POLICY BRIEF Integrating irrigation realities into climate solutions Evidence from Takeo, Cambodia 2 HIGHLIGHTS Cropping season definitions are highly variable due to diverse local agro- hydrologies that inhibit attempts at synchronized crop calendars Adjusting cropping calendars has been recognized as an important solution to climate change; however, adaptation efforts are still limited. Survey data from Takeo shows 20– 50% of farmers operate outside conventional seasonal definitions, independently choosing cropping windows due to binding constraints such as local water access, flood exposure, and micro-elevation differences. This variability demands a shift towards integrated decisions for local water security through location-specific climate-smart adaptation tools; one-size-fits-all solutions will not work. Wide disparities in irrigation cost signal the need for targeted interventions Pumping costs vary substantially, from USD 2 to 589 per hectare, reflecting inequities in energy access and system efficiency. This range indicates the need for targeted support to optimize water use in high-cost zones, improve pump efficiency, and bring in affordable clean energy to lower production risks and emissions. The development of a high-value agriculture cluster (such as modern AC) needs to consider the spatial structure of these varying water costs and availability. Irrigation decisions are governed by risk management rather than crop water demand consideration, leading to high cost and emissions Despite high pumping costs (up to USD 589 per hectare or 23% of total profit), 98% of farmers do not see a need to save water. More than 70% of farmers prefer to irrigate just after ponding disappears due to perceived unreliability of irrigation water services. Accordingly, there is a wide potential for promoting Alternate Wetting and Drying (AWD). To encourage water savings and limit emissions, policies may reframe water-saving as a cost-reduction strategy while investing in improved irrigation service performance to reduce risks in system-wide AWD implementation. Irrigation delays affect nearly 30% of farmers, disrupting in-season water management Irrigation delays caused by water shortage (35%), labor shortages (20%), and pump failures (18%) reduce farm-level synchrony in planting and affect productivity. Policies to prioritize pump access, infrastructure upgrades, and labor-efficient irrigation solutions can help rice farmers. Sustainable irrigation investments to provide more reliable and flexible irrigation services can address the delays and production impacts. 3 SUMMARY The climate crisis is fundamentally a water crisis. It is marked by rising temperatures, floods, droughts, unpredictable rainfall, declining water quality, and worsening water scarcity. These disruptions deeply affect agriculture by altering both the quantity and timing of irrigation water. For countries like Cambodia, where agriculture underpins food security and livelihoods, these shifts pose a serious threat. They also undermine progress toward the Sustainable Development Goals (SDGs), especially SDG 6 on Clean Water and SDG 2 on Zero Hunger. According to the Climate Risk Index 2024, Cambodia ranked among the top 20 countries most affected by climate-related disasters from 1993 to 2022 (Adil et al., 2025). Projections indicate that climate risks will intensify in the coming decades (WBG, 2024). In response, the Cambodian government has made numerous efforts to promote climate resilience over the past thirty years and is increasing its investment in climate-resilient agriculture. Strengthening rice- based systems is critical for Cambodia, considering the importance of rice systems, which occupy 82% of the country’s cultivated land and employ 36% of the agricultural labor population. Policies and research studies have focused on identifying adequate climate adaptation approaches. These underscore that improved irrigation, cropping calendar shifts, and crop diversification are the most important strategies for building farmers’ resilience to climate change. Amongst these, sustainable irrigation expansion and alternate wetting and drying (AWD) emerged as critical adaptation and mitigation practices. However, translating from recommendations to practical action is not straightforward. All climate-smart recommendations rely on water availability, including reliable irrigation, effective scheduling, and predictable flooding and rainfall patterns, and require a transition from current water use practices. Therefore, better insights are needed on local water use, irrigation reliability, and the practices or decisions of farmers. Despite their importance, current water-use patterns remain insufficiently understood and incorporated in climate action plans. This policy brief presents findings from research