agriculture · water · family: success's shadow
the harvest that hollowed the land
Green revolution's yield success degrades the soil and water systems it depends on
Problem statement
India's Green Revolution more than tripled wheat production — from 10.3 million tonnes (1960) to 36.3 million tonnes (1981), reaching 55.1 million tonnes by 1991 (USDA data) — eliminating chronic food deficit and achieving food self-sufficiency. But the high-yield variety (HYV) monoculture that produced this success has progressively degraded its own resource base. More than 90% of Punjab's intensively cultivated soils are now low to medium in soil organic carbon (Sharma et al. 2016). India's fertilizer response ratio collapsed from 13.4 kg grain per kg NPK in 1970 to 3.7 kg/kg in 2005 (Biswas & Sharma 2008, as cited in the nutrient-management literature). Punjab's water table fell at roughly 0.49 m per year (CGWB assessment through 2017), 109 of the state's 138 blocks (79%) are officially "over-exploited," and the groundwater farmers are chasing downward is contaminated — uranium exceeded the 30 ppb permissible limit in 62.5% of Punjab's post-monsoon samples in CGWB's 2025 annual report, the worst rate in India. Yields have stagnated in approximately 72% of rice and 85% of wheat long-term rice-wheat experiments across South Asia and China, in the plots given recommended fertilizer rates (Ladha et al. 2003). The system that created India's food security is undermining the foundations on which that food security rests.
Why this matters
Punjab accounts for approximately 15% of India's wheat production and 10% of its rice on just 1.5% of the country's geographical area (Sonker & Kallummal 2025). Approximately 100,000 (1 lakh) indigenous rice varieties were lost to HYV adoption (John & Babu 2021). Micronutrient deficiencies are measurable: 11% of sampled Punjab soils are low in zinc, 15% in manganese, and 12% in iron (Sharma et al. 2016). The crisis is not future-tense — yield stagnation is measurable in existing long-term field trial data, and groundwater depletion has moved from declining aquifers to contaminated aquifers.
What’s been tried and why it hasn’t worked
Fertilizer is one of the subsidized inputs, and Punjab's fertilizer use is heavily skewed toward nitrogen: an N:P:K ratio of 23:5:1 in the latest year reported, against a recommended 4:2:1 (Sonker & Kallummal 2025). Minimum support prices (MSP) for wheat and rice lock Punjab into the monoculture rotation that degrades soil — diversification would mean abandoning guaranteed procurement prices. Crop residue management (related to existing brief agriculture-india-crop-residue-valorization-gap) addresses one symptom but not the systemic monoculture driver. The fundamental barrier is policy lock-in: the MSP/procurement system, fertilizer subsidies, and free electricity for pumping all reinforce the wheat-rice monoculture.
What would unlock progress
Reforming the MSP system to include crops beyond wheat and rice — assured procurement of the alternative crops — is the condition that a farm leader and an agriculture expert quoted by Mongabay-India name for phasing out paddy (Gupta 2022). Outcome-based incentives (soil health payments, water-use efficiency payments) rather than input subsidies. Precision nutrient management calibrated to actual soil deficiency profiles rather than blanket NPK application. Crop diversification into pulses, oilseeds, and millets: pulses fix nitrogen in the soil (Sonker & Kallummal 2025), and millets need little water (John & Babu 2021).
Entry points for student teams
The plot-level records of the All India Coordinated Research Project's long-term fertilizer experiments are held by ICAR and the state agricultural universities and are released only on institutional request, so a team should work from what those experiments have already published: the per-treatment yield and nutrient-balance summary tables in Ladha et al. (2003), the largest such synthesis for the rice-wheat belt, carry enough of the series to fit nutrient-depletion trajectories and test alternative cropping scenarios against them. FAO's FAOSTAT (https://www.fao.org/faostat/) supplies the open India-level fertilizer-by-nutrient and cereal-production series a team needs to check whether a fitted trajectory reproduces the national response-ratio decline. A second door needs no data access at all: an economic model comparing farmer returns under current MSP procurement against a diversified rotation paid through soil-health and water-use incentives, built from the published CACP cost-of-cultivation and MSP schedules. Agricultural economics, soil science, and policy analysis skills apply.
