Exploring the water-energy-food nexus among farming households in northern Bangladesh

The interconnections among water, energy, and food systems—commonly known as the Water-Energy-Food (WEF) Nexus—are gaining attention as critical components of sustainable development, particularly in regions where farming households depend only on agriculture and agriculture-related activities for their livelihoods. This study investigates the WEF nexus at the farming household level in Northern Bangladesh, a region characterized by intensive agricultural activity, seasonal water scarcity, and growing energy demands. As agriculture accounts for 70% of global freshwater withdrawals and over a quarter of global energy use, managing these resources through an integrated “system of systems” is critical for sustainability. The research aimed to determine how household-level interactions between water and energy consumption influence food security and whether integrated management of this nexus can generate synergistic effects for long-term development. Using a log-linear (OLS) regression model and structured surveys of 80 farming households, the study analyzed the determinants of monthly household food expenditure, treating the cost ratio (the proportion of energy cost relative to total irrigation cost) as a primary indicator of the nexus dynamics. Key findings reveal that energy costs represent 75% of total irrigation costs in the study area, primarily due to the energy-intensive nature of groundwater pumping for rice production. The empirical results confirmed the central hypothesis: the cost ratio has a significant negative impact on food expenditure (β = −0.225, p = 0.037), indicating that as energy costs dominate irrigation expenses, household food security declines. Additionally, farm size and household size were positively associated with food expenditure, while households in the study area faced a food access gap of roughly 30%, significantly higher than the national average. The study concludes that climate-driven stressors, such as declining rainfall and rising temperatures, have created a causal chain of increased production costs and reduced livelihood security. To break this cycle, the authors recommend transitioning to solar energy for irrigation, adopting drought-tolerant crops, and implementing integrated policy frameworks that move beyond traditional “sector-based” management. The findings highlight the potential importance of integrated WEF management in supporting SDG-related sustainability goals.

Sustainability is defined as the long-term goal of maintaining the Earth's life-sustaining mechanisms—including clean water, healthy soil, and energy—across generations, thereby establishing an ecological balance that preserves the quality of life in modern societies (Costanza and Patten, 1995; Sabau, 2024). In contrast, sustainable development refers to the processes and actions undertaken to achieve this goal (UN, 2018; Wolff and Ehrström, 2020). Both concepts are fundamentally linked to the interconnected pillars of economy, society, and the environment. Critically, access to natural resources such as water, soil, and energy is essential for food security and human well being (Biggs et al., 2015). Recent research further indicates that governance of the Water-Energy-Food (WEF) nexus offers a promising opportunity to enhance sustainability outcomes, including within urban systems (Jones-Crank, 2024).

Earth holds roughly 1.386 billion cubic kilometers (333 million cubic miles) of water, of which 97.5% is saline and just 2.5% is freshwater (Oksana and Dmytro, 2021). Rising water consumption is driven by a growing global population, an expanding middle class, and technologies that make water extraction and distribution easier. Alongside climate change, which further degrades water supplies, these factors boost freshwater use in sectors like agriculture and industry, placing immense strain on this vital resource (Gbandi, 2022). By 2050, global water demand across all sectors is projected to rise by 20%−30%, though this increase will vary greatly by region (Baggio et al., 2021). Additionally, water consumption for energy production is forecast to increase by 85% globally by 2050, and food demand—driven by population growth—is set to rise by 60%, which will directly boost the water needs of agriculture (Baggio et al., 2021).

Agriculture currently accounts for 70% of global freshwater withdrawals, supplies nearly 40% of the world's food through irrigation, and consumes over a quarter of global energy for food production and distribution (FAO, 2011; Ghosh et al., 2024; Halder and Islam, 2015). As for food security, FAO has reported that about 746 million people were food insecure in the world, particularly in African and Asian countries (FAO, 2020). According to the United Nations, food security defines a situation when all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food which meets their dietary needs and food preferences for an active and healthy life (FAO, 2002).

