How to Lose Three Quarters of a Lake Without Anyone Draining It

There is no dramatic moment in the disappearance of Vellayani Lake. No barrage was closed, no channel was cut, nobody decided to drain it. It simply became smaller, every year, for a century, and it is still doing so.

The numbers, and why they disagree

A geospatial reconstruction of the water spread reports 558.93 hectares in 1973, 267.89 hectares in 1992 and about 222 hectares in 2019. A separate line of accounts begins earlier and higher — roughly 750 hectares in the 1920s, of which about 165 hectares belonged to what became Kerala Agricultural University and the rest to private holders — and continues to roughly 184 hectares by 2024.

Those two series cannot both be exactly right, and it is worth being honest about why rather than picking the more alarming one. A lake like this has no hard edge. It has open water, then seasonal marsh that is water for four months a year, then wet paddy that was marsh a decade ago, then dry paddy, then houses. Every study has to choose where the lake stops, and moving that line by a hundred metres around a shallow basin moves the total by tens of hectares.

What survives the disagreement is the shape of the curve: a steep loss through the second half of the twentieth century, and a continuing loss since. Between half and three quarters of this lake is gone.

The ratchet

The mechanism is a ratchet, and each click is individually defensible. A shallow margin dries for part of the year and is treated as seasonal land. Someone bunds it to hold a crop. The bund makes it dry for longer, so the next owner fills it another half-metre, at which point it is buildable. When the boundary is next surveyed it is drawn where the water actually reaches — which is now inside the old line — and the newly exposed margin becomes the shallow edge for the next round.

Nothing in that sequence looks like destroying a lake. Each step is a farmer or a householder making an ordinary decision about land they hold. And the ratchet only turns one way: a filled plot is never returned to marsh, because the value moves in one direction too.

What is lost is not mainly area

The headline is always hectares, but the consequential loss is depth. Siltation from a catchment that has been progressively built over settles across the whole bed, and a shallow lake that loses a metre of depth across its remaining area loses far more storage than the surface figure implies.

Storage is what a lake is for, in the dry months. A wide, shallow, weed-covered sheet in March holds much less usable water than a smaller, deeper one, and it evaporates faster and heats further, which in turn favours the weed. That is why the Irrigation Department’s de-silting programme, running since around 2019, is framed around raising the mean depth by about 1.5 metres rather than around recovering the old boundary. The boundary is not coming back. The depth might.

Why it is hard to stop

Because there is no defendant. Enforcement against a wetland boundary requires proving where the boundary was, against holders with documents, over encroachments made incrementally across decades by people who have since sold. Kerala has wetland and paddy conservation law; applying it retrospectively across a basin of hundreds of small holdings is a different proposition from applying it to a single reclamation.

It is also hard because the loss is invisible from the road. Vellayani in the early morning is beautiful. The wide green margins that photograph so well are, substantially, the lake — converted. A visitor sees a landscape; the satellite record sees an absence.

Why it now matters more than it did

A shrinking rural lake is a slow, local loss. A shrinking lake with a deep-water port drawing on its supply is something else. Vellayani feeds irrigation, the Kerala Water Authority’s local scheme, and the drinking-water supply for the Vizhinjam International Seaport — which means its condition has stopped being a village matter.

That is the argument the restoration is really making, and it is a more durable one than aesthetics. A lake protected because it is pretty gets protected as far as the budget allows. A lake protected because a port needs it tends to get protected.