Determinants of Climate-Smart Agriculture Adoption Among Rice Farmers: Enhancing Sustainability

Abstract

The use of conventional farming methods, excessive reliance on fertilizers and inputs, and abrupt shifts in climate have raised significant concerns regarding global agricultural production, particularly in developing countries like Nepal. Agriculture products such as rice hold significant importance in Nepal’s agriculture and economy, serving as a staple food and a crucial source of livelihood for its population. Sustainable cultivation and enhancing productivity are imperative for ensuring food security and economic stability in the country. Adoption of climate-smart agriculture (CSA) practices can minimize detrimental effects, promote sustainability, and enhance resilience towards climate change. We surveyed 200 farmers across four municipalities in the Chitwan District of Nepal to explore the prevalence and socio-economic drivers of the adoption of CSA practices, which include stress-tolerant varieties, efficient water management, and diversified cropping, among others. The results revealed that the adoption of pest-resistant plant varieties was a common CSA practice in the study area. Logistic regression results revealed that the adoption of CSA practices increases with an increase in the education of farmers and membership of climate-related organizations. Similarly, the adoption of CSA practices is negatively associated with an increase in farm size, farmers’ farming experience, and their access to credit facilities. Short-term courses and training could be initiated as a complement to formal education to maximize the adoption of CSA practices. Similarly, climate and farmer-related organizations should be further strengthened to maximize their capacity to facilitate more farmers and provide need-based, timely information flow. This study highlights the potential of CSA to promote sustainability and enhance resilience to climate change, but also identifies barriers such as credit access and the need for tailored policy interventions. Our findings contribute to understanding the dynamics of CSA adoption in vulnerable agricultural settings and can guide future strategies to promote sustainability and climate resilience in smallholder farming communities in developing countries.

1. Introduction

Rice is a primary staple crop that feeds more than half of the world’s population and provides livelihoods for millions of smallholder farmers globally [1,2,3,4]. In Asia, where over 90% of the world’s rice is grown and consumed, rice is a major source of nutrition and is deeply intertwined with culture and traditions [4,5,6]. However, despite rice’s importance for global food security, climate change poses a significant threat to sustainable rice production in the coming decades [7]. Climate change events like shifting weather patterns and rising temperatures will impact rice yields, especially in vulnerable tropical developing countries across Asia and sub-Saharan Africa [8,9]. Out of the top five most “climate-vulnerable nations” in the world, Nepal ranks in the fourth position [10].

Nepal is a South Asian nation where rice is the predominant staple crop, cultivated across diverse agro-ecological regions from the Terai plains to the Himalayan high hills [11]. In Nepal, rice is the primary cereal crop in terms of production and cultivated areas. It holds an immense role in the economy of the country and food security [12].

Rice contributes over 20% to Nepal’s agricultural GDP and provides about 40% of the daily calorie intake for the population [11,13,14]. Given rice’s crucial role in Nepal’s food security, livelihoods, and cultural identity, climate change impacts pose major concerns [15]. Evidence suggests Nepal is already experiencing climate change through erratic monsoons, extreme temperatures, prolonged droughts, and unseasonal extreme rainfall [8,15,16,17,18,19]. Such events can significantly hamper rice yields, as over 60% of Nepal’s rice area relies on rain-fed production [14,20]. Declining rice harvests due to climate change will threaten Nepal’s smallholder farmers’ income and subsistence needs. To address these challenges, climate-smart agriculture (CSA) or sustainable agriculture has emerged as an integrative approach that aims to sustainably increase productivity, strengthen farmers’ climate resilience, and reduce greenhouse gas emissions from agriculture [21,22,23,24,25]. CSA encompasses a wide range of practices tailored to local contexts, including stress-tolerant varieties, efficient water management, minimum tillage, diversified cropping, agroforestry, and many others [23,26,27,28,29,30]. Evidence from across Asia and Africa indicates that CSA practices can effectively buffer climate risks in rice systems and even increase yields and incomes compared with conventional methods [31,32,33,34,35,36,37,38]. However, the adoption of CSA practices remains limited, especially among smallholder farmers with constraints like poor access to information, credit, inputs, and markets [14,22,39]. Moreover, farmers’ risk perceptions, education, resource endowments, collective action, and other socioeconomic factors influence their adoption decisions [40,41]. Therefore, understanding the multi-scale drivers of CSA adoption is critical to promoting climate-resilient farming.