Approximately one-third of global annual end-use energy is attributed to food supply and consumption, with the agri-food industry relying heavily on fossil fuels for production, transport, processing, and distribution—a demand that will continue rising steadily in line with increasing food needs (Sims et al., 2015). As a result, the increase in energy demand causes an increase in carbon dioxide emissions, one of the leading sources of environmental degradation also (Ahmad and Majeed, 2019). The nonrenewable energy resources are expected to last between 30 and 150 more years; however, the production of gas and oil will drop to roughly 40%−60% by 2030 compared to 1970s (Finley, 2012; Kahia et al., 2016). More importantly, the world is in an accelerated process of energy transition, which means that globally electricity generation by renewable energy sources will grow (Yang et al., 2024); hydroelectricity and wind energy are projected as the largest shares in total renewable electricity generation at a percentage of 54 and 26%, respectively over the period 2007–2035 (Apergis and Payne, 2012).

Many countries around the world are increasing their domestic food production to improve food security (Haji et al., 2022). As a result, demand for water, food and energy is increasing, driven by a rising global population, rapid urbanization, changing diets and economic growth (FAO, 2020). To meet these rising demands with limited access to these resources it is necessary to address water, energy and food issues jointly, because choices and actions in any of these domains can significantly affect the others (Halstead et al., 2014). From an economic, environmental, and social standpoint, sustainable management of natural resources must take into account a reasonable balance between their availability and use (Fabiani et al., 2020). Sustainable agricultural development increasingly demands a complete understanding of how critical resources—water, energy, and food—interact. The objective of sustainable agriculture is to develop new and creative farm equipment and services that support the intensive farming industry in maximizing productivity and input management (water, energy, and food; Fabiani et al., 2020) while also protecting nature.

The sustainability triangle (or the WEF nexus) is a conceptual framework introduced by Rabi H. Mohtar in the late 1990s (formally cited as Mohtar, 2022) to describe a strategic and holistic framework for managing the three most essential resources for human development (Adeel and Böer, 2024; Ansari et al., 2023; Mohtar, 2022). It is built on the fundamental argument that water, energy, and food are cross-linked and intertwined, forming a triangular relationship where a change or disturbance in one sector inevitably forces a response in the other two (Ansari et al., 2023). The triangle identifies Water, Energy, and Food as three cross-linked services essential to economic, social, and environmental development as well as regional development (Mohtar, 2022). The Water-Energy-Food (WEF) nexus has profound impacts on the livelihoods, health, and social stability of farming households, particularly in regions where a large portion of the population is employed in agriculture (Adeel and Böer, 2024).

Albrecht et al. (2018) conceptualize the Water-Energy-Food (WEF) nexus in three interconnected ways: as an analytical tool employing qualitative and quantitative methods to examine resource interactions; as a conceptual framework supporting coordinated policymaking across sectors; and as a discourse that frames complex problems and encourages collaborative solutions. The ultimate goals of nexus thinking, they explain, are to improve resource-use efficiency, address social and environmental impacts across multiple scales, increase synergies, and reduce trade-offs. Critically, Albrecht et al. (2018) highlight that many nexus studies remain largely theoretical and advocate for mixed-methods, transdisciplinary approaches that integrate social and political dimensions while actively involving stakeholders and decision-makers. Complementing this view, Howarth and Monasterolo (2016) argue that addressing the interconnected nature of the WEF nexus requires a co-produced approach to knowledge development, which can inform timely and effective decision-making, strengthen societal resilience to shocks, and move beyond the sector-based focus of traditional research practices. Together, these studies underscore that effective WEF nexus governance depends not merely on technical solutions but on participatory, context-sensitive methodologies that bridge disciplinary and institutional boundaries.

In addition, Endo et al. (2017), Farmandeh et al. (2024), and Jones-Crank (2024) defined the Water–Energy–Food (WEF) Nexus as an important analytical and integrated governance framework that captures the complex interconnections among water, energy, and food systems. They emphasize its particular relevance in regions vulnerable to climate change, environmental degradation, and resource scarcity, where it can improve resource efficiency and support long-term system sustainability. In Bangladesh, agriculture accounts for a substantial share of water withdrawals and energy consumption, while simultaneously being the backbone of food security and rural livelihoods. In Northern Bangladesh, farming households face a complex interplay of challenges: seasonal water shortages, rising energy costs, and food production pressures. So, it is necessary in order to ensure food security of all people with cost-efficiency and in a sustainable way, that less water and energy resources are used for more production of food and for preserving the integrity of ecosystem functions and processes. This study aims to investigate whether there is a nexus between households' water and energy consumption and their food security, and whether managing this nexus can lead to sustainability outcomes. Based on the formulated thesis statement of this study, the research questions are: (a) What are the patterns of water and energy consumption, and the food security status at the farming household level? (b) How these three factors are related to each other? (c) Is their combined implementation necessary to generate synergy effects? This study seeks to answer these research questions.