to build a water-based framework to inform Cambodia’s adaptation and mitigation efforts. Our goal is to support policy and investment decisions that enhance the impact of climate action. The study focused on Takeo province, in the Cambodian Mekong floodplains. We used existing climate-risk and adaptation maps to guide a landscape analysis across nested scales. This was combined with a household survey, stratified by key climate vulnerability zones. Our findings highlight three main insights: 1. Rice cropping systems, whether single, double, or triple, differ across regions based on local rainfall, flood exposure, and irrigation access. These differences create real constraints on the flexibility to shift cropping calendars. Infrastructure and management interventions must reflect these local realities and anticipate future changes under a changing climate. 2. Farmers manage irrigation primarily to reduce risk, often by keeping standing water in their fields as insurance against dry spells. In practice, 70% of farmers decide to irrigate when surface water disappears. While this is understandable from a risk perception, it leads to higher irrigation costs, lower water productivity, and greater greenhouse gas emissions. There is potential for AWD to replace current practices and deliver multiple benefits. 3. Most irrigation water is delivered through small diesel pumps, with highly diverse costs, reflecting inequality in irrigation access and poor service reliability. Over 30% of farmers report irrigation delays due to water shortages, labor gaps, or pump failure. These issues impact rice production and limit system resilience. To address these constraints, climate financing should focus on local solutions. Improvements should not only target canal water control but also address pump systems that lift water from the canals. Promising actions include improving pump selection, expanding support for pump repair services, and scaling service provision models. Small water impoundments could also increase buffering during droughts or delivery delays. In parallel, further research is required to examine how shifting rainfall patterns and associated climate risks influence cropping-system timing, rice productivity, and opportunities for diversification and value-chain upgrading. A coordinated agro-hydrological research and investment program is essential to scale these diagnostics and innovations nationally, support long-term resilience, and foster high-value agriculture aligned with ecosystem service needs. 4 RESEARCH APPROACH A preliminary study was conducted in Takeo Province to gain contextual insights into local irrigation practices. Data collection involved key informant interviews with 19 stakeholders, including two government officials, representatives from three community-based water user associations, and fifteen rice farmers. In addition, field observations were carried out to understand the physical and operational characteristics of the irrigation infrastructure. This combination of qualitative methods provided a grounded overall understanding of the prevailing irrigation patterns within the province’s rice-based agricultural systems. Building on these initial findings, a structured questionnaire survey was employed to collect detailed data on irrigation water use across climate risk zones (Box 1) of the Takeo province. The survey captured in-depth information on cropping seasons, water sources, the temporal scale of water availability, irrigation practices and farmer perceptions, irrigation water use quantification, and pumping mechanisms. To ensure representativeness, one commune was randomly selected from each of the 23 climate risk classes in Takeo Province. Within each selected commune, ten rice farmers were randomly chosen, yielding a total sample size of 230 respondents. The face-to-face survey using digital tool (Kobo Toolbox) was administered over the first two weeks of December 2024. Figure 1: Climate-Smart Map showing climate risk classes in Takeo province BOX 1: Climate smart maps (CS-Maps) The CS-Maps – a tool to inform climate risk classes and climate smart strategies – was developed by the International Rice Research Institute (IRRI) using a participatory mapping approach in complement with climate and terrain data. The Ministry of Agriculture, Forestry and Fisheries, Cambodia collaborated with IRRI in developing CS-Maps in four provinces in Cambodia including Takeo. Figure 1 presents the CS-Map for Takeo province under the potential drought risk for rice in a normal year. These maps serve as a decision-support tool to delineate climate risk zones within selected provinces and propose tailored, zone-specific adaptation strategies to inform climate-resilient agricultural planning. In a normal year, areas with the highest climate risk for