What Cities Must Keep for Leopards, Civets and Jackals to Move Through

Every night, across the edges of Pune, Mumbai, Bengaluru and Delhi, animals that most city residents have never seen are attempting to cross the city. A leopard slips under a highway overpass. A small Indian civet threads through a storm drain. A golden jackal trots along a dry nullah bed before a construction light stops it cold. These movements are not random wandering, they are the remnants of migration routes that predate the cities themselves, still being attempted by animals whose internal maps have not been updated for concrete.

A corridor is not a park. It is a connected strip of habitat, forest, scrub, grassland, riverbank, or some combination, that allows an animal to move from one patch of viable territory to another without crossing open ground where it is exposed to traffic, dogs, or people. The distinction matters because cities tend to fund parks and call the job done. A park with walls on all four sides is an island. An island cannot sustain a breeding population of any medium or large carnivore for more than a few generations. Genetic isolation follows, and then local extinction, quietly, without anyone noticing until the animal simply stops appearing.

A leopard's home range in fragmented urban-edge habitat runs between 15 and 50 square kilometres depending on prey availability, which means a single animal needs multiple connected patches to survive. Research from Sanjay Gandhi National Park in Mumbai, where leopards have been studied continuously for over two decades, shows that the park's leopard population persists partly because of the Aarey Colony green buffer and the Thane creek mangroves, neither of which is formally a protected area. Remove either and the effective corridor narrows to a point where road mortality becomes the primary cause of adult death. Leopards use culverts, dry riverbeds and dense lantana thickets as crossing points. They avoid lit areas and will detour several kilometres around a well-lit road rather than cross it. This means that streetlight placement along forest edges is not a neutral infrastructure decision.

Small Indian civets and golden jackals are useful indicator species precisely because they are more tolerant of human proximity than leopards. If civets are gone from an urban green patch, the patch has already crossed a threshold of disturbance that larger animals crossed long before. Civets are nocturnal, solitary, and rely on dense understorey, leaf litter, low shrubs, fallen logs, for cover and foraging. They are eliminated not by hunting but by habitat tidying: the municipal habit of clearing undergrowth, burning dry leaves, and maintaining parks as mown lawns destroys the structural complexity they depend on. Jackals, which are more adaptable, persist longer but shift their behaviour under pressure. Urban jackals in Bengaluru's Hesaraghatta grassland fringe have been documented moving almost entirely between 11 pm and 4 am to avoid human activity, a behavioural compression that reduces foraging time and increases stress-related mortality.

Four physical features determine whether a corridor functions or becomes a barrier. The first is canopy continuity: a gap of more than 30 to 50 metres of open ground is enough to deter most medium mammals from crossing, because open ground means exposure. The second is water: seasonal streams and nullahs are not drainage problems to be concreted, they are the linear corridors that animals have used for centuries, and their banks carry the dense riparian vegetation that provides cover. The third is wall permeability: compound walls along forest edges that run unbroken for hundreds of metres function as hard barriers. Gaps of 30 to 60 centimetres at ground level, or culverts kept clear of debris, restore passage for civets, mongoose, and smaller cats. The fourth is light discipline: high-mast lighting along forest-edge roads suppresses movement across those roads for the entire duration of the night. Shielded, downward-directed lighting that does not spill into adjacent vegetation reduces this effect substantially. None of these interventions requires relocating a city. They require treating the existing infrastructure as a variable rather than a fixed constraint.

The pattern of failure is consistent: a city identifies a green patch as ecologically sensitive, designates it as a reserve or a biodiversity park, and then permits road widening, a metro depot, or a logistics hub on the connecting strip that made the patch viable. Pune's Katraj-Sinhagad corridor has been documented as a functioning leopard movement route, and sections of it have been proposed for development in successive master plans. The pattern of partial success is equally consistent: Bengaluru's lake network, when maintained with natural bunds and vegetated margins rather than concrete, functions as a secondary corridor for jackals, mongoose and birds. Mumbai's Flamingo Sanctuary in the Thane creek mudflats survives because the mudflats are too difficult to build on, not because of active protection, but the outcome is the same. The animals that remain in Indian cities are there because some piece of the built environment failed to be fully built. That failure is the corridor.