The WEF framework

The Water-Energy-Food (WEF) Nexus was first conceptualized at the Bonn 2011 Nexus Conference (Hoff, 2011; Mahlknecht et al., 2020), as a pathway for understanding complex and dynamic interlinkages between issues related to water, energy and food security (Nhamo et al., 2020). The WEF nexus is a conceptual framework that can illustrate sustainability in its three dimensions, ecological, economic, and social, and is expected to make the implementation of the Sustainable Development Goals (SDGs), particularly Goal 2 (Zero hunger), Goal 6 (Ensuring availability and sustainable management of water and sanitation for all) and Goal 7 (Ensuring access to clean and affordable energy) more efficient (Biggs et al., 2015; Brandi et al., 2014; Yumkella and Yillia, 2015). Knowledge of the interactions between WEF resources temporally and spatially can enhance resource security and facilitate inter-sector and policy making, which can eventually lead nations toward sustainability (Mahlknecht et al., 2020). Figure 1 outlines an analytical framework for studying the Water-Energy-Food (WEF) nexus, structured around key components and their interconnections. It identifies water and energy as foundational inputs, each broken down into sources (e.g., groundwater, rainwater for water; liquid fuel, solar for energy) and associated costs (e.g., irrigation/storing for water; pricing, maintenance for energy).

These inputs converge in Water for Irrigation, which directly drives agricultural production (crop yield). The framework further incorporates other factors at two scales: the household level (income, farm size, education, experience, etc.) and the national level (price, food availability, and public policy). This framework enables a multi-dimensional analysis of how water and energy costs and access mediate crop output, while socio-economic and policy contexts moderate these relationships. It is particularly suited for empirical studies on resource efficiency, food security, and trade-offs in agricultural systems, allowing researchers to model both biophysical constraints and human/institutional drivers within the WEF nexus. Based on this analytical framework, we investigated the Water-Energy-Food (WEF) nexus for rural farming households in Northern Bangladesh. Due to budget and time constraints, our study did not include all national-level factors identified in the framework; we incorporated only pricing—specifically as it relates to household food expenditure—given its direct and measurable impact on the nexus.

Data collection

The research area (Figure 2) has been selected purposively to reveal the food security status of farming households with consumption of water and energy in the Northern part of Bangladesh by using the WEF nexus framework (Figure 1). The study area is characterized by high irrigation dependency, seasonal energy scarcity, and vulnerable farming livelihoods, making it highly relevant for examining Water-Energy-Food nexus dynamics. The study is based on primary and secondary data. Primary data were collected through face-to-face structured questionnaire surveys administered to 80 farming households in the Rajshahi district of Bangladesh from June 2019 to December 2020. The survey focused on water usage patterns, energy sources, crop yields, and household income. Secondary data has been collected from various sources, such as the government of Bangladesh's economic reviews, the household income and expenditure survey, other related publications, peer-reviewed articles and journals, and reports. Descriptive analysis and log-linear regression were employed to analyze the primary data, while figures were drawn from both primary and secondary data sources.

The study area

Rajshahi district is located in the north-western part of Bangladesh and covers about 1.64% (2425.4 km 2) of the total area of the country. The primary source of the people's livelihood in the study area is agriculture (BBS, 2018). This district consists of nine sub-districts (upazillas) and Godagari is the largest one, according to the land area and population (BBS, 2018). Rajshahi district is also a part of the Barind Tract.

“The tract is criss-crossed by a number of rivers and underlined by distinct oxidized red soils and sediments of Plio-Pleistocene and Holocene ages.” (Rashid et al., 2014, p. 1). As a result, hard red soil, typical dry climate with comparatively high temperatures and less rainfall are common characteristics of this area (BMDA, 2018).

