The Corridor or the Carriageway

Framework.

• Wildlife (Protection) Act, 1972. Section 29 prohibits destruction in a sanctuary without a permit; Section 27 restricts entry. Critically, an elephant corridor is usually not a protected area - which is precisely why it is easy to nibble at. Elephant reserves are notified by states under Project Elephant, but the corridor itself has weak legal standing.
• Forest (Conservation) Act, 1980, as amended in 2023. Diversion of forest land needs central approval, but the amendment created exemptions for certain strategic and linear projects. The first question is whether this stretch falls within an exempted category or needs full clearance and compensatory afforestation.
• EIA Notification, 2006, entry 7(f). Highway widening attracts prior environmental clearance - Category A or B depending on length, right-of-way and whether it is a new alignment.
• National Board for Wildlife. A project within 10 km of a protected area boundary requires Standing Committee clearance. Connecting two protected areas almost certainly triggers this.
• Eco-Sensitive Zone notifications around both protected areas, and NGT and Supreme Court jurisprudence on linear infrastructure through corridors, crossing structures and night restrictions.

Gaps.

1. Two underpasses sited on roadkill data, which shows where animals die, not where they cross - two is almost certainly too few.
2. No geotechnical assessment of the 14 m cut.
3. Tree felling without a tree-by-tree enumeration, a translocation plan for the endangered patches, or a Ficus retention protocol.
4. No induced-traffic assessment: four-laning raises speeds, and speed kills more elephants than width.
5. No comparison with the alternative on total cost, including land, time, litigation and carbon.

Substantive point. Wildlife clearance is the only forum where the corridor's value is formally weighed against the road's. Your job is to make that weighing explicit, not to supply a signature.

Decision: reject the proposal as submitted; recommend approval of the existing alignment only with a redesigned mitigation package.

Realignment is not automatically greener - 19 km of new alignment, 22 hectares of private land, acquisition time and litigation will fragment something else and may still kill elephants. The better question is not where the road goes but how it behaves.

Conditions.

1. Crossings sized to the animal. A twelve-month camera-trap and dung/DNA survey before siting, then at least five structures - underpasses of 30-50 m width and 5-6 m clearance, plus overpasses where the road runs in cutting. Two underpasses on roadkill data is mitigation by gesture.
2. Speed governance. A hard 50 km/h limit through the corridor with rumble strips, reflectors and intelligent speed enforcement; night restriction for heavy vehicles during peak movement months if the survey supports it. Speed reduction is the cheapest intervention per life saved.
3. Reroute or tunnel the 1.4-km cut. the extra cost is smaller than a landslide that closes the only link for 48 estates and blocks the corridor.
4. No net loss. Compensatory afforestation with native species on equivalent corridor land, protection for the 40 Ficus, and a corridor conservation fund drawn from the project's CSR and CAMPA provisions.
5. Independent monitoring - the DCF, NHAI, both field directors and a wildlife biologist, reporting quarterly to the State Board for Wildlife.

The test for each condition. Does it reduce animal-vehicle collisions per crore of expenditure, and can it be verified? Signage alone fails both tests. Crossings plus speed control pass on the evidence. Plantation without monitoring fails the second.

What I would refuse to do. Approve two underpasses and a plantation plan and call it mitigation. That is the standard failure mode, and it is the one that ends up in a comptroller's report.

What I would insist on, and why.

1. Geotechnical investigation before any cutting - borehole data, slope stability analysis for June-September rainfall, and a factor of safety. If it falls below 1.5, the cut must be redesigned with benching, subsurface drainage and retaining structures, or avoided altogether by a short tunnel.
2. A construction window. No slope cutting between 1 June and 30 September. A cut slope left exposed through the monsoon is a landslide waiting for a date.
3. Instrumentation and triggers. Rain gauges and piezometers on the slope linked to the district control room, with a published protocol - for example, 100 mm of rain in 24 hours triggers closure of the stretch and diversion of traffic.
4. A published response plan. Diversion route, ambulance access through the ghat, a muck-disposal site outside the stream course, and a named nodal officer.
5. A pre-monsoon joint inspection and one mock drill before the first June.
6. Compensation and restoration obligations for any private land used for debris or diversion.

The reasoning. A landslide on a ghat cut is rarely a natural disaster. It is a design decision deferred to the monsoon, and the district's legal duty under the Disaster Management Act, 2005 covers mitigation and preparedness, not merely relief. The cheapest moment to mitigate is before the cut - after the debris has blocked the only road to 48 estates, every option costs ten times more and the options are emergency ones.

Where I would compromise. On the timing of instrumentation, not on the slope design. Instruments tell you when to close a road; they do not stop the hill from moving.