (Quarshie et al., 2021). Our research expands this body of literature by illustrating that the influence of PE is mediated by contextual elements—such as market accessibility and input affordability—emphasizing the necessity for supplementary support mechanisms (e.g., subsidies, training) to convert awareness into action.

Effort expectancy (EE), indicating the perceived ease of implementing new practice, demonstrated a more significant impact (β=0.206, p=0.008) than performance expectancy (PE), so confirming Davis et al. (1989) Technology Acceptance Model (TAM), which asserts that usability precedes utility. Farmers preferred low-effort measures, such as vaccination scheduling, over complex technology like genomic selection, despite the latter’s potential for greater returns. This corresponds with Gigerenzer and Selten (2001) “effort-accuracy tradeoff,” wherein decision-makers experiencing cognitive burden (e.g., time-constrained smallholders) choose satisficing solutions over optimal ones. Our findings diverge from urban technology adoption research (Caffaro et al., 2020; Sugandini et al., 2018; Kamal et al., 2020), indicating that the role of ease of use declines with time, whereas agrarian environments demonstrate a sustained reliance on ease of use due to lower technological literacy and resource instability. This highlights the necessity of streamlining interventions—such as visual guides or audio-based applications—to minimise cognitive obstacles.

In contrast to the fundamental principles of UTAUT, social impact (SI) proved to be nonsignificant (β=0.136, p=0.133), hence contradicting Venkatesh et al. (2003) claim that peer pressure and expert opinions drive adoption. This disparity can be ascribed to three context dependent factors:

This finding requires a reassessment of UTAUT’s universality and indicates that social influence may be context-dependent, diminishing in significance under low-connectivity, high autonomy environments. The primary contribution of this study is the empirical validation of habit (HA) as the most significant predictor of behavioural intention (β = 0.420, p< 0.001), representing a substantial deviation from the traditional UTAUT paradigm. Mineral supplementation and clean milk production developed into standard practices that promoted the adoption of broader innovations, supporting Cilliers et al (2022) habit-intention nexus, which posits that repetitive actions in consistent environments become automated. This corresponds with Linder et al. (2022) findings in enabling pro-environmental habits.

This study suggests that farmers’ adoption of technology is driven more by entrenched habits than by rational evaluation. Technologies that align with existing routines, rather than disrupt them, tend to see greater uptake. For example, a study of subclinical mastitis diagnostics among dairy farmers in Punjab revealed that when simple farm-level tests—such as the Bromothymol Blue or California Mastitis Test—were integrated into regular milking routines, awareness and adoption significantly increased compared to standalone interventions (Nimbalkar et al., 2020). Similarly, regional field-level initiatives in Punjab, found that embedding health diagnostics within routine livestock health camps led to sustained engagement and higher farmer compliance (Singh et al., 2023). These examples reinforce the importance of designing technology adoption strategies that dovetail with habitual practices, such as pairing pregnancy testing with regular deworming schedules or embedding digital record-keeping within milk quality monitoring routines.

While the model successfully captured individual behavioural predictors, it did not account for broader institutional and macro-level barriers that critically shape adoption decisions. Structural issues—such as inconsistent access to veterinary services, limited availability of quality inputs, unstable milk procurement systems, and climate-induced fodder scarcity—pose systemic constraints that individual-level models often overlook. These elements can moderate the effects of behavioural constructs like performance expectancy and effort expectancy. Future studies should consider incorporating multi-level frameworks that blend micro behavioural data with macro-institutional variables to better reflect the lived realities of smallholder farmers. Comparative, longitudinal, and multi-level studies would help to reduce these limitations, increasing the application of the UTAUT model in agricultural innovation research and, hence, the practical relevance of results for policymakers and extension agencies.

Although previous research has explored UTAUT in agriculture, such as precision farming (Nguyen et al., 2024; Larasati et al., 2024; Ronaghi and Forouharfar, 2020), none have investigated the significance of habit in livestock contexts. Our structural equation modelling (SEM) lends empirical support to theoretical assertions regarding routine-driven adoption. The observed non-significance of Social Influence (SI) in predicting Behavioural Intention (BI) contrasts with the theoretical assumptions of the original Unified Theory of Acceptance and Use of Technology (UTAUT), which posits SI as a salient predictor of intention, particularly in collectivist societies or structured environments. However, this deviation may be explained by the unique socio-cultural and economic context of smallholder buffalo farmers in rural India. In contrast to corporate or institutional settings—where decisions are often shaped by peer dynamics, supervisory expectations, or organizational norms—rural farming communities tend to be characterized by a high degree of individual autonomy in decision-making. Farmers often base their choices on personal experience, perceived effectiveness, or trusted technical advice, rather than on the practices or opinions of peers. Consequently, the normative pressure assumed by the SI construct may be comparatively weak or functionally absent in these settings. Furthermore, the informal and fragmented nature of information dissemination in many rural areas may limit the extent to which social norms are perceived or internalized. The absence of strong peer networks, cooperatives, or group-based decision-making structures potentially reduces the influence of community adoption trends. In such a context, constructs such as performance expectancy, habit, and effort expectancy may exert a stronger influence on adoption behaviour than social influence.