Empirical model specification

To examine the determinants of household food expenditure within the WEF nexus, we specify a multiple linear regression (ordinary least squares—OLS) model because of its simplicity and popularity (Ozturk and Acaravci, 2013). The relationship between a dependent variable (Y) and an independent variable (X) can be postulated as a linear regression (Gujrati, 2004, pp. 58–62):

where β 0 and β 1 are regression coefficients and parameters while u is an error term. The estimated least squares regression is given by

This is known as simple linear regression which is used to estimate coefficients for the equation. For multiple linear regression with n explanatory variables (X 1 …….. X n), the estimated least squares regression is written as follows:

Food expenditure is treated as a continuous dependent variable capturing household consumption of purchased and home-produced food items (converted to monetary value using local market prices). The general form of the model is:

This model estimates how household income, family size, landholding size, and the energy-to-irrigation cost ratio affect household food expenditure. The energy-to-irrigation cost ratio represents the share of energy expenses within total irrigation costs. Since energy is the major component of irrigation expenditure in the study area, a higher ratio indicates greater energy dependency and higher production pressure on farming households. Household electricity use is generally limited to basic purposes such as lighting, fans, and televisions, while frequent load-shedding further constrains non-agricultural electricity consumption. Based on questionnaire responses, we added a lump-sum amount to total energy cost to account for household electricity consumption. Consequently, rather than including energy cost and irrigation cost as separate variables in the regression, we used the ratio of energy cost to irrigation cost as a single explanatory variable.

All dependent and explanatory variables were converted into natural logarithms. This log-linear specification was employed to (i) reduce the influence of outliers, (ii) transform positively skewed data toward normality, (iii) linearize nonlinear relationships between variables, and (iv) allow direct interpretation of coefficients as elasticities. The empirical equation is modeled as follows:

where β 0 and ϵ i represent ln (β 0) and error term respectively of the ith household respectively. Most importantly, β 1,………………, β 4, represent the long-run elasticities of the dependent variable with respect to the independent variables. It is hypothesized that the coefficients signs of all the explanatory variables are expected to be positive. The description of the variables is provided in Table 1.

The WEF nexus is particularly relevant in the context of climate change, resource degradation, and increasing population pressure, which intensify competition for resources across the water, energy, and food sectors (Endo et al., 2017). Although the global population growth rate has been steadily declining, Bangladesh continues to experience comparatively higher population growth (Worldometer, 2025). This sustained demographic expansion, together with a youthful age structure and very high population density, exerts substantial pressure on the country's water, energy, and food systems. There are three major components of the WEF nexus, namely water, energy and food systems and these factors are interrelated (Figure 1). The WEF results are analyzed using these components, based on collected primary and secondary data.

Pattern of water- energy-food consumption

Water

Water supplies in northern Bangladesh, particularly in regions like Rajshahi, are under increasing stress due to dropping groundwater levels from intensive irrigation and diminishing surface water during the dry season. Climate change exacerbates this problem by altering rainfall patterns and increasing the frequency of droughts and floods, thereby endangering food and water security (Misra, 2014). Temperature extremes further constrain crop production, as deviations from optimal ranges negatively affect soil organic matter and contribute to food insecurity (Anik et al., 2012). In the study area, annual maximum temperatures rose by 0.013–0.032°C per year between 2001 and 2018, with monthly averages exceeding 30°C from March to October and peaking in summer (April–June). Notably, minimum temperatures during the harvesting period of the main rice crop (March–October) increased from 18.6 to 23.05°C, posing additional risks to agricultural productivity.

Rainfall is critical for crop production in Bangladesh, where only 56% of cropland is irrigated, leaving much agriculture dependent on rain (Uddin et al., 2024). In the study area, annual total rainfall decreased significantly between 2001 and 2018—by 18.55% in 2006, 20.68% in 2009, and 24.06% in 2010—accompanied by sharply declining groundwater levels that peak in depth from March to May (BBS, 2018; BMDA, 2018). Heavy rains from late June–September frequently cause flooding, while low or no rainfall from November to February creates water scarcity, leading to drought, high temperatures, and increased pressure on groundwater extraction. This declining rainfall exacerbates Water-Energy-Food (WEF) nexus challenges by raising irrigation demand and energy costs for groundwater pumping.