This study examines CSA adoption trends and dynamics among rice farmers in the Chitwan district, one of Nepal’s major rice-producing regions. Using survey data from 200 farmers across four municipalities, it investigates the prevalence of different CSA practices and analyzes socioeconomic determinants influencing adoption. The findings contribute empirical insights on CSA adoption patterns from a climate-vulnerable context dominated by smallholder, subsistence farmers.

2. Materials and Methods

We conducted this research in the Chitwan district of Bagmati Province of Nepal (Figure 1). It is the leading rice-producing district with sub-tropical climatic conditions which are suitable for rice cultivation. Situated at latitudinal and longitudinal coordinates of 27°40′59.99″ N and 84°24′59.99″ E, Chitwan spans a diverse elevation range from 144 to 1947 m above sea level. Geographically, the Chitwan district is divided into three climatic zones, each influencing agricultural practices in the region. The Lower Tropical Climate Zone, covering 58.2% of the district, is characterized by warm temperatures and is highly suitable for rice cultivation and other crops due to its favorable climate. The Upper Tropical Climate Zone, which accounts for 32.6% of the area, experiences slightly cooler temperatures but remains favorable for agriculture, particularly for growing rice during the spring and monsoon seasons. Lastly, the Sub-Tropical Climate Zone covers 6.7% of Chitwan and features cooler temperatures and higher elevations, supporting a variety of cropping systems despite its smaller geographical area. These diverse climatic zones contribute to the region’s agricultural productivity and the adoption of different farming practices. The Ratnanagar, Rapti, Khairahani, and Kalika municipalities were purposively selected for the study as these are the major regions for rice production within the district, based on the number of rice growers and the field area under rice cultivation.

2.1. Data Collection and Analysis

We employed multistage, purposive, and random sampling to select districts, municipalities, and farmers. The farmers for the survey were randomly selected for the study. We selected 200 respondents from four different municipalities, 50 from each municipality, representing at least 10% of the total farmer’s population in each municipality. We chose this sampling frame to maintain consistency and statistical power across the municipalities, facilitating a comparison of adoption rates and factors influencing CSA practices. Primary data were collected from farm households using a pretested semi-structured interview questionnaire. We conducted two focus group discussions in each of the municipalities with a group of farmers to capture information based on consensus and verify individual interview responses with the survey. Before conducting the survey, all respondents provided verbal consent, with assurances that their data would remain confidential and anonymized throughout the research process. To comply with ethical standards, only researchers directly involved in the study were granted access to raw survey data. Additionally, data will only be shared for legitimate academic purposes to protect participant’s privacy. Secondary data and other relevant information were collected from governmental and non-governmental institutions’ bulletins, books, and publications.

2.2. Variables in Model and Data Analysis

We employed a logit model to determine the factors affecting the adoption of CSA practices among rice farmers. The CSA practices being considered in this study include a range of techniques aimed at improving resilience to climate change while boosting productivity (Supplementary Materials). These practices include pest-resistant plant varieties, drought-tolerant crop varieties, crop residue mulching, water management and harvesting, zero-tillage or minimum tillage, composting, and planting legumes among crops. The study also considered practices like integrated nutrient management and riparian vegetation conservation. Some of these have been explicitly adopted as part of modern CSA initiatives, such as pest-resistant varieties, water management systems, and zero-tillage practices, specifically promoted in response to climate variability and resource conservation goals. On the other hand, traditional practices such as composting and crop residue mulching have been in use for a long time by farmers, but they are now recognized as contributing to CSA due to their role in enhancing soil health and water retention. A logit model is a statistical model that estimates the probability of an event occurring by expressing the logarithm of the chances for the occurrence as a linear combination of one or more independent variables. The logit model is preferred to the conventional linear regression models in analyzing the factors influencing the adoption of CSA practices because the parameter estimates from the logit model are asymptomatically consistent and efficient. The conditional probability of adopting CSA technology lies between zero (0) and one (1). The logit model offers advantages in estimating the non-linear relationships between the probability of adoption of CSA practices and the explanatory variables [42]. While dealing with the extreme probabilities in cases of some agricultural practices, the logit model is preferred because it constrains the predicted probabilities to the [0, 1] range, providing consistent and more appropriate results. Moreover, several similar studies have widely used the logit model as a robust methodological approach. For instance, [43] used the logit model to study the determinants of CSA adoption in Pakistan and [44] studied CSA adoption in southern Africa using the logit model. Ref. [45] used a logit model to find the factors influencing farmers’ adoption of CSA technologies in Vietnam. The logit model is structured as