rice cultivation are indicated in red, while those with the lowest risk are marked in green. Each color-coded zone corresponds to a distinct climate risk class; with a total of 23 risk-classes, each associated with a specific adaptation strategy. Suggested adaptations include cropping calendar adjustments, crop diversification, animal husbandry, aquaculture, irrigation infrastructure upgrades, and new water source construction (IRRI, 2024). Box 1: Climate smart maps (CS-Maps) 5 RESEARCH INSIGHTS Why do farmers have dynamic seasonal definitions? Section highlights: Variability in how farmers define cropping seasons is shaped by differences in water availability, micro-elevation, flooding, and related factors. Recognizing these local conditions is key to effective adaptation efforts such as planting date adjustments. In Cambodia, rice cultivation primarily occurs across two main cropping seasons: the dry and the wet season. The wet season is further subdivided into early wet, main wet, and late wet seasons in Takeo province. These classifications are shaped not by fixed calendars, but by water availability informing crop establishment decisions. The survey results reveal that 44% of farmers cultivated during the dry season, 28% during the early wet season, 60% during the main wet season, and 58% during the late wet season. However, there is no uniform agreement among farmers on the precise timing or duration of these seasons (Figure 2). The most preferred seasonal definitions in Takeo are:  Dry season as December–March or January–April  Early wet season as May–August,  Main wet season as June–December, and  Late wet season as September–December. Between 20% and 50% of farmers operate outside conventional seasonal definitions, which creates a significant risk of misalignment if climate-smart programs adhere to rigid seasonal calendars or inflexible recommendations regarding seasonal changes. The specific seasonal definition in rice fields is influenced by multiple factors, including water access, rainfall, differences in micro-elevation, flood exposure, cropping history, risk perception and experience, crop variety selection, market dynamics, labor availability, and household priorities. This reflects the diverse water access conditions across the province and highlights the dynamic and location-specific nature of cropping decisions. Furthermore, it affects the timing of water delivery, crop inputs, and risk communication. The inconsistency in seasonal definitions poses a challenge; however, mapping structural patterns and tailoring climate-smart policies to these patterns offers a practical solution. When climate policies or programs adopt generalized seasonal categories without accounting for localized realities, they risk being interpreted differently by farmers, ultimately reducing efficiency, uptake, and overall impact. Accordingly, strategic climate-smart interventions are needed to align with the core determinants of cropping seasons (Table 1). Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec % 10.00 10.00 4.78 4.35 3.04 2.61 1.30 1.30 0.87 0.87 0.87 0.87 0.87 0.87 0.43 0.43 0.43 0.43 Dry season Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec % 7.39 6.96 2.17 2.17 1.74 1.30 0.87 0.87 0.87 0.43 0.43 0.43 0.43 0.43 0.43 0.43 0.43 0.43 Wet season Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec % 15.22 10.43 6.52 4.35 3.48 2.61 2.17 2.17 2.17 1.74 1.74 0.87 0.87 0.87 0.87 0.87 0.43 0.43 0.43 0.43 0.43 0.43 0.43 0.43 Early wet season Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec % 9.57 8.26 6.96 6.52 6.52 4.78 1.74 1.74 1.30 1.30 1.30 0.87 0.87 0.87 0.87 0.87 0.43 0.43 0.43 0.43 0.43 0.43 0.43 0.43 Late wet season Figure 2: Farmer defined cropping seasons and percent respondents for each definition 6 Table 1: Key entry points for incorporating seasonal differences into climate-smart planning: Addressing leading factors that define agricultural seasons Season-Defining Factor Entry points for climate-smart planning Expected outcome Rainfall timing and reliability  Make weather forecasting easily accessible for farmers  Promote advisory systems that translate forecasts into actionable decisions  Pre-preparedness for rainfall variabilities & make climate advisory more adoptable Water access and irrigation control  Invest in long-term water resource development plans  Upgrade and maintain irrigation infrastructure  Build institutional and farmer capacity for field level water governance  Enhances reliability of irrigation water sources Micro-elevation differences across paddy fields  Tailor climate advisory to local topography  Bring more adaptable climate smart recommendations Flood exposure (timing, duration, and extent)  Improve access to reliable, area-specific flood forecasts - Integrate flood risk zones into seasonal planning and crop insurance schemes  Pre-preparedness for flood variabilities & make climate advisory easily adoptable Risk perceptions and past experiences  Promote risk literacy and awareness campaigns  Strengthen access to credit and safety nets for climate-related shocks  Efficient and productive water use Crop variety selection (short-, medium-, long- duration)  Ensure availability and marketability of diverse rice varieties - Strengthen seed systems and bring awareness to market chains to support variety choices adapted to seasonal variations  Bring market chains to enable incentives for climate smart decision making Cropping history and land-use patterns  Encourage crop rotation and diversification strategies that align with seasonal conditions  Support field-level data systems for historical cropping records  Facilitate climate-smart and profitable farming systems Market dynamics (input access, product pricing)  Improve market information systems  Improved farm profitability Labor availability (seasonal constraints)  Promote labor-efficient climate-smart technologies  Support mechanization and service delivery models during peak labor shortages  Enhance agriculture labor timely available Household priorities  Integrate gender- and youth-responsive planning in seasonal adaptation strategies  Inclusion and social sustainability 7 Farmer perceptions shaping irrigation practices This section highlights two key findings: (1) irrigation practices are often inconsistent and delayed, and (2) many farmers view irrigation as a risk management strategy and do not consider water-saving or greenhouse gas emission reduction techniques, leading to high water use and high irrigation costs. Rice farming in Takeo Province consists of both irrigated and rainfed systems, with 74% of surveyed farmers relying on irrigation. Delayed irrigation emerged as a critical constraint, with 30% of irrigated farmers affected. This directly affects cropping schedules and input application, raising concerns about planting synchrony, pest risks, and farm-level efficiency. While direct yield impacts were not measured, such delays contribute to overall variability in productivity and profitability. The primary causes of these delays include untimely water availability (35%), labor shortages (20%), and pump breakdowns (18%). These findings point to an urgent need for targeted infrastructure or service delivery interventions. To meet farmers’ needs for timely water access, future investments should explore new and reliable irrigation sources, such as sustainable groundwater development, as part of a long-term water resource strategy. In the Mekong Delta considerations for transboundary ground water management are being discussed, such investments should also consider these (FAO, 2025). In the short term, interventions could focus on making high-quality water pumps and spare parts more accessible by trainings of local repair shops to strengthen market supply chains. In parallel, to address labor constraints, policies should support adopting labor-efficient irrigation services through community, public, and private sector modalities. Decision-making around irrigation timing also reveals another important behavioral insight: Survey results show that 71% of irrigated farmers decide to re-irrigate when ponding water is no longer visible on the soil surface, often leading to over-irrigation and continuously saturated fields. This practice prevents proper drainage in the root zone, highlighting strong potential to introduce AWD as a practical, yield-safe strategy to save water and reduce emissions. Despite the well-recognized importance of efficient water use, 98% of farmers in the study area do not perceive a need for water- saving practices. This perception is driven by several factors: 27% believe water is abundant, 30% use excess water as a weed control method, and 36% intentionally over-irrigate as a precautionary strategy against anticipated water shortages. This behavioral insight, combined with high irrigation costs, represents one of the most significant opportunities for climate-smart gains in the region. Since Takeo heavily relies on mechanical pumping, the irrigation application has two major downsides: it significantly raises input costs, reduces profitability, and contributes to carbon emissions. The results revealed that irrigation pumping costs can be substantial, reaching up to 589 USD/ha, and may account for as much as 23% of a farmer’s total profit (see Table 2). Table 2 also reveals a wide variation in irrigation costs, ranging from as little as 2 USD/ha to as much as 589 USD/ha. At the same time, the majority (65 – 80%) of the irrigation events use only one pump to irrigate the same plot, while others use two pumps. This variability highlights stark disparities in access to low-cost and low-emission irrigation and energy and reflects unequal exposure to input-related profit risks, especially for low-margin farmers. Such cost variation underscores the urgent need for targeted support in high-cost zones, alongside efforts to:  expand access to efficient pumping (pumps, spare parts, and relevant services, trained mechanics);  facilitate capacity building for low-emission/low-cost energy alternatives  encourage local irrigation service providers as a means of reducing production costs and increasing climate resilience. “ Water isn’t something we save—when it’s there, we use it as much as needed. You never know if there’ll be any tomorrow.” 