Drought also poses a major problem for agricultural production in the northwestern part of the country (Chan et al., 2016). Groundwater is the main source of irrigation in northwest Bangladesh, and the higher daily pumping rate in dry seasons may aggravate the situation of groundwater scarcity in the region (Shahid, 2011). This situation is very common both for water for domestic use and for agriculture use, which comes mostly from ground water via deep tube wells and shallow tube wells (Anik et al., 2012; CCBVO, 2018). In the study area, rice is the main food crop that provides nearly 90% of rural employment and about two-third of the total calorie supply of an average person (CCBVO, 2018). The total production cost of rice can be used to explain the relation between water and food because irrigation is the main input of rice production. Figure 3 provides a graphical presentation of the total production cost of rice, with irrigation as the first position in the total production cost and the amount is around 30,000 Taka (Tk.) per acre. This high percentage of 60% of the total production cost may be due to higher temperatures and low rainfall amount which have had negative effects on soil organic matter (Anik et al., 2012). The cost of fertilizer is the second item representing one-fifth of the total production cost, followed by the cost of labor and other costs, respectively, in the study area. All survey respondents believed that the cost of irrigation has increased two times more (on an hourly basis) compared to the past 10–15 years. However, increasing the rice monoculture cropping (and dependency of rice) is an important cause of the upward trend of irrigation costs as well as of total production cost for rice.

Energy

Bangladesh relies heavily on fossil fuels for electricity-−64% from natural gas and 20% from furnace oil—while renewables account for only 3% (Das et al., 2020). Although 80% of the population has electricity access, supply remains unreliable and unaffordable, especially in rural areas, due to infrastructural challenges and declining fossil fuel availability (Bangladesh Clean Energy Summit, 2019; Islam et al., 2014; Khanam et al., 2018). The share of agriculture and industry sectors in electricity consumption is increasing gradually (Masuduzzaman, 2013), though only a small percentage of electricity (3.09%) is used in the agricultural sectors (Das et al., 2020). In the study area, 52% of the total population has partial access to electricity; that barely meets limited usage for light, ceiling fans, refrigeration, and irrigation (BBS, 2018). All respondents believed that energy cost is the main input in the cost of irrigation. From the respondents' point of view, the cost of energy use is significantly affecting the irrigation cost as well as the total production cost of rice.

Figure 4 shows that the cost of energy use represents 75% of the total irrigation cost and the rent for equipment cost is one-fifth of it, while a small percentage of the cost (5%) is covered by other spending such as purchase of plastic pipes, labor, etc. Most of the farmers operate on small-scale farms and they have no access to funding for covering fixed costs to build infrastructure for irrigation and to purchase irrigation machinery; as a result, they rent irrigation equipment from others. In addition, 80% of the irrigation system is operated by unconventional energy such as diesel due to lack of access of electricity; as a result, irrigation cost is higher than for electricity supported irrigation systems.

Household food security

The results of food security are analyzed based on food access and utilization because the primary produced food is rice, which does not reflect the overall food availability in terms of a balanced diet comprising all essential food items. However, food access is ensured when households and all individuals within them have adequate resources to obtain appropriate foods for a nutritious diet. Stable sources of income (with sufficient amount) are crucial for ensuring adequate food security at both the individual and household levels (Dev et al., 2020; Talukder, 2005). Monthly access to food and the actual food expenditure scenario are shown in Table 2. The table shows that Bangladeshi citizens spend less than 50% of their average monthly income to purchase food, whereas in the study area people spend nearly 70% of their average monthly income to cover their basic needs for food. The required food expenditure of households per month at national level is Tk. 7,879.81 with a food expenditure gap of Tk. 524.47 and an access gap of 8.93% whereas this access gap is nearly 30% at the household level in the study area.

In Bangladesh, many people cannot afford the variety of foods needed to meet their daily calorie and nutritional requirements, largely due to low incomes. Figure 5 shows the source of food value from different food groups at the national level. Cereal products are the main source of calories, which is the sum of rice, wheat and pulses. Vegetable products, namely potato and other vegetables, are the second contributor of these food values. The next important source is protein (it consists of fish, meat, and eggs) followed by fruit, milk, oil (fat), and sugar.

The daily per capita calorie intake is a key measure of nutritional status, with the minimum dietary energy requirement for an active adult being about 2,100 kcal (Dev et al., 2020; Teklemichael and Mekuriaw, 2025). In Bangladesh, the upper poverty line is set at 2,122 kcal and the lower at 1,805 kcal based on the Direct Calorie Intake method (Hossain, 2020). Figure 6 provides the graphical presentation of the sources of food value at the household level in the study area. Cereals, being the staple food, are the highest providers particularly of energy, which secure, on average 78% of the calories, and edible oil is the second source at the household level. This dietary imbalance reflects insufficient domestic production of non-cereal foods like pulses, oilseeds, fruits, meat, milk and eggs, along with low incomes, and lack of nutrition education (Hossain et al., 2005; Talukder, 2005).