Genes
Structural cousins: stuck the same way, in other fields
Sources
John, Daisy A. & Babu, Giridhara R. (2021), "Lessons From the Aftermaths of Green Revolution on Food System and Health," Frontiers in Sustainable Food Systems 5: 644559, doi:10.3389/fsufs.2021.644559; Ladha, J. K., et al. (2003), "How extensive are yield declines in long-term rice-wheat experiments in Asia?", Field Crops Research 81(2–3): 159–180, doi:10.1016/S0378-4290(02)00219-8; Sharma, B. D., Rajkumar, Manchanda, J. S., Dhaliwal, S. S., Thind, H. S. & Yadvinder-Singh (2016), "Mapping of Chemical Characteristics and Fertility Status of Intensively Cultivated Soils of Punjab, India," Communications in Soil Science and Plant Analysis 47(15): 1813–1827, doi:10.1080/00103624.2016.1208756; Central Ground Water Board data as reported in Gupta, Vivek (2022-06-14), "Accelerating rate of groundwater depletion in Punjab, worries farmers and experts," Mongabay-India, and The Tribune (2025-12-12), "Uranium exceeding safe limit in 62.5% of Punjab groundwater samples: Satnam Sandhu in Rajya Sabha," Sonker, Akanksha Pratik & Kallummal, Murali (2025), "Trade and Sustainability Challenges in India: Possibilities for Paddy Crop in Punjab," CRIT/CWS Working Paper Series No. 87, Centre for WTO Studies, USDA India wheat production series via IndexMundi, All India Coordinated Research Project on Long-Term Fertilizer Experiments (published summary tables; plot-level dataset held by ICAR/state agricultural universities, not open). Accessed 2026-08-21 go to source 1 ↗ go to source 2 ↗ go to source 3 ↗ go to source 4 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
This is a "problems of success" case: the Green Revolution's yield gains were genuine, but the system that produced them is self-undermining. Distinct from existing brief agriculture-india-crop-residue-valorization-gap (which covers the residue burning symptom, not the underlying soil/water/policy system). The "infrastructure lag after success" sub-type applies: fertilizer subsidies, MSP, and power subsidies for pumping took shape as part of the Green Revolution package and still give rice and wheat a competitive edge over other crops in Punjab (Sonker & Kallummal 2025). Related to agriculture-realtime-soil-organic-matter-sensing (which addresses the sensing gap for SOM measurement, not the policy drivers of SOM decline).
Reconciliation 2026-08-21: Drift confirmed — the anchor paper is real but was carrying numbers it does not contain, and several of those numbers were wrong. John & Babu 2021 (Frontiers in Sustainable Food Systems 5: 644559, fetched in full) confirms the ~100,000 ("almost 1 lakh") lost indigenous rice varieties but contains none of the brief's other figures. Wheat production: "11.4 MT (1960) to 36.3 MT (1990)" mis-dated the endpoint by a decade — USDA series (via IndexMundi) shows 10.3 MT in 1960, 36.3 MT in 1981, and 55.1 MT by 1991; corrected. The yield-stagnation figure verifies cleanly where triage might have suspected it: rice ~72% and wheat ~85% of long-term rice-wheat experiments show stagnant yields, from Ladha et al. 2003, Field Crops Research 81(2–3): 159–180 — now cited explicitly. The fertilizer-efficiency figure "80.9 kg grain per kg NPK (1966) to 16.0 (2004)" could not be found in any source; replaced with the widely cited 13.4 → 3.7 kg grain/kg NPK (1970→2005) decline attributed to Biswas & Sharma (2008) — the original is paywalled, so the attribution is flagged in-text as literature-cited