$$Z_i=l_n\left[{\displaystyle \frac{P_i}{1-P_i}}\right]=a+b_1{X}_1+b_2{X}_2+\dots \dots ..+b_{13}{X}_{13}+U$$

where $P_i$ = the probability of adoption and non-adoption of one of the CSA practices:

• $P_i$ = 1 = adoption;
• $P_i$ = 0 = non-adoption.
• The dependent variable is $Z_i$:
• $Z_i$ = Probability of adoption of one of the CSA practices.
• The independent variables are as follows:
• $X_1$ = Age of the household head (continuous);
• $X_2$ = Gender of the household head (categorical);
• $X_3$= Family member involvement (continuous);
• $X_4$ = Education of the household head (continuous);
• $X_5$ = Income (continuous);
• $X_6$ = Off-farm income (categorical);
• $X_7$ = Membership in climate-related organization (categorical);
• $X_8$= Membership in farmers’ organization (categorical);
• $X_9$ = Leased land (categorical);
• $X_{10}$ = Farming experience (continuous);
• $X_{11}$ = Farm size (continuous);
• $X_{12}$ = Access to extension (categorical);
• $X_{13}$ = Access to credit (categorical);
• a = Intercept;
• $b_1$ to $b_{13}$ = Regression coefficients of the dependent variables;
• U = Error term.

Before running the model, we performed a multicollinearity test on the interactions between the selected variables. The recorded data were first coded, tabulated, and analyzed using Microsoft Excel 2019, and further analysis was carried out with the help of IBM SPSS Statistics V.25.0 and Stata/SE 12.1.

3. Results

Our results revealed that around 72% of respondents had adopted at least one type of climate-smart agriculture (CSA) practice (

Table 1. Descriptive statistics for the independent variables used in the analysis to assess factors influencing farmers’ adoption of climate-smart agriculture technologies.

Variables	Description	Mean	S.D.
Age	Age of the respondent (years)	49.51	14.22
Gender	Gender of the respondent (=1 if male, 0 female)	0.51	0.50
Involved members	Number of family members engaged in rice farming	2.91	1.57
Education	Formal education of the respondent (years)	5.7	4.73
Income	Annual income of household (USD)	2935.50	3363.93
Off-farm	=1 if the respondent has an off-farm income source, 0 otherwise	0.35	0.47
Climate member	=1 if the respondent has climate organization membership, 0 otherwise	0.22	0.42
Group member	=1 if the respondent has farmer organization membership, 0 otherwise	0.55	0.49
Lease	=1 if the respondent has leased in land for rice farming, 0 otherwise	0.19	0.39
Exp	Rice farming experience (years)	25.61	15.48
Farm size	Land used for rice farming (kattha)	16.46	13.26
Extension	=1 if the respondent has regular contact with the extension worker, 0 otherwise	0.15	0.36
Access to credit	=1 if the respondent has access to a credit facility, 0 otherwise	0.69	0.46
Adoption	=1 if the respondent has adopted at least one CSA practice, 0 otherwise	0.72	0.45

Note: 1 ha = 29.58 kattha.

). The demographic profile revealed an experienced cohort of farmers, with an average age of 49.51 years and a male majority of 51%, reflecting gender dynamics in Nepal’s agricultural sector. Respondents possessed an average of 5.7 years of formal education and 25.61 years of rice farming experience, reflecting the importance of traditional knowledge. The results also highlighted the familial involvement in agriculture, with an average of 2.91 members per household actively engaged in farming activities. This suggests a ready labor force despite the challenges posed by an average farm size of 0.013 hectares and 19% leasing land.

