evidoria

← Back to browse

Good practice Imported

Loess Plateau Watershed Rehabilitation Project — Terracing and Grazing Bans Cut Sediment 57 Mt/Year (China)

China · Jixian County · See the China profile

Evidence: Observational / pre–post Top 14% 60/100 · Ask Evidence Copilot about this practice

World Bank-financed terracing, afforestation and grazing bans across 1.56M ha of China's Loess Plateau (1994-2002) cut Yellow River sediment by 57Mt/year and lifted farm incomes, though later research found the wider afforestation drive it inspired cut regional water yield ~43%.

-57 Mt/year
Annual Yellow River sediment discharge reduction (1994-2002)
427,000 to 700,000 tonnes
Grain output (1993-2001)
360 to 1,263 yuan
Farmer per-capita net income (1993-2001)
59% to 27%
Project-area poverty rate (1993-2001)
-43.3%
Regional water yield change (wider afforestation campaign) (2000-2015)

Details

Maturity
Established
Promoter
China Ministry of Water Resources / World Bank (IDA)
Period
1994-2002
Keywords
watershed management, soil and water conservation, agriculture, afforestation

Context

China's Loess Plateau Watershed Rehabilitation Project, financed by a US$150 million World Bank (IDA) credit against a total cost of about US$250 million, ran from 1994 to 2002 across 1.56 million hectares in 21-22 counties of Shanxi, Shaanxi, Gansu and Inner Mongolia -- among the poorest, most sediment-eroded parts of the Yellow River basin.

Activities

The project terraced 90,446 ha of cropland, afforested roughly 291,000 ha of arbor trees, shrubs, economic trees and orchards, built 149 key dams plus over 3,000 warping and check dams, and backed local grazing bans on goats and sheep. Disbursement was tied to unit prices per hectare verified against 1:10,000 land-use maps; farmers received 30-50 year land contracts, and roughly 60% of costs were recovered from beneficiaries.

Results

Annual sediment discharge to the Yellow River fell by 57 million tonnes; grain output rose from 427,000 to nearly 700,000 tonnes; farmers' per-capita net income climbed from 360 to 1,263 yuan (1993-2001); project-area poverty fell from 59% to 27%. The project's economic rate of return rose from 22% (sediment benefits only) to 29% once carbon-sequestration benefits were added.

Conclusions

A 2022 peer-reviewed meta-analysis of the wider afforestation campaign this project helped popularise (2000-2015, whole 640,000 km2 plateau) found forest cover up about 36% and soil erosion down 13.3%, but regional water yield down 43.3% as new vegetation consumed more water than the grasses/crops it replaced -- a genuine erosion-versus-water trade-off later restoration phases have had to address.

Implementation

Indicative cost
Very high (> €5M) — US$250M total cost, of which US$150M was a World Bank IDA credit; about 60% of costs recovered from beneficiaries.
Time to results
Long (> 3 years) — 8-year implementation (1994-2002); downstream afforestation effects studied through 2015.
Staffing & skills
China Ministry of Water Resources project office, county-level water and land bureaus, World Bank IDA supervision team, participating farmers under 30-50 year land contracts

Conditions for success

  • disbursement tied to verified per-hectare unit prices against 1:10,000 land-use maps
  • long-tenure (30-50 year) land contracts for farmers
  • about 60% beneficiary cost recovery to sustain incentives
  • locally enforced grazing bans to let vegetation regenerate

Common failure modes

  • afforestation at wider regional scale can cut water yield by over 40% via evapotranspiration if species and scale are not matched to local hydrology

Where it fits

Governance type
national government with multilateral (World Bank IDA) financing
Scale
regional watershed (1.56M ha)
Income level
lower-middle income (China, 1990s)

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful