Indias cotton farms need better soil, not just new GM seeds
Written by: Keshav Kranthi
5 min readJul 2, 2026 06:20 AM IST
First published on: Jul 2, 2026 at 06:20 AM IST
The article by Ashok Gulati, Ayushi Gupta, and Ritika Juneja (‘In cotton fields, policy is undoing what science built’, IE, June 8) reads more like an advocacy piece for the seed industry than a balanced analysis of what ails Indian cotton. The authors credit Bt cotton, especially Bollgard-II, for the 88 per cent productivity increase between 2002 and 2013. They then argue that seed price control decisions killed the introduction of new GM events, including BG-II RRFlex and Bollgard-III, causing a decline in production and productivity since 2014-15. Had India continued its earlier trajectory, they claim, production would have reached 65.3 million bales by 2026, implying a productivity of 967 kg/ha. This diagnosis is flawed. The authors appear to believe that GM seed technology alone drove India’s productivity gains, and that approving BG-II RRFlex, Bollgard-III or the seven-gene Thryvon stack would restore that trajectory. Neither premise is supported by evidence. BG-II and BG-III protect crops against bollworm damage; RRFlex enables better weed management. These traits are agronomically useful, but there is no credible evidence from anywhere in the world that any of these traits directly increased yields. Framing GM trait approvals as the primary lever for productivity recovery deflects attention from the agronomic, soil science, breeding and management interventions that are decisive for sustainable yield improvement.The authors compare India’s yields of 441 kg/ha with China’s 2,311 kg/ha and credit technology adoption for the difference. But China grows neither BG-III nor HT cotton nor Thryvon. It deploys exactly the same BG-II, Bt genes as India. The critical difference is that China grows Bt varieties rather than Bt hybrids. The authors also omit a telling counterexample. Turkey has not approved GM cotton, yet its yields rose from 1,100 kg/ha in 1999 to 1,728 kg/ha in 2026. Over the same period, India’s yields rose from 304 kg/ha to 458 kg/ha — they peaked at 566 kg/ha in 2013-14 and then declined. Turkey, without any approved GM variety, now achieves yields nearly four times that of India, whose cotton area is almost entirely planted with GM hybrids. China and Turkey offer compelling evidence that sustained productivity improvement depends on varietal development, agronomic investment and soil health, not trait approvals. Data from the Ministry of Textiles also complicates the authors’ narrative. India’s productivity had already reached 521 kg/ha in 2006 and 554 kg/ha in 2007, when BG-II was approved, but occupied only a negligible share of cotton area, as seed production and distribution had yet to scale. Crediting BG-II with the gains is not defensible. Productivity then declined to 449 kg/ha by 2018, even as BG-II adoption peaked. The productivity gains of 2002-2011 are better explained by a confluence of factors: Effective bollworm control by Bt cotton, hybrid seed adoption surging from 38 per cent to 92 per cent, fertiliser use increasing 2.3-fold, and irrigated area expanding from 2.5 million to 4.4 million hectares. Heavier farmer investment in crop management compounded these gains. Once inputs approached saturation, it was the law of diminishing returns, not seed price policy, that drove decline. The root cause of India’s cotton productivity collapse goes unmentioned in the article. According to FAO, nearly 32 per cent of Indian land is degraded and 25 per cent faces desertification. Soil organic carbon in Indian croplands averages just 0.3-0.6 per cent, far below the globally accepted minimum of 1-1.5 per cent for productive soils, with critically low levels below 0.25 per cent recorded across cotton-growing regions of Rajasthan, Gujarat, Madhya Pradesh and Maharashtra. No technology stack, GM or oth
Discussion (0)