Terai Arc Landscape

Map key

Indicative range from habitat rules — not sightings

Terai Arc Landscape outline

Terai Arc reserves13 of 15 drawn, and 3 more

Priority speciestap a species ↗

White-rumped vulture

Bengal florican

Bengal tiger

Gharial

One-horned rhino

Swamp deer

Gaur

Leopard

Asian elephant

East–West Highway & trunk roads

Katerniaghat WS and Jhilmil Jheel CR are members of the landscape and are named but not drawn: OpenStreetMap carries no boundary for either, and a forest-cover polygon is not a gazetted boundary. Tap the label to read why. Kishanpur, Nandhaur and Sonanadi are drawn too, as the lead author’s build draws them; SM Table 1 does not list them.

Human settlement · 2020

People per 100 m cell

scatteredtown

Terai Arc reserves

The Terai Arc carries among the highest rural human population densities on Earth — about 550 inhabitants per km² — on the same alluvial flats that carry its ungulate biomass. Nepal legally recognised buffer zones in 1996, with parks sharing 30–50% of revenue. Density figure: the paper's Supplementary Material, Appendix A (the authors' revision received 2026-09-29). Reading the glow. Nepal's five reserves are holes in it: under 1% of their area carries any. On the Indian side the grid fills some districts edge to edge, forest included — parts of Dudhwa, Suhelwa and Sohagi Barwa read as settled — so there it cannot tell a reserve from the farmland beside it (measured by us from the grid, 2026-09-30). Two national grids also meet at the border, so a step in the glow there is not a comparison. Population: WorldPop (University of Southampton), CC BY 4.0 — 2020 constrained (BSGM), Nepal, India and China.

Tiger corridors · least-cost model

Cheapest 10% of the landscape to cross

Next 10%

Next 10%, the buffer

Terai Arc reserves

Measured by us, 2026-09-27: least-cost corridors between the reserves after the 2004 design of the landscape, re-run on 2021 land cover. The land cover has no sugar-cane class, so the model is run twice: cropland costed as other crops, and as sugar cane, which tigers use as cover. Percentile bands of the cost surface, not observed tiger routes. Land cover: © ESA WorldCover project 2021 / Contains modified Copernicus Sentinel data (2021) processed by ESA WorldCover consortium

Altitude bands · inside the reserves

Shivalik crest, dry ridge900 m +

Shivalik / Churia slope600–900

Sal forest & bhabar350–600

Alluvial grassland · the phantas200–350

Floodplain & river< 200

Bands are altitude, not vegetation: sal forest grows on the valley floor and on the Shivalik slope alike, and the wet grassland this index cares most about is a floodplain feature rather than an elevation one. The tint is clipped to the reserves — grey ground is unprotected. The frame reaches the Greater Himalaya; no reserve does.

The Terai Arc Landscape

Fifteen national parks, wildlife sanctuaries and tiger reserves — ten in India, five in Nepal — across 51,002 km² and 900 km of the Himalayan foothills, managed as one tiger metapopulation. The pale line around them is the landscape’s own outline, as supplied to the paper’s authors by one of its leads. It measures about 82,400 km² on this map, against the 51,002 km² the paper states: the two delineate the Terai Arc differently, so both are given and neither is adjusted to match.

Landscape

51,002 km² SM Appendix A

Length

900 km stated, end to end

Reserves

15 10 India · 5 Nepal

Drawn here

13 OpenStreetMap

Those 13

10,946 km² we measured the rings

Same 13, stated

11,636 km² SM Table 1

Drawn extent

694 × 349 km we measured it

Adopted

2001 Nepal

Government-adopted, not a designation we invented — Nepal adopted the Terai Arc in 2001 as its first landscape-scale conservation initiative, conceived from a tiger dispersal model, and has run two ten-year strategy and action plans (2004–2014, 2015–2025) and approved a third, the TAL Strategy and Action Plan 2026–2035. In India the constituent tiger reserves are statutory under the Wild Life (Protection) Act 1972 as amended in 2006.

UN World Restoration Flagship, 2024 — one of seven named under the UN Decade on Ecosystem Restoration.

The monitoring is a government’s, not ours — Nepal’s Department of National Parks and Wildlife Conservation has run national tiger censuses in 2009, 2013, 2018, 2022 and 2026, and national rhinoceros counts in 1994, 2000, 2005, 2008, 2011, 2015 and 2021. India’s National Tiger Conservation Authority and Wildlife Institute of India run the quadrennial All-India Tiger Estimation, reporting the Indian Terai Arc within the Shivalik Hills and Gangetic Plains unit.

429 tigers in 2026, against 121 in 2009 — Nepal’s national census, across Chitwan, Bardiya, Parsa, Banke and Shuklaphanta (National Tiger Census 2026, DNPWC/MoFE, released 2026-07-29; earlier totals from the 2009, 2013, 2018 and 2022 national censuses). Every tiger counted in Nepal is a tiger in the Terai Arc.

Potential landscapes in India and Nepal — the paper counts “potential Priority Species Landscapes” by country. India: 85 larger than 1,000 km² (454,189 km² in total), 49 larger than 2,000 km² (402,937 km²) and 24 larger than 5,000 km² (326,426 km²). Nepal: 13 larger than 1,000 km² (58,379 km²), 10 larger than 2,000 km² (53,369 km²) and 3 larger than 5,000 km² (30,268 km²). Their outlines are not drawn, and the paper states no human-density rank for this landscape among them, so none is shown. SM-Table 4, derived from the WDPA (UNEP-WCMC), used for this paper by permission.

Why a landscape and not a park

One population, not fifteen. That is the premise the whole thing rests on: the landscape was conceived from a tiger dispersal model, so the fifteen are managed as one metapopulation and the corridors between them are the infrastructure. SM Appendix A: “Passive rewilding through existing corridors (tigers, elephants) and active translocation (greater one-horned rhinoceros) have re-established resident populations across several reserves, with further translocations expected.” Banke, the most recent of the fifteen at 2010, is one of them: Nepal counted 25 tigers there in 2022 and 51 in 2026 (National Tiger Census 2026, DNPWC/MoFE).

Three species live nowhere else. The greater one-horned rhinoceros, the swamp deer and the gharial occur in no country other than India and Nepal. For those three this landscape is not one place among many where they could be tracked — it is effectively one of two.

The index rose; its 2020 peak is one species. The paper’s Priority Species Index for this landscape runs 100.0 in 2012, 104.8 in 2014, 137.1 in 2016, 157.6 in 2018, 214.0 in 2020 — and 159.1 in 2022. The 2020 point rests on a single species’ survey: the dataset records one species behind it, and 214.0 is exactly the white-rumped vulture’s value from its 2019–2020 survey — in the paper’s words, “only the white-rumped vulture was surveyed during 2019–2020; therefore, the 2020 PSI value is based on this single species.” So the step from 214.0 to 159.1 compares one species with six; it is not a measured decline, and this page does not present it as one. Where a species’ most recent survey is carried forward to a later index year, that is the authors’ rule as they state it: carried forward from the previous survey, case by case, with the completed records in the Supplementary Material’s tables.

Interventions, dated. Rhino translocations to Bardia and Shuklaphanta from 1987. Buffer zones legally recognised in 1996, with parks sharing 30–50% of revenue with them. The landscape recognised in 2001. A Wildlife Crime Control Coordination Committee at national and district level in 2010. The paper’s own caveat travels with these: many interventions predate intensive multi-species monitoring, so this is association, not demonstrated cause.

Among 838. The paper counts 838 landscapes larger than 5,000 km² worldwide, 61.1 million km² in total, with a median of 16,199 km², across 501 ecoregions and 125 countries. At 51,002 km² the Terai Arc is a little over three times that median.

Boundaries: © OpenStreetMap contributors (ODbL), one ring per reserve — Shuklaphanta rel 10458351, Bardiya way 294497796, Banke way 1124634291, Chitwan rel 6083166, Parsa rel 4815272, Rajaji rel 8815465, Corbett way 682980216, Pilibhit rel 9474608, Dudhwa rel 9474607, Suhelwa rel 9302654, Sohagi Barwa rel 1765042, Valmiki rel 2640060, Hastinapur way 668611261. Note, and it matters: we DRAW with OpenStreetMap and STATE areas from the paper, and the two disagree reserve by reserve — Sohagi Barwa’s rings measure 42% under its stated area, Suhelwa’s 38% over, Chitwan’s 26% over. Those are digitisation and scope differences, not errors in either source, and every per-reserve card prints both figures side by side. No cross-national protected-area database was used as a source or as a cross-check. Two of the fifteen are named and not drawn: OpenStreetMap carries no protected-area boundary for Katerniaghat WS (only a landuse=forest polygon) or Jhilmil Jheel CR (only an unnamed leisure=nature_reserve). Landscape figures, species, categories and the index: the paper's Supplementary Material, the authors' revision received 2026-09-29 — Appendix A (landscape description), Table 1 (per-reserve survey data and areas), Appendix C (method), Appendix F (policy interventions). Credit line, exactly as registered in provenance.json — Priority Species Index, from the paper's Supplementary Material — the same line the Priority Species Index panel prints beside its measurement badge, repeated here because this is where the landscape figures are. Generated from that document, not transcribed by hand. IUCN Red List categories only, transcribed from the Supplementary Material; this page does not carry the individual assessments.

3.97% of the Terai Arc plate was never observed by any scene in the search window and is filled rather than imaged; those pixels are not mapped separately, so fine detail on that plate is not a measurement everywhere. Imagery: Contains modified Copernicus Sentinel data 2026 (the Terai Arc plate); Contains modified Copernicus Sentinel data 2023–2025 (the Sauraha inset) · Elevation: Copernicus DEM GLO-30 — credited in full above · Boundaries & water: © OpenStreetMap contributors (ODbL) · Protected-area boundaries: © OpenStreetMap contributors (ODbL) · An independent visualization by Landseed PBC. Not affiliated with, endorsed by, or produced for DNPWC or NTNC.