Farmers in Madhya Pradesh’s water-scarce Dewas district have built an estimated 30,000–35,000 farm ponds since 2006, improving rainwater storage and irrigation security.
The ponds have helped farmers diversify into wheat, oilseeds, chia seeds, dairy farming and fisheries, generating substantial additional income.
The initiative began with support from the district administration, which provided technical guidance and helped farmers access bank loans.
However, smaller farmers often cannot afford to set aside cultivable land for ponds, leaving the benefits unevenly distributed.
Despite widespread rainwater harvesting, groundwater extraction remains high, raising questions about the model’s long-term sustainability.
Pope Singh of Nipania village says he catches about 50 tonnes of fish, worth nearly Rs 1 crore, every alternate year. Balramji Patel of Khoob village says he sold chia (Salvia hispanica) seeds worth Rs 23 lakh last year. Subhash Jalotiya of Dhaturiya village says he sold wheat worth Rs 60 lakh in 2025. “Farmers have bought four-wheelers, taken up dairy farming with improved breeds of cattle, even built houses with the earnings generated from their ponds,” says Ramesh Singh Khinchi, a farmer from Tonk Kala village.
Since 2006, farmers of Madhya Pradesh’s Dewas district have built thousands of farm ponds, which have opened new avenues for farm income, besides helping retain rainwater, secure irrigation and expand agriculture. There is no official figure for the number of farm ponds built, but experts put the count at around 30,000. Some estimates say the number exceeds 35,000, partly because different assessments cover different years, areas and categories of water- harvesting structures.
Situated in the perpetually water-scarce Malwa region, Dewas faced acute water shortage till about two decades ago. But things started to change around 2006 when the local administration took steps to assist farmers set up ponds on their land. “I would tell village residents that before asking what the government can do for you, ask what can the village do for itself,” says Umakant Umrao, district collector of Dewas from 2006 to 2008, who is now retired. “His initiative encouraged us to set aside part of our land for farm ponds,” says Manoharlal Mandloi, a farmer lea-der based in Padlia village, Dewas.
Prior to the pond construction, tubewells were the main source of irrigation. A study by International Water Manage-ment Institute (IWMI), a non-profit based in Colombo, Sri Lanka, found that farmers in the district had 3,887 tube wells in 1990. By 1998, the number had risen to 14,172—more than fourfold in eight years. By 2009-10, the district had 23,119 tube wells and 36,531 wells. About 82 per cent of its irrigation depended on groundwater.
The result was a large-scale, unchecked extraction of groundwater. The Central Ground Water Board’s (CGWB's) 2023 assessment puts the district’s groundwater extraction at 79.99 per cent and classifies Dewas as overexploited. In Dewas block, the groundwater extraction is 110.88 per cent, which means the annual groundwater extraction exceeds the as-sessed annual recharge.
“As Dewas became increasingly dependent on groundwater, tube wells went deeper, costs and risks rose, while rain-water continued to run off without adequate recharge,” says Kailash Kundal, a former member of the state Legislative Assembly from Khategaon constituency, Dewas. “The only change we brought about was the decision to stop letting rainwater run off, and hold it on the farm,” says Umrao.
To aid farmers in building ponds, the administration provided technical advice and helped farmers access bank loans. Large farmers were encouraged to allocate one-tenth or one-fifth of their land, creating wider benefits. Technical experts monitored the work while the administration stayed active in the fields. The campaign’s message was simple—“Save wa-ter, earn profits”. Beginning with 40 large farmers in 2006, the initiative soon led to construction of around 2,000 farm ponds within the next 14 months. Dhaturiya and Gorwa villages alone recorded more than 200 ponds. IWMI’s AgWater Solutions project found that water-harvesting structures occupied an average of 10.04 per cent of cultivated land in Khategaon and 2.2 per cent in Tonk Khurd. Early ponds covered about 0.2-1.6 hectares and were mostly 9-12 m deep.
“Rainwater stayed on the farm, collected in ponds, and some seeped into the ground, potentially contributing to groundwater recharge. The ponds strengthened rainwater harvesting and irrigation security,” explains Sunil Chaturvedi, a Dewas-based hydrogeologist. As ponds filled, wells regained water, fields turned green, and farmers began growing a second crop. Wheat, oilseeds, chia seeds and fish farming strengthened farm incomes. Because farmers owned both the land and the ponds, they developed a stronger sense of ownership and responsibility towards water and its usage.
The scale of water storage in the district has since become striking. Dhaturoia village reportedly has more ponds than households. Mahavir Singh of Tonk Khurd owns what is regarded as the largest and deepest pond in Dewas, which continues to hold water through the peak summer months, helping ease sea-sonal water stress. “Our objective was not merely to build ponds, but to make farmers see their water as their own responsibility,” says Umrao.
Despite its success, the model has exposed a structural limitation. Investment depends on land ownership, credit and the ability to absorb a temporary loss of cultivable land—resources that are not equally available to all farmers. Its success among larger farmers has also raised a crucial question—how to implement the idea for marginal farmers? The economic gains from the pond model are also not necessarily distributed equally. A large farmer may be able to sacrifice 10-20 per cent of cultivated land for a pond because the remaining land can still generate substantial income. "For someone with 2 acres (0.8 ha), a pond is not a tenth of the farm, it might be a third of the family's food security," says a farmer in Tonk Kala who has held off building a pond for exactly this reason. That makes the real test of the model less about whether ponds raised aggregate farm income and more about whether farmers with the smallest holdings can gain reliable access to water without sacrificing the land on which their livelihoods depend.
For the next phase of the Dewas story, the question is no longer about the number of ponds built, but what happens to the aquifer beneath them. A meaningful assessment must compare historical and current groundwater levels, pumping depths, annual extraction, rainfall, cropping intensity and the spatial spread of ponds. Without this, the number of ponds demonstrates the scale of conservation, but not groundwater sustainability.
CGWB's 2025 Dynamic Ground Water Resources Assessment's 2023 baseline groundwater extraction rate of 110.88 per cent for Dewas block, and about 80 per cent for the district should serve as a reference point against which newer data is as-sessed. The January 2025 CGWB bulletin shows shallow groundwater levels at 2-5 m below ground level in certain parts of Dewas, but these single-year figures cannot establish ponds’ impact without historical trends.
“The Dewas model has provided practical, on-ground answers to water availability, irrigation security, farm incomes and farmer participation. Now, the questions before us are: how much has water level in wells risen due to surplus water stored in ponds; whether pumping depths have decreased; how extraction has changed alongside increased cropping intensity; and to what extent benefits have reached small farmers,” says Shashi Poonam Indwar, Scientist D, National Institute of Hydrology, Bhopal. Dewas' story has critical lessons on how to hold water. The harder test is whether the district can recognise the limit of what it holds. “The next test may not be how much water it can hold, but how wisely it can use what it has,” Umrao says.
This article was originally published in the October 1-15, 2026 print edition of Down To Earth