Empirical results

Descriptive analysis

In this section, we present the results of the empirical analysis on the Water-Energy-Food (WEF) nexus at the household level. For this research study, data were collected from households focusing on their socioeconomic conditions, along with information related to the variables specified in the equations. Table 3 presents the basic characteristics of the collected data, commonly referred to as descriptive statistics.

Table 3 presents descriptive statistics for the key study variables. The average monthly household food expenditure was 4,999.70 BDT (SD = 2,134.56), ranging from 2,500 to 12,800 BDT, indicating substantial variation in food consumption across the 80 farming households. The mean monthly household income was 11,346.00 BDT (SD = 6,789.23), with a wide range from 5,000 to 35,000 BDT, reflecting considerable economic heterogeneity in the sample. The average farm size was 2.45 acres (SD = 1.56), ranging from 0.50 to 8.00 acres, which is typical for smallholder farming systems in northern Bangladesh. Household size averaged 3.85 adults (SD = 1.34), with a minimum of one and maximum of seven adults per household. The cost ratio (energy cost divided by irrigation cost) had a mean of 1.02 (SD = 0.08), ranging from 0.85 to 1.35, indicating that energy costs are nearly equivalent to irrigation costs with very limited variation across households. This low variation in the cost ratio suggests that in this context, energy and irrigation costs are tightly linked, reflecting the energy-intensive nature of groundwater pumping.

Correlation analysis

Table 4 presents the Pearson correlation matrix examining bivariate relationships among the study variables. Food expenditure shows statistically significant correlations with all predictors except household size. Specifically, food expenditure is positively correlated with farm size (r = 0.362, p < 0.05) and household income (r = 0.487, p < 0.05), indicating that households with larger farms and higher incomes spend more on food. Most notably, the cost ratio (energy cost divided by irrigation cost) exhibits a strong negative correlation with food expenditure (r = −0.507, p < 0.05), providing compelling bivariate evidence for the Water-Energy-Food (WEF) nexus hypothesis. Specifically, households facing higher energy costs relative to irrigation costs tend to spend significantly less on food. Household income is strongly correlated with household size (r = 0.729, p < 0.05), indicating potential multicollinearity concerns, which are subsequently addressed through variance inflation factor (VIF) diagnostics in the regression analysis. The cost ratio also shows moderate negative correlations with farm size (r = −0.316, p < 0.05), while household size exhibits no significant correlation with food expenditure (r = −0.087, p > 0.05). All correlations are in the expected directions, and the strong negative relationship between the cost ratio and food expenditure (r = −0.507) provides robust preliminary support for the central hypothesis of this study.

Table 5 presents the log-linear regression results examining the determinants of household food expenditure among farming households in northern Bangladesh. The model is statistically significant F (5, 74) = 6.71, p < 0.001 and explains 48.5% of the variance in food expenditure (R 2 = 0.485). The adjusted R 2 of 0.458 accounts for the number of predictors in the model. The Durbin-Watson statistic of 1.94 indicates no autocorrelation, and variance inflation factors (VIFs) range from 1.18 to 1.38, confirming the absence of problematic multicollinearity.

Among the independent variables, three achieved statistical significance at conventional levels. Farm size shows a positive and significant association with food expenditure (β = 0.238, p = 0.032), indicating that households operating larger farms spend more on food. Household size is also positively and highly significantly associated with food expenditure (β = 0.157, p = 0.004), reflecting the greater food requirements of larger families.

Critically, for the Water-Energy-Food nexus hypothesis, the cost ratio (energy cost divided by irrigation cost) exhibits a statistically significant negative effect (β = −0.225, p = 0.037). This finding suggests that households facing higher energy costs relative to irrigation costs reduce their food expenditure, providing empirical support for the nexus framework. Household income shows a positive but only marginally significant association (β = 0.187, p = 0.091). While this trend aligns with economic expectations, the lack of full significance may stem from income measurement challenges in agrarian settings or the modest sample size (N = 80).

