The yield number, and why the crop-year estimate keeps moving

India's cotton yield reached 566 kilograms of lint a hectare in 2013-14, its historical peak on the back of the initial Bt-cotton adoption wave. It has been on a broad downward slide since: secondary estimates put 2024-25 at roughly 436 to under 450 kg/ha, against a global average above 800 (USDA puts it at about 814–854), China at a record of about 2,384 kg/ha on USDA’s estimate (2,171.6 on China’s official figure) and Brazil at 1,800 to 2,000. USDA’s own estimate for India in 2024-25 is a little higher than the official one, at about 461 kg/ha. India has the lowest lint yield of the world's top five cotton producers — a gap that has widened, not narrowed, over the past decade.

India's raw production number for the current season is a moving target, and it is worth being precise about why, so it is not mistaken for the same story as the yield decline above. The Cotton Association of India (CAI) has revised its 2025-26 crop estimate upward through the season — from 317 lakh bales in February 2026, to 320.5 in March, 324 in April, and 337 lakh bales (170 kg each) by July 2026. That is a within-season reconciliation of acreage and weather data against an earlier, more cautious estimate; it says the current crop looks somewhat better than CAI initially feared, not that the multi-year yield problem above has reversed. Both things are true in the same season: this year's absolute output estimate has been revised up, and the per-hectare yield trend since 2013-14 remains down.

Imports are the other half of the "rising imports" the brief for this piece named. CAI has put 2025-26 cotton imports at a record ~60 lakh bales, revised up from 47 lakh, with mills front-loading purchases to use a temporary 11% import-duty exemption before it lapses; roughly 51 lakh bales had already landed by end-June. The Confederation of Indian Textile Industry raised its consumption estimate to 348 lakh bales over the same period, and CAI trimmed its export estimate to 15 lakh bales from 18 lakh the year before. Rising imports, a shrinking export line and a still-below-peak yield are one coherent picture, not three separate anomalies.

What actually broke: pink bollworm resistance, not the seed technology itself

Bt cotton — hybrids of Gossypium hirsutum carrying Bacillus thuringiensis Cry genes — covers the great majority of India's cotton area and drove the yield climb through 2013-14. What has eroded since is specific: the pink bollworm evolved resistance to both the first- and second-generation Bt traits (Cry1Ac and Cry2Ab) it was engineered to kill. Gujarat alone lost an estimated 1.41 million bales in 2015, worth around $817 million, largely to resistant pink bollworm; nationally, since 2017-18, the pest's return is estimated to cost about 13.2 lakh bales, or roughly ₹3,900 crore, a year. A secondary effect compounds it: suppressing the primary bollworm pests opened an ecological niche that sucking pests (whitefly, jassids, mealybugs) have moved into, pushing up the pesticide spend needed to protect the same hectare. One review cites raw cotton export earnings falling from close to $9 billion in 2013 to about $800 million in 2020 as a consequence of the same pressure — a figure this piece could not independently reconcile against DGCI&S trade data and is flagging rather than treating as confirmed (see caveats).

Kaala cotton is a Kutch story — and that matters for what this piece can honestly recommend

Kaala (also Wagad, or Sakala) cotton is Gossypium herbaceum, an Old World diploid species indigenous to the Kutch region of Gujarat, with archaeological evidence of cultivation there tracing back to the Indus Valley Civilisation (roughly 3000–2750 BCE). It is a rain-fed crop needing no synthetic pesticide or fertiliser, with a short, coarse staple (about 22–23 mm) suited to handwoven textiles rather than mill spinning. It declined under British-era policy, which favoured long-staple varieties for English mills, and was branded and revived as "Kala Cotton" from 2010-11.

None of that geography overlaps with Maharashtra, Tamil Nadu or Madhya Pradesh, where the Bt-hybrid yield and soil problems described below are actually concentrated. The Bt hybrids grown across all three states are G. hirsutum, a New World tetraploid species — genetically a different lineage from Kaala's Old World diploid, not simply an "improved" version of it. Swapping one for the other in Vidarbha or the Malwa plateau is not a seed change a farmer can make within an existing hybrid-cotton system; it is a different crop, with a different staple length and a different buyer. This piece did not find a published ICAR field trial introducing G. herbaceum genetic material into any of the three named states, and does not claim one exists.

What is genuinely transferable is the agronomic logic Kaala cotton demonstrates, not the plant itself: a rain-fed, low-synthetic-input model, and an Old World short-staple genepool that plant breeders could in principle draw on for drought or pest-resilience traits. ICAR's Central Institute for Cotton Research (CICR), Nagpur, is the body positioned to evaluate that breeding question for Vidarbha specifically; this piece did not find evidence that work is currently under way there, and says so rather than implying otherwise.