Furthermore, 35% of respondents reported that their households rely on alternative income sources outside farming, supporting the narrative on the significance of income diversification in rural settings. The data also show strong community ties, with 55% of respondents being members of agricultural organizations or cooperatives and 22% belonging to climate-focused groups, illustrating the communal approach to farming prevalent in Nepal. Additionally, only 15% of the respondents reported regular contact with extension workers facilitating knowledge dissemination, while more than two-thirds (69%) had access to credit facilities.

The logistic regression analysis explored the factors influencing the adoption of climate-smart agriculture (CSA) technologies among farmers in Central Nepal. With a sample size of 200, the model provides insightful findings into the determinants of CSA adoption, evidenced by an LR chi² (13) of 46.98 and a highly significant Prob > chi² of 0.000, indicating that the model is fit and the variables collectively explain the adoption of CSA technologies effectively. The Pseudo R² value of 0.491 suggests that the model explains a fair proportion of the variance in CSA adoption (

Table 2. Logistic regression of farmers’ adoption of climate-smart agriculture (CSA) technologies in rice production with selected independent variables.

Variables	Coefficient	p-Value	SE
Age	0.226	0.155	0.016
Gender	0.457	0.277	0.420
Involve members	0.181	0.183	0.135
Education	0.195	0.000	0.050
Income	0.136	0.023	0.021
Off-farm	−0.355	0.411	0.433
Climate member	1.272	0.012	0.508
Group member	0.118	0.757	0.383
Lease	0.188	0.792	0.714
Exp	−0.037	0.005	0.014
Farm size	−0.005	0.016	0.021
Extension	0.293	0.596	0.554
Access to credit	−0.874	0.048	0.442
Constant	−1.104	0.677	2.648
Summary Statistics
N	200
LR chi2(13)	46.980
Prob > chi2	0.000
Pseudo R2	0.491
Log Likelihood	−94.143

).

Among the independent factors analyzed, the academic qualification of farmers (Education) emerged as a significant predictor of CSA adoption, with a coefficient of 0.195 (p < 0.01), indicating that an increase in education level significantly increases the likelihood of adopting CSA technologies (Table 2). A farmer in the Khairahani municipality mentioned that after completing high school, education helped them to read newspapers and advertisements, which helped them to understand the long-term benefits of different CSA practices. They mentioned that education helped them see how crop rotation and mulching could improve soil health and increase yields. Involvement in climate-related organizations (Climate members) also significantly influences farmers’ adoption of CSA technologies, with a coefficient of 1.272 (p < 0.01). Farmer 23 stated that “joining a farmers’ group in my village helped me learn about water conservation techniques like water harvesting. I use it now because we face water shortages during critical times, and this method helps me save water for when I need it the most” (female, 52, Kalika municipality). Conversely, the factor access to credit (Access to credit) was significantly and negatively associated with farmers’ adoption of CSA technologies with a coefficient of −0.874 (p < 0.05). This suggests that access to credit facilities might deter the adoption of CSA technologies, possibly due to the conditions attached to these credit facilities or farmers’ risk perceptions. A farmer (male) from the Ratnanagar municipality stated “Even though I have access to credit, I don’t think it’s the right time to borrow for CSA practices. The process of taking a loan from a bank/cooperative is complicated, and I would rather invest in things I already know, like seeds and pesticides, which give immediate results.” Another factor, farmers’ years of farming experience (Exp), was significantly and negatively associated with the adoption of CSA technologies [coefficient of −0.037 (p < 0.05)], indicating that increased farming experience slightly reduces the likelihood of adopting CSA technologies. A farmer from the Rapti municipality mentioned, “I’ve been farming for over 30 years, and the traditional ways have always worked for me. I don’t see the need to change my practices, especially when I’m not sure these new methods will really help.” The farm size variable was significantly and negatively associated with the adoption of CSA practices (coefficient −0.005 (p < 0.05)), indicating that farmers with smaller farm sizes are more likely to adopt CSA practices. Other factors, including age, gender, involvement of family members in farming, off-farm income sources, and the lease of farmland, though explored, did not show statistically significant effects on the adoption of CSA technologies at a significance level of 0.1 or less (Table 2).