This finding reinforces the notion that the applicability and predictive strength of UTAUT constructs are context-dependent, and that social influence may be less relevant in decentralized, self-reliant systems such as smallholder livestock farming. Accordingly, adaptation of theoretical models to reflect these contextual specificities—particularly institutional and structural constraints—is essential for accurately capturing behavioural dynamics in rural technology adoption research.

The results of this study have significant theoretical and practical implications for both innovation adoption literature and agricultural extension operations. This approach theoretically requires a recalibration of the UTAUT framework for agricultural contexts by prioritising habit as a fundamental construct and reducing the significance of social impact, marking a notable divergence from Venkatesh et al. (2003) original model. This modification corresponds with recent challenges of UTAUT’s universality (Xue et al., 2024; Blut et al., 2022; Bayag and Madimabe, 2024) and highlights the framework’s context-dependence, especially in environments characterised by traditional behaviours and diminished institutional confidence. The strong empirical support of habit’s priority (β=0.420, p<0.001) addresses a significant gap in behavioural agriculture research, which has historically conjectured about routine-driven adoption (Ouellette and Wood, 1998; Gardner et al., 2024) but has been devoid of quantitative evidence from cattle systems. These observations align with nudge theory (Thaler and Sunstein, 2021), illustrating that minor, routine-compatible interventions—such as including mineral supplements into daily feeding schedules—are more beneficial than isolated, high-effort technologies. This method leverages established cognitive pathways, so decreasing the activation energy necessary for behavioural change and lessening resistance to interruption. This divide is particularly pertinent for policy formulation, as top-down agricultural extensions frequently falter due to misaligned incentives. Our findings support habit-centric extension programmes that commence with the mapping of farmers’ established practices (e.g., milking routines, seasonal grazing patterns) prior to integrating innovations into these routines. For example, instructing veterinarians to include AI breeding techniques during regular health assessments—rather than as standalone workshops—could leverage normal interactions to enhance adoption. Likewise, low-effort toolkits, such as pictorial deworming calendars or voice-activated reminders for vaccination schedules, may alleviate the cognitive burden that weakens effort expectancy. These guidelines collectively transition the focus from technology driven to behaviour-first interventions, ensuring that scientific developments are not only theoretically superior but also realistically applicable within the limitations of smallholder systems.

This study offers critical insights into the multifaceted nature of technology adoption in smallholder buffalo husbandry systems. While confirming the prominent role of habitual practices in shaping adoption intentions, our analysis reveals a more complex behavioral landscape where performance expectancy and effort expectancy continue to play significant roles. The findings suggest that effective intervention strategies must move beyond singular approaches to embrace a holistic understanding of farmer decision-making. Technologies demonstrating clear productivity advantages require robust evidence-based demonstrations to establish their performance value, while complex innovations demand careful attention to usability through hands-on training and simplified implementation protocols. The non-significance of social influence challenges conventional extension paradigms, indicating a need for more individualized, farmer-centric support systems rather than reliance on peer diffusion mechanisms. These results collectively advocate for integrated intervention designs that simultaneously address habitual integration, proven benefits, and ease of implementation while remaining sensitive to individual farm contexts. Such multidimensional approaches promise to bridge the innovation-adoption gap more effectively than strategies focusing exclusively on any single behavioral determinant. The study ultimately provides a differentiated understanding and facilitating technology adoption that acknowledges the interplay of multiple behavioral factors in smallholder production systems, offering valuable guidance for both researchers and practitioners working to enhance livestock productivity through sustainable innovation adoption.

This work improves understanding of behavioural adoption in buffalo husbandry; yet, some constraints have to be acknowledged along with opportunities for more investigation. As purposive sampling was used, the findings may not be generalizable to all smallholder buffalo farmers in India. However, care was taken to include participants from diverse production systems to ensure a breadth of perspectives and practices.

The spatial focus limits the relevance to other agricultural environments with different socioeconomic or ecological contexts. To prove its cross-cultural relevance, subsequent research should assess the modified UTAUT model in different settings (e.g., smallholder farms in Africa or intensive systems in Europe). Second, depending too much on self-reported data could cause response bias, particularly in relation to socially desired activities like clean milking. Including objective adoption measures (such as technology use logs and agricultural output records) in future research would improve the results. The cross-sectional technique prevents the analysis of long-term behavioural modification. By means of longitudinal analysis of habit development and technology adoption patterns, one may find fundamental temporal dynamics, including the change from initial adoption to continuous use. The study mostly focused on personal farmer decisions, so it mostly ignored institutional barriers such as legislative frameworks, market access, and climate variability. Future models might combine macro-level elements with micro-level behavioural data to provide a more complete knowledge of adoption challenges. In the end, more modern technologies—such as precision livestock farming tools—introduce new adoption problems that call for study using this altered paradigm. Comparative, longitudinal, and multi-level studies would help to reduce these limitations, increasing the application of the UTAUT model in agricultural innovation research and, hence, the practical relevance of results for policymakers and extension agencies.