rather than directly fetched. Soil organic carbon "0.4–0.7% from historical 2–3%" was unsupportable (no evidence Punjab's alluvial soils ever held 2–3% SOC); replaced with the verified characterization from Sharma et al. 2016 (Communications in Soil Science and Plant Analysis 47(15): 1813–1827): >90% of intensively cultivated Punjab soils low-to-medium in OC. The same paper corrects the micronutrient claim — zinc-low soils are 11% of samples (not 22%), Mn 15%, Fe 12%; the unsourced sulfur clause was dropped. Water table "75 cm–1 m per year" overstated the assessed statewide rate: CGWB's assessment through 2017 works out to ~0.49 m/yr, with 109 of 138 blocks (79%) over-exploited (Gupta, Mongabay-India, 2022-06-14, fetched). Contamination re-anchored: uranium above the 30 ppb limit in 62.5% of post-monsoon samples per CGWB's Annual Ground Water Report 2025, the highest rate in India (The Tribune, 2025-12-12, fetched); "lead and arsenic" dropped from the headline sentence — arsenic is documented in Punjab districts but the lead attribution could not be verified to a primary source. Production shares "~20% wheat / ~12% rice" corrected to ~15% of national wheat production and 10% of rice on 1.5% of geographical area, copy-quoted from Sonker & Kallummal 2025 (CWS Working Paper No. 87, PDF fetched: "Punjab, today, accounts for approximately 15% of the country's total wheat production"; rice "Punjab (10%)"; "only 1.5% of the country's geographical area"). Soil Health Cards launched 2015, MSP/procurement lock-in, subsidized-urea imbalance, and free-electricity claims verify as standard policy facts and stand unchanged. All Source-line URLs fetched live 2026-08-21.
Reconciliation 2026-08-21: Entry-point realism (C37 triage, score 1) — flag CONFIRMED. The first door asked teams to "analyze the 50-year Long-Term Fertilizer Experiment dataset"; the AICRP-LTFE plot-level records are held by ICAR and the state agricultural universities and are released on institutional request, not as an open download, and the ICAR-IISS host for the AICRP-LTFE could not be reached at all from this session (DNS failure on `iiss.icar.gov.in`), so no AICRP annual-report URL is cited here — the door was rebuilt on what those experiments have already published, chiefly the per-treatment summary tables in Ladha et al. 2003 (Field Crops Research 81(2-3): 159-180), which the brief already cites and which carry enough series to fit depletion trajectories. Added as the open cross-check: FAOSTAT's India fertilizer-by-nutrient and cereal-production series (https://www.fao.org/faostat/, fetched). Unflagged-door check: the second door (MSP-versus-diversification economic model) was checked against the ≥2-doors rule and passes as the facility-free door — it needs only the published CACP cost-of-cultivation and MSP schedules — so it was kept and labelled as needing no data access rather than rewritten. Declined to cite, for lack of a verifiable fetch from this session: the AICRP/ICAR-IISS annual reports, India-WRIS (`indiawris.gov.in`, connection refused), CGWB (`cgwb.gov.in`, TLS failure), DES/desagri.gov.in (certificate failure), and the ICRISAT district-level database (`data.icrisat.org`, connection refused); the CGWB and Tribune anchors already in the Source line were verified in the earlier citation round and are untouched. The Source line's AICRP entry was corrected from "(long-term dataset)" to name the published summary tables and flag the plot-level data as not open.