4. Discussion and Conclusions

This study provides insights into the dynamics of climate-smart agriculture (CSA) adoption among rice farmers in Nepal. The results revealed both the potential and the limitations of existing frameworks in promoting sustainable agriculture under changing climatic conditions. Our results indicated that 72% of respondents had adopted at least one CSA practice, pointing to a considerable engagement with these techniques among the local farming community. These high adoption rates indicate some knowledge to test or adopt these methods. If these CSA methods make cropping and production resilient to climate variability, then their adoption will enhance resilience to climate variability in the region. Out of multiple independent variables, education emerged as a key determinant of the adoption of CSA practices by farmers. The positive correlation between the education level of the household heads and CSA adoption highlights the importance of awareness and understanding of CSA practices and climate change impacts, which education tends to foster [46]. Therefore, enhancing educational opportunities could be a pivotal strategy in increasing CSA uptake, as more educated farmers are likely to have better access to information and more capacity to implement complex agricultural innovations that can enhance resilience to climate change [43].

Furthermore, membership in climate-related organizations also significantly influenced CSA adoption [46,47]. This finding emphasizes the role of collective action and knowledge sharing facilitated by such groups, which can provide critical support and resources for farmers transitioning to more sustainable practices. Strengthening these organizations and expanding their reach could thus serve as an effective tool for promoting wider adoption of CSA practices [14,22,30,43,47]. Conversely, our results indicated a negative association between CSA adoption and farmers’ experience, access to credit, and farm size.

The negative relationship between experience and CSA adoption may reflect a resistance to change among more experienced farmers, who might rely more on traditional farming methods. This suggests that targeted intervention programs that address the specific needs and concerns of experienced farmers might be necessary to encourage the adoption of innovative practices among this demographic. Another variable, access to credit, was found to negatively impact CSA adoption, possibly due to the stringent conditions attached to such financial products or a strong dislike of the perceived risks associated with new agricultural technologies. The failure of credit mechanisms to support CSA promotion can be traced to several systemic issues. In Nepal, various financial institutions provide loans to farmers to support agricultural activities. The Agricultural Development Bank (ADB/N) plays a key role, specializing in financing agricultural and rural development projects. Additionally, commercial banks such as Nepal Bank Limited and Rastriya Banijya Bank (National Commercial Bank) offer agricultural loans to farmers. Microfinance institutions (MFIs), including Sana Kisan Bikas Bank (Small Farmers Development Bank) and Nirdhan Utthan Laghubitta Bittiya Sanstha Limited (a microfinance providing service to the poor), provide smaller-scale loans to marginalized farmers. Rural cooperatives are also significant, extending credit often at lower interest rates, while regional development banks support smallholders and cooperatives through specialized agricultural loan programs. However, farmers often prefer informal credit sources over formal institutions due to bureaucratic obstacles and unfavorable loan conditions [48]. Furthermore, high collateral demands, inflexible repayment schedules, and rigid eligibility criteria make formal credit inaccessible to smallholders [49]. These structural barriers mean credit institutions inadvertently create financial stress, discouraging investment in CSA, especially for marginalized groups, such as women and youth, who face additional hurdles in accessing credit and extension services, limiting their potential for CSA adoption [49]. These financial institutions might need to adjust their credit/loan terms and policies to better support the adoption of CSA practices, perhaps by offering more favorable terms or by incorporating insurance schemes that mitigate the perceived risks associated with their uptake. In addition, structural changes within these financial institutions, such as developing specialized CSA lending teams or training, could improve the credit system’s alignment with CSA goals. Moreover, formal credit and loan systems are often unsuited for CSA, lacking tailored support that could mitigate associated risks. Adoption of CSA requires financial inputs and confidence in long-term benefits, which credit systems could support better by offering flexible terms, incorporating crop insurance, or aligning with existing governmental or non-governmental organizations’ extension services [50]. As seen in successful models like Payments for Environmental Services (PES), programs integrating credit with environmental incentives have potential but require careful design and public funding to make CSA accessible and appealing to smallholder farmers [50,51,52]. Therefore, adapting financial institutions’ policies to align with CSA’s unique needs through specialized credit and loan products or risk-reducing mechanisms is critical for promoting CSA adoption in developing countries like Nepal.

