Forty check dams, 220 hectares of sugarcane, and four villages with dry wells
A 4,200-hectare watershed in a 550 mm rainfall taluka of Marathwada was treated between 2010 and 2014 with 40 loose-boulder check dams, 18 farm ponds and contour trenching, at a cost of Rs 6.2 crore. Between 2015 and 2023, dug wells in the upper two villages rose by 3.5 m, and 220 hectares shifted to sugarcane and banana, mostly on drip. The lower four villages report wells going dry by April; the percolation tank at the outlet has not filled beyond 30 per cent since 2019; and the last 2.5 km of the stream is now dry through the year. A review is being held under the state watershed guidelines, which require benefits to be shared across the watershed. Sugarcane in the area uses about 1,900 mm of water and jowar about 500 mm. Two sugar factories within 30 km offer harvest contracts and drip subsidies to farmers who switch.
1. The project file describes this as successful watershed development. How can the same numbers support the opposite conclusion?
Read the worked answer
Because a rising water table proves recharge happened, not that water was conserved. The check dams deliberately slow the stream and push water into the upper aquifer. Farmers then respond rationally: they dig wells, install drip, and plant a crop that pays more per hectare. Drip raises efficiency per hectare but does not cap total use, so a 220-hectare shift to a crop needing 1,900 mm is a large new consumptive demand sitting on top of the recharge that was created. The upper two villages are effectively being subsidised to intercept water that used to move down the stream.
The lower villages' symptoms fit: a percolation tank that no longer fills, an April dry-out, and a stream dry for 2.5 km are signs of reduced baseflow, not of a failed monsoon alone.
Before concluding, two controls are needed. First, a water balance: 4,200 hectares at 0.55 m of rainfall is about 23 million cubic metres a year, of which most leaves as evapotranspiration; the question is how the remainder is now partitioned. Second, a control: long-term hydrographs from an untreated neighbouring watershed, and a check on whether April dry-out predates 2010 and simply tracks the 2019-2023 dry years. Without those, the case is strong but not proven.
2. What institutional rules would you propose for sharing the water, and what must you know before restricting sugarcane?
Read the worked answer
The rules should operate at the watershed level, not the farm level. First, define a minimum downstream release from the percolation tank and a no-new-check-dams limit on the upper reach until baseflow recovers. Second, shift future recharge work to the middle and lower reaches, so recharge happens where the wells are shallow and the stream still runs. Third, move from crop bans to water budgeting: publish an annual water budget for the watershed, allot a volumetric ceiling per well on the notified aquifer, and tie drip subsidy to a cap on area under cane rather than to the switch alone. Maharashtra's groundwater law does provide for notifying stressed aquifers and regulating extraction and crop plans, and this is the instrument to use. Fourth, convert farm ponds from private storage into managed recharge structures with desilting and defined filling rules.
Before restricting cane, I would need a well census and the area-irrigated data per village; the number of households whose income depends on cane and how many are tenants or labourers; the length and terms of the factory contracts, since a five-year contract is a different problem from a spot purchase; and evidence on whether drip actually reduced total use or expanded the area. A restriction imposed without that will be evaded and will hit the smallest farmers hardest.
3. The upper villages argue they are being punished for investing in their own land. How would you answer, and what would a fair compensation or transition package look like?
Read the worked answer
The argument is not frivolous. Those farmers took a subsidy and a loan on the promise that watershed treatment would raise their yields, and the rules should not retroactively criminalise a crop choice that the same state departments encouraged.
But the counter is equally clear: the rainwater that recharged their wells fell on the whole 4,200 hectares, and the downstream villages have as much claim to it as the upper ones. Capturing recharge is not the same as owning it. The equitable principle is that everyone in the watershed should reach at least the pre-project level of access before anyone expands.
A fair package would include: a guaranteed minimum flow to the lower villages, monitored and published; compensation for the lower villages where wells failed, funded from a watershed-level water fund rather than from the upper farmers alone; a shift of the drip and crop-diversification subsidy towards lower-water crops and protective irrigation, with the sugar factories asked to share the cost through contract terms; help with debt rescheduling where cane loans were taken on state advice; and a transition period of three to five years so that perennial crops already planted are not simply written off. Punitive enforcement without a transition will fail, and so will a transition without enforcement.