Interconnection analysis

Khan et al. (2009) highlight that reducing environmental footprints in food production requires coordinated improvements in water and energy management, as shown through an Australian case study on rice, wheat, and barley. Enhancing water productivity and energy efficiency can lower the pressure on natural resources while maintaining agricultural output, addressing challenges from population growth, climate change, and resource scarcity. Gomez-Zavaglia et al. (2020) note that agriculture's global economic contributions are highly climate-dependent, with rising temperatures and CO 2 levels threatening nutrients, soil moisture, and water availability, while more frequent droughts and floods disrupt farming and food production. Syaukat (2011) shows that climate change—especially rising temperatures—threatens rice production in Indonesia, with a projected 90-million-ton shortfall by 2050, requiring integrated adaptation strategies. This supports the water–energy–food nexus approach in Northern Bangladesh by demonstrating that coordinated resource management can boost agricultural productivity and household food security. This finding also highlights the importance of transdisciplinary research, as addressing interconnected challenges across water, energy, and food systems requires the integration of knowledge from multiple disciplines, stakeholders, and policy domains.

The study area is highly dependent on irrigation for agriculture especially during summer and winter. The climate change impact, resulting in drought and decreasing ground water levels, has become a main concern to present agricultural activities in this region (Anik et al., 2012). The Figure 7 presents a structured interconnection framework showing how climatic problems cascade through water-energy-food (WEF) systems, leading to production, socio-economic, and environmental impacts, and ultimately producing negative outcomes that feed back into greater vulnerability in the study area. The figure (Part A) shows the main climatic problems and environmental impacts of agriculture production in the study area. According to respondents' opinions, droughts and irregular rainfall together are the main climatic problems and drought only is the second main climatic problem. It is also mentioned that all respondents believed that raising temperatures, decreasing rainfalls, increasing droughts, reduced access to groundwater and surface water were the main aspects of climate change, both separately and together also. These climatic stressors are the starting point of the causal chain.

These climate drivers directly affect the Water, Energy, and Food (WEF) Nexus Systems by disrupting the interconnected resources essential for agricultural livelihoods (Part B of this figure). On the water side impacts include precipitation variability, surface water availability, groundwater recharge, and irrigation water supply. Energy is affected through increased demand for pumping, irrigation, and the production and transport of agricultural inputs. Food systems, in turn, experience disruptions in crop production, livestock production, and fisheries and aquaculture. Together, this illustrates how climate variability undermines the critical resource flows that sustain farming and food security.

Disruptions to the water, energy, and food (WEF) systems lead to clear production impacts (Part C of the Figure), as reported by respondents: 37% faced higher production costs and 31% experienced actual production losses. More specifically, irrigation costs rose for 41%, chemical costs (pesticides and fertilizers) for 32%, and labor costs for 29%. These figures demonstrate that farmers are simultaneously dealing with reduced yields and rising input expenses, which ultimately squeezes their profitability, and threatens their livelihoods and food security.

The production impacts described above give rise to broader socio-economic and environmental consequences (Part D of this figure). Survey results show that 78% of respondents experienced reduced access to drinking water troughs, while 52% reported diminished access to fish in open water resources. Additionally, 78% confirmed a decline in livestock numbers. On the household front, 90% faced rising medical costs, and 41% saw an increase in overall household expenditure. Together, these findings illustrate how the effects of climate-driven disruptions extend far beyond the farm, adversely affecting community health, household financial stability, and access to essential natural resources.

The cumulative effect of these climatic and production-related disruptions leads to a set of severe outcomes, including reduced livelihood security, food insecurity, water insecurity, health impacts, and increased poverty and vulnerability (Part E of this figure). These outcomes directly undermine the core pillars of sustainability—environmental integrity, social equity, and economic viability. When communities cannot rely on stable food and water supplies or maintain their health and incomes, long-term sustainable development becomes unattainable. Breaking this cycle requires integrated mitigation strategies that restore resource security, build resilience, and lift vulnerable populations out of poverty, ensuring that both people and the environment can thrive over time. At the bottom, the framework lists governance, policy, and adaptation measures that can intervene in this cycle, including climate-smart agriculture, water management, risk reduction, diversification, social protection, and early warning systems. Without effective government mitigation policies, these outcomes increase vulnerability, which further worsens climate impacts—a self-reinforcing downward spiral. In addition, transdisciplinary research is highly relevant, as sustainable management of the WEF nexus requires the integration of scientific knowledge, local experience, and policy perspectives. Such an approach is vital for ensuring the sustainability of water, energy, and food systems not only for the present population but also for future generations.