The variable measuring farm size was also negatively associated with the adoption of CSA practices, indicating that farmers with smaller farm sizes were more likely to adopt CSA practices [32]. As small-scale farms are always looking for ways to minimize risks and maximize returns [53], they might perceive CSA practices as risk mitigators. Not all CSA practices considered in this study are equally adopted by all farmers. Small farm sizes can indeed limit the types of CSA practices that farmers can implement, particularly those requiring larger areas for effectiveness, such as certain agroforestry systems or large-scale water harvesting. However, space-efficient and labor-intensive practices, such as crop residue mulching, pest-resistant varieties, and zero-tillage, are more feasible for smaller farms. These practices require minimal land but can significantly improve soil health and productivity, which may explain why smaller farms are more likely to adopt them. The analysis also shows that smaller farms adopt CSA practices more frequently, likely because these farmers are more vulnerable to the impacts of climate change and are thus more motivated to adopt adaptive strategies. In addition, from a cost-benefit analysis perspective, CSA practices are considered to have significant economic advantages for smallholder farmers, particularly in developing countries. CSA practices not only enhance productivity but also improve profitability, making them an economically viable option for farmers facing climate-related challenges. Studies indicate that CSA practices provide positive net returns, with revenues surpassing costs and promoting financial sustainability for small-scale farmers in the long run. Practices such as drought-tolerant crop varieties and intercropping have proven financially viable [54]. CSA adoption also enhances household and farm income, which improves rural welfare [55,56,57]. Furthermore, improved market access and extension services are essential for maximizing profit efficiency, highlighting the need for supportive infrastructure to enhance the economic impact of CSA technologies [49,50,55,56,57]. Therefore, with limited land, any decline in productivity can have severe economic consequences, pushing smallholders to seek resilient methods [53]. In addition to this, smaller farms tend to have fewer resources for large-scale mechanized agriculture, making CSA practices that are low-cost and resource-efficient more attractive [35,36]. For larger farms, the adoption of CSA could include the logistical difficulties of managing larger areas or the higher initial costs involved in transforming larger operations.

Policy interventions that provide scalable solutions tailored to different farm sizes could help in addressing these barriers. The findings of this research have substantial implications for policy and practice. Policies focused on improving educational attainment and organizational support are essential due to their significant influence on CSA adoption. Effective investments in education and training programs tailored to sustainable agriculture and strengthening farmer organizations and networks could facilitate broader adoption of CSA practices. Moreover, revisiting credit schemes to better support sustainable practices and addressing the specific needs of both experienced and large-scale farmers can remove significant barriers to CSA adoption. A tailored outreach and extension approach that considers different farmer groups’ diverse needs and conditions will likely be more effective in promoting sustainable agricultural practices across varying contexts. The widespread adoption of CSA in developing countries like Nepal will provide disproportionate benefits as it offers significant benefits for soil health and climate resilience, addressing the challenges posed by climate change [58,59]. By integrating sustainable practices, CSA not only improves agricultural productivity but also fosters environmental sustainability [58]. The adoption of CSA practices not only increases the adaptive capacity of farmers but also helps farmers stabilize production by minimizing climate-induced risks and ensuring food security [58]. Consequently, long-term CSA adoption can create a robust foundation for sustainable agricultural systems, though supportive policies and financial incentives remain essential for broader adoption [26,27,29,58,59].

In conclusion, while CSA practices have gained traction among farmers in developing countries, the full potential of these practices has yet to be realized. Addressing educational, financial, and structural challenges through targeted policies could support more widespread adoption of CSA, contributing to enhanced resilience against climate change impacts in Nepal’s agricultural sector. Future research could include a temporal analysis to examine how long certain CSA practices have been in use, explore adoption intensity, distinguish between different stages of adoption of CSA practices, explore the long-term impacts of CSA adoption on productivity, sustainability, and livelihoods, and incorporate a system-level analysis of land and soil factors to provide deeper insights into the complex interplay of factors that influence sustainable agricultural practices in regions vulnerable to climate change.