28 March 2003

YIELDS OF DISCONTENT - 'SCIENCE' ARTICLE "FULL OF HOLES"

'Abhijit Sen, professor of