8 In parallel, promoting integrated weed management methods and introducing climate-smart risk mitigation approaches such as AWD can help shift farmer perceptions toward more efficient and sustainable water use. Table 2: Reported irrigation water pumping costs in Takeo across different seasons Pumping cost (USD/ha) Pumping cost as a percentage of total profit (%) Dry Season Early wet season Long wet season Late wet season Dry Season Early wet season Long wet season Late wet season Maximum pumping cost 439 589 196 554 23 9 15 23 Minimum pumping cost 2 5 29 11 0.14 0.87 0.15 0.03 Median pumping cost 112 59 52 70 1.5 1.4 1.2 2.2 Water Access and Source Selection: A Dynamic and Adaptive Process Section highlight: Dynamic nature of water source requires improved and agile delivery methods Effective climate-smart water use planning must begin with a thorough understanding of the dynamic and complex interactions that shape current irrigation practices: water sources, source characteristics, delivery mechanisms, scheduling, water availability in terms of time and quantity, and irrigation water levels. Such contextual understanding forms the foundation for climate-smart recommendations that are practical, scalable, and readily adoptable by farming communities. In Takeo Province, rice farmers typically rely on one or a combination of the following irrigation systems:  Rainfed systems: Plots depend solely on rainfall, with water retained or drained based on bund height. Some areas also include drainage canals to remove excess water during the rainy season.  Canal irrigation: In areas near the river, an extensive canal network delivers water from the Mekong River and its tributaries. These canals are regulated by gates. Mechanical pumping is widely used—either to lift water from canals to fields or to move it from main canals to sub- canals—making it the main water control method before gravitational flows.  Reservoir-fed irrigation: Rainwater is collected in reservoirs and distributed through canal systems. Some of these reservoirs also contain large drainage canals, managing substantial runoff during the rainy season.  Groundwater use: Although not prominent overall in Takeo, localized groundwater pumping exists in specific areas. Water source selection is not static; it is an adaptive process shaped by access, seasonal availability, and reliability. Some farmers use fixed sources and delivery methods for each plot throughout the year. However, 30% of the farmers shift between water sources and methods at least once during a cropping year. In fact, fixed water sources do not ensure a reliable irrigation supply and may indicate a limited availability of alternative water-source options. Water delivery performance also varies across time and space, influenced by source type, distance of plots to water sources, canal water level (e.g., gravity flow becomes possible when canal water levels are high), micro-topography, and irrigation service access. Climate-smart policy design must recognize and address the variability of local conditions. It is crucial to identify the specific irrigation scenarios that a proposed intervention aims to tackle and to tailor strategies accordingly. Policies and investments can increase their effectiveness if they cater to these local dynamics and target context-specific solutions that can fill these gaps to increase irrigation service reliability and performance. Figure 4: Extreme water withdrawal from canal amid seasonal water shortage 9 10 POLICY IMPLICATIONS AND RECOMMENDED ACTIONS To ensure that climate-smart agriculture interventions in Cambodia are grounded in field realities of the current water management and irrigation systems, we propose attention across the following three levels: the landscape, the irrigation system, and the farm. Across these three levels, the following specific recommendations provide entry-points for efforts towards sustainable rice landscape and agricultural and food system transformation: Location-specific climate adaptation plans tailored to diverse agro-hydrology of the country:  Develop spatial zoning for a high-value and resilient agricultural development strategy. This strategy should consider irrigation accessibility, high-value crops and value chains in water insecure areas, as well as highly productive rice areas that also provide ecosystem services such as flood regulation.  