This study provides robust empirical evidence confirming the existence of a Water-Energy-Food (WEF) nexus between rural households' water and energy consumption and their food security in the drought-prone, irrigation-dependent region of northern Bangladesh, demonstrating that managing this nexus can contribute to sustainability outcomes. The log-linear regression analysis confirms that farm size and household size are significant positive determinants of food expenditure, while critically, the energy-to-irrigation cost ratio exerts a statistically significant negative effect (Table 5), indicating that rising energy costs directly constrain household food budgets. Complementing this quantitative finding, the interconnection framework (Figure 7) validates the cascading effects of climate stressors—including drought, irregular rainfall, and declining groundwater access—which simultaneously disrupt water availability, increase energy demand for pumping, and elevate production costs. Specifically, 41% of respondents reported higher irrigation costs and 31% experienced production losses. These disruptions further cascade into severe socio-economic consequences, with 78% of households reporting reduced drinking water access and declining livestock numbers, while 90% faced rising medical costs, ultimately generating a self-reinforcing cycle of food insecurity, water insecurity, and heightened vulnerability. Collectively, the results confirm that variations in energy consumption and water availability directly affect food-related outcomes in rice-based production systems, underscoring that effective WEF nexus management can stabilize resource flows and enhance sustainability outcomes.

Beyond its empirical findings, this study also contributes to the practical advancement of WEF nexus theory by demonstrating how household-level expenditure patterns can serve as simplified proxy indicators for understanding WEF interactions in resource-constrained settings. In particular, the cost ratio among water, energy, and food expenditures provides an empirical basis for assessing interdependencies and trade-offs at the household scale. Although the study primarily applies existing WEF nexus concepts, it supplements the theoretical discourse by proposing a simplified and context-sensitive approach for evaluating WEF linkages where detailed resource flow data are unavailable. Therefore, the findings extend the applicability of WEF nexus analysis to micro-level socio-economic assessments and support future research on household-centered sustainability evaluation. However, the study's focus on a single drought-prone region, exclusive emphasis on rice cultivation, and modest sample size (N = 80) limit the broader generalizability of the findings, suggesting the need for comparative and larger-scale future studies across diverse agroecological contexts.

Addressing the identified WEF nexus dynamics demands integrated, multi-sectoral interventions rather than siloed approaches. First, given that the energy-irrigation cost ratio is a primary transmission mechanism linking climate stress to reduced food expenditure, transitioning from diesel-powered pumps to solar-powered irrigation systems is the most strategically sound intervention for this region. Solar energy aligns with local weather patterns, can stabilize energy costs, reduce groundwater withdrawal pressures, and simultaneously buffer households from fossil fuel price volatility—directly protecting food budgets. Second, climate-smart agricultural practices must be prioritized, including drought-tolerant crop varieties, rainwater harvesting, and conservation tillage, to reduce irrigation dependence while maintaining productivity. Given that groundwater plays a critical role in adapting soil moisture conditions for high-yielding variety crops and increased cropping intensity, governance interventions should include stricter groundwater regulation alongside community-based water user associations to manage aquifer recharge equitably. Third, social protection mechanisms—such as indexed crop insurance, emergency food assistance during drought years, and health fee waivers—are essential to buffer the 90% of households facing rising medical costs and prevent climate-driven poverty traps. Fourth, livelihood diversification programs (climate-resilient aquaculture, homestead gardening, and off-farm enterprise development) can reduce reliance on groundwater-dependent cropping. Future research must broaden geographic scope, employ longitudinal designs to capture climate variability impacts, and integrate qualitative approaches to understand local decision-making and socio-cultural dimensions of resource use. Crucially, a transdisciplinary sustainability approach—integrating hydrology, agronomy, energy systems, and food security with active participation of local stakeholders—is necessary to design effective policies that enhance resource efficiency, reduce climate vulnerability, and support long-term development of farming communities in northern Bangladesh and analogous semi-arid agrarian settings.