Reconciliation 2026-10-01: Fact repair after reading the whole write-up against the sources it names, all fetched 2026-10-01; a second reader went over the repairs on the same day, re-fetched every source, and reworded two of them and restored one word. Read in full: John & Babu 2021 (https://www.frontiersin.org/journals/sustainable-food-systems/articles/10.3389/fsufs.2021.644559/full); the Mongabay-India article; the Tribune article; the Sonker & Kallummal working paper (PDF, 44 pages, full text searched and the soil, water, diversification and conclusion sections read); and the IndexMundi wheat table. Read as abstract and citation record only, because the publishers block access and no open copy or archive snapshot was found: Ladha et al. 2003 (abstract from Europe PMC, record from https://api.crossref.org/works/10.1016/S0378-4290(02)00219-8) and Sharma et al. 2016 (abstract from OpenAlex, record from https://api.crossref.org/works/10.1080/00103624.2016.1208756). Not opened: Biswas & Sharma 2008 (no link on the page, no open copy found) and the AICRP long-term fertilizer experiment tables. (1) The text said the water table fell 0.49 m per year "statewide". The Mongabay-India article gives 0.49 metre per year as the average rate in a Central Ground Water Board report, without the word statewide; the word was removed. The word "officially" before "over-exploited" stays: the 109 of 138 blocks (79%) figure is the Board's 2017 categorisation, reproduced in Table 3 of the working paper (GWRA 2017: 109 over-exploited of 138 assessment units, 79.0%). (2) The uranium figure of 62.5% is for samples taken after the monsoon (53.04% before it); the sentence now says post-monsoon. (3) The yield-stagnation sentence read as if it were about Punjab trials. The Ladha abstract covers 33 long-term rice-wheat experiments in South Asia and China, and the 72% and 85% figures are for treatments given recommended rates of N, P and K; the sentence now says so. (4) The text said subsidized fertilizer policy "perversely incentivizes over-application of cheap urea relative to P and K." No source named on the page mentions urea or makes that causal claim. The working paper lists fertilizer among the subsidized inputs and reports Punjab's N:P:K ratio as 23:5:1 against a recommended 4:2:1; the sentence now gives that. (5) The sentence on soil health cards (launched 2015) was removed; none of the sources mentions them. (6) The text said reforming MSP "would break the monoculture lock-in." The sources do not promise that; the Mongabay-India article quotes one farm leader and one agriculture expert saying assured procurement of other crops is the condition for phasing out paddy, and the sentence now says that, in the singular. (7) The text said pulses, oilseeds and millets "rebuild soil organic matter and break pest cycles." Neither claim is in the sources. The working paper says pulses fix nitrogen in the soil, and John & Babu say millets have low water requirements; the sentence now says that. (8) The working note that subsidies, MSP and water pricing were "never updated" is not in the sources (the working paper records that free electricity for agriculture was announced in 1997 and withdrawn in 2002, and lists several diversification drives); it now says fertilizer subsidies, MSP and power subsidies for pumping took shape as part of the Green Revolution package and still give rice and wheat a competitive edge over other crops, which is the working paper's wording ("input subsidies on fertiliser, credit and power"; "helped rice and wheat gain a competitive edge over the other crops"). Confirmed in the sources and left as written: wheat production of 10.3, 36.3 and 55.1 million tonnes in 1960, 1981 and 1991; the move from importing grain to self-sufficiency; more than 90% of soils low to medium in organic carbon; 11% of samples low in zinc, 15% in manganese, 12% in iron; 0.49 m per year and 109 of 138 blocks (79%); the 30 ppb limit and Punjab as the worst-affected state; about 15% of wheat, 10% of rice, 1.5% of area; almost 1 lakh indigenous rice varieties lost; MSP, input subsidies and free electricity as the forces holding the rice-wheat pattern in place; and the Source-line citations for John & Babu, Ladha et al., Gupta, The Tribune and the working paper. Could not be checked: the 13.4 to 3.7 kg grain per kg NPK figure (Biswas & Sharma 2008 not opened); the issue number for Sharma et al. (Semantic Scholar gives volume 47, pages 1813-1827 and the author "Rajkumar"; the Crossref record gives no volume or pages and lists the second author as Raj Kumar); and whether Ladha et al. is the largest such synthesis and prints per-treatment tables (only its abstract was read). One thing a reader should know: the working paper cited here for Punjab's production shares concludes that the situation is serious but not an imminent threat to the rice-wheat pattern, that over-exploited blocks had fallen to 75% of assessment units by 2024, and that diversifying away from paddy is not feasible under current policy. No title changed. Genome tags untouched.