Develop and deploy targeted risk management strategies for rice farmers across different rice ecologies, irrigation practices, and risk zones (floods, drought). Promote good agronomy practices, especially in low-risk zones, to support high-value production and integrate these with weather advisory services. Strengthen irrigation infrastructure and service reliability:  Prioritize investments in canal maintenance, participatory management, and pump stations to increase irrigation reliability and performance and reduce irrigation delays.  Explore sustainable groundwater development where feasible, especially for bridging dry-season water gaps  Enable supply chains for irrigation technology to expand irrigation delivery through incentives for local dealers, service providers, or rural repair hubs.  Introduce and co-finance smart pumps or community-based irrigation services to address labor shortages and reduce delays in irrigation. Promote low-emission rice irrigation as a low-cost and resilient farming practice:  Develop and deploy awareness campaigns for low-emission and climate-resilient irrigation techniques in rice to reduce emissions and save irrigation water e.g. simplified alternate wetting and drying (AWD)  Highlight farmer-preferred benefits of climate-smart irrigation techniques (e.g., simplified AWD) to reduce costs and shield farmers from price volatility (e.g., fuel for pumps), and risks such as lodging Landscape Climate adaptation plans that take into account the diverse agro-hydrology of the country. Irrigation system Strengthen irrigation infrastructure and service reliability and performance. Farm Promote low-emission rice irrigation (e.g. AWD) as a low- cost and resilient farming practice. 11  Ensure communications with farmers tailored to the local language and experience to enhance understanding and uptake, especially to avoid misunderstanding between different types of flooding, waterlogging, drying, and drainage. Authors Jayasiri, M.M.J.G.C.N., International Rice Research Institute, Phnom Penh, Cambodia; n.jayasiri@cgiar.org Flor, R.J.– International Rice Research Institute, Phnom Penh, Cambodia; r.flor@cgiar.org Keo, Sokheng – International Rice Research Institute, Phnom Penh, Cambodia; S.Keo@cgiar.org Chuong, Thort – International Rice Research Institute, Phnom Penh, Cambodia; t.chuong@cgiar.org Then, Rathmuny – International Rice Research Institute, Phnom Penh, Cambodia; t.rathmuny@cgiar.org Urfels, A, International Rice Research Institute, Los Baños the Philippines; Wageningen University and Research; Cornell University; a.urfels@cgiar.org Citation Jayasiri, M.M.J.G.C.N., Flor, R.J., Keo, S., Chuong, T., Then, R. Urfels, A. (2025) Integrating irrigation realities into climate solutions. Policy Brief. International Rice Research Institute, Philippines Acknowledgement Takeo Provincial Department of Agriculture, Forestry, and Fisheries (PDAFF), Provincial Department of Water Resources and Meteorology (PDOWRAM), All the village leaders and farmers in Takeo province for their support during field work. This work was supported by the Asian Mega Delta (AMD) initiative, the OneCGIAR Sustainable Farming Program, and OneCGIAR Scaling Program. The authors are grateful for the support of the CGIAR Trust Fund contributors. See www.cgiar.org for more details. References Adil, L., Eckstein, D., Künzel, V., & Schäfer, L. (2025). Climate Risk Index 2025: Who suffers most from extreme weather events? Germanwatch e.V. https://www.germanwatch.org/en/cri IRRI. (2024). Climate-Smart risk maps and adaptation plans for Cambodian Mekong Delta. International Rice Research Institute. FAO. (2025). New GEF project in preparation to improve sustainable use and governance of groundwater in Cambodia-Mekong River Delta Transboundary Aquifer. Food and Agriculture Organization of the United Nations. https://www.fao.org/cambodia/news/detail/New-GEF-project- in-preparation-to-improve-sustainable-use-and-governance-of-groundwater-in-Cambodia-Mekong- River-Delta-Transboundary- Aquifer/en#:~:text=With%20a%20budget%20of%20USD%C2%A015,and%20Ho%20Chi%20Minh %20City WBG. (2024). Climate Risk Country Profile: Cambodia. World Bank Group (WBG). https://climateknowledgeportal.worldbank.org/sites/default/files/country-profiles/16814- WB_Cambodia%20Country%20Profile-WEB.pdf mailto:n.jayasiri@cgiar.org mailto:r.flor@cgiar.org mailto:S.Keo@cgiar.org mailto:t.chuong@cgiar.org mailto:t.rathmuny@cgiar.org mailto:a.urfels@cgiar.org http://www.cgiar.org/ Farmer perceptions shaping irrigation practices Water Access and Source Selection: A Dynamic and Adaptive Process