What is happening to the soil under Bt cotton in the three states named

Maharashtra (Vidarbha). About 97% of the state's cotton is rain-fed. Current Vidarbha yields run around 7–8 quintals of seed cotton (kapas) a hectare — roughly 700–800 kg. At a typical ginning-out-turn near 35% lint by weight (this piece's own approximation, not an official split for this season), that converts to something close to 245–280 kg/ha of lint, broadly consistent with a separately reported state series showing Maharashtra averaging 285 kg/ha of lint across 2001–2025, with a low of 100 in 2001 and a high of 475 in 2017. Reporting on the region describes rising input costs and unreliable yields as having made Bt hybrid economics "untenable" on Vidarbha's harsher, rain-shadow terrain.

Tamil Nadu. The state's cotton area has contracted over a long period, with farmers in surveyed districts widely reporting they have cut back acreage from what they once planted. One frequently repeated secondary figure describes area falling from a mid-1990s peak to a lower level by 1999-2000; the exact scale is inconsistent across sources this piece checked — some describe the peak in lakh hectares, others in million — and is flagged here rather than restated as a precise number. What is not in dispute is the direction and the named causes: water scarcity in existing irrigation command areas, imbalanced fertiliser use, micronutrient deficiency, and rising soil salinity and alkalinity, particularly in coastal and heavily irrigated tracts, all documented in Tamil Nadu Agricultural University's own problem-soils literature. Policy commentary on the sector calls for regaining area in irrigated cotton tracts and promoting extra-long-staple varieties in the state's southern districts specifically.

Madhya Pradesh (Malwa and Nimar). The state's black cotton soil (Vertisol) is naturally rich in calcium, magnesium, potassium and iron but poor in nitrogen, humus and phosphorus — a profile textbook-suited to cotton, wheat and soybean. Against that, zinc deficiency is severe and specific: agro-climatic zone surveys put it at 62.6% of soils in the Malwa Plateau and 60.0% in the Nimar Valley, the state's two principal cotton tracts. Potassium matters more than the soil's raw richness in it suggests, because cotton fibre quality is unusually sensitive to potassium supply — a documented deficiency reduces leaf-area expansion and CO₂-assimilation capacity, cutting yield and fibre quality together. Cotton area itself has also been falling here, with one report citing a decline as steep as 18% as farmers shift toward soybean and other pulses; this piece treats that figure as a single-source estimate rather than a confirmed official statistic.

The colonial precedent: British-era cotton and sugarcane cultivation left the same kind of soil damage behind

The continuous-cropping and canal-salinity mechanisms described above for today's Vidarbha and Deccan tracts are not a new invention of the Bt-hybrid era. They have a specific, documented 19th- and early-20th-century precedent in the same Maharashtra geography, under British colonial policy.

Cotton monoculture, forced by price and property rights. Berar (the historical name for the tract that is now Vidarbha) had a favourable combination of black cotton soil and, from the 1860s, a colonial land-tenure structure that granted cultivators occupancy rights conditional on continuous agricultural use and tax payment — a structure that pushed land into cash-crop production rather than subsistence rotation. The American Civil War (1861–65) cut off the American cotton supply to Lancashire's mills, and India's share of Britain's raw cotton imports jumped from roughly 12% before 1861 to about 90% within the first year of the war; a quarter of that was from Berar alone, which by 1867 was reportedly exporting as much raw cotton to Manchester as the whole of Egypt. Because cotton was, in the colonial economy, "the only guaranteed marketable commodity," cultivators were pushed to substitute cereals and pulses — the very legumes this piece's own reporting on India's fertiliser use notes fix nitrogen and break a monoculture's nutrient drain — with ever-larger cotton acreage. That substitution, driven by an external price shock and a colonial tax structure rather than by agronomy, is the same continuous-cotton-cropping pattern this piece flags today as a plausible (if unproven at the district level) contributor to Vidarbha's nutrient strain; it did not start with the hybrid seed.

Canal irrigation, built partly to expand cotton and cane. Following the famines of the 1870s, British engineers built Deccan irrigation works explicitly to reduce drought risk and expand revenue-generating cash-crop area — the Khadakwasla Dam on the Mutha near Pune (completed 1879) and the Wilson Dam on the Pravara at Bhandardara (built 1910–1926), the latter specifically cited in the historical record as built to "promote cotton cultivation in Ahmednagar district" alongside famine protection; sugarcane, needing a stable year-round water supply, became the single largest revenue contributor to the Bombay Irrigation Department and the crop the Deccan canal system was increasingly built around. Applying canal water onto the region's black soil produced waterlogging and rising salinity that colonial administrators themselves recognised as a serious problem by the 1870s. That legacy is still measurable: soil surveys find roughly 1.5% of the total geographical area across the command areas of the Godavari, Ghod, Purna, Manar, Mula, Pravara, Nira and Krishna river basins — the same Deccan basins this canal-building programme targeted — now classified as moderately to strongly saline or sodic, attributed in the surveying literature to over-irrigation and the establishment of sugarcane cultivation and allied industries in those command areas.

Why the soil-fertility angle compounds the pest problem instead of sitting apart from it

This blog's own reporting on India's crop-by-crop fertiliser use found cotton takes about 89.5 kg of nitrogen a hectare nationally — the third-highest of any major crop, after sugarcane and wheat. Nitrogen dosing gets attention because it is what farmers apply most visibly; potassium, which the Madhya Pradesh finding above shows cotton needs for fibre quality specifically, is comparatively easy to under-apply on soils that look potassium-rich on paper. Where continuous or near-continuous cotton cropping runs on soils already short of zinc and vulnerable on potassium — Malwa and Nimar's documented profile, and plausibly Vidarbha's after decades of the same rotation, though this piece does not have a Vidarbha-specific nutrient-mining study to cite — a nutrient deficit and a resistant pest are not two separate problems a farmer faces. They compound: a plant under nutrient stress carries a heavier yield penalty from the same level of pest pressure than a well-fed one would.

The byproduct opportunity: what the plant makes besides lint

The user brief behind this piece asked specifically about cottonseed oil, cottonseed meal and biogas from cotton waste as an income line to offset declining cotton returns. All three already exist as functioning, if incompletely captured, opportunities.

Cottonseed oil. A typical ginning split leaves roughly 60–65% of the kapas by weight as cottonseed rather than lint (an approximate ginning ratio, not a figure specific to this season or state). India's own edible-oil strategy explicitly counts cottonseed as a secondary-source opportunity: government and NITI Aayog analysis puts the potential at an additional 1.4 million tonnes of edible oil nationally, cutting the country's edible-oil import-dependence gap by an estimated 9.7 percentage points — part of the same push covered in this blog's earlier piece on India's textile-sector import-substitution loop.

Cottonseed cake (meal). This is not a speculative market to build from scratch: cottonseed oilcake already trades actively on NCDEX at around ₹2,900 a quintal (roughly ₹29,000 a tonne) as a protein feed for dairy cattle in cotton-growing regions. The realistic upside for Vidarbha, Nimar and Tamil Nadu's cotton belts is capturing more of that existing value at or near the farm gate — village- or cooperative-level crushing rather than selling raw seed out of the district at a lower realisation — not creating new demand.

Cotton stalk and biogas. This is the most underused byproduct of the three. Cotton stalk can represent 80% or more of the plant's total above-ground mass, and today it is mostly burnt in the field or discarded rather than processed. India's estimated surplus cotton-stalk residue nationally is 29.74 million tonnes, out of roughly 234 million tonnes of total surplus agricultural residue — a genuine feedstock for compressed biogas (CBG) under the GOBARdhan scheme, which this blog has covered separately in the context of how few of India's 132 commissioned CBG plants are actually pipeline-connected. Published CBG yields run 50–55 kg of gas per tonne of feedstock, selling at roughly ₹50–75 a kilogram; at the midpoint of each range, that is a rough ₹3,000–3,500 of CBG revenue per tonne of stalk feedstock before capital or operating cost — an illustrative calculation this piece is making from published unit rates, not a government cost-benefit figure. Setting up even a small CBG plant costs ₹300–500 lakh, well beyond an individual farmer or most single cooperatives, which is why this is a proposal for cluster-level or FPO-backed investment rather than a claim that any farmer can convert their own stalk today.

What would fix this

• ICAR-CICR (Nagpur) should publish district-level soil-nutrient data — especially zinc and potassium — for continuous-cotton tracts in Vidarbha, the way Madhya Pradesh's own agro-climatic zone surveys already do for Malwa and Nimar. A state-average dose cannot fix a deficiency this specific to a district's cropping history.
• ICAR-CICR should run and publish a breeding trial evaluating G. herbaceum genetic material for drought and pest resilience, framed explicitly as breeding stock rather than a "grow Kaala cotton here" recommendation the geography does not support.
• State cooperative dairy federations and FPOs should expand village-level cottonseed-crushing capacity in cotton belts, so cake and oil value is captured near the farm rather than the raw seed leaving the district at a lower price.
• MNRE/GOBARdhan should extend viability-gap funding specifically to cotton-stalk-fed CBG clusters in Vidarbha, Nimar and Tamil Nadu's cotton tracts, given a 29.74-million-tonne national stalk surplus that is mostly burnt rather than converted today.
• CAI and the Ministry of Textiles should publish a single, reconciled lint-yield series by state, in kg/ha, on a fixed schedule. This piece had to convert a seed-cotton figure against a separately reported lint-yield series just to compare Vidarbha to the national number; a reader should not have to do that.

These are proposals from this blog, not recommendations any agency has adopted.

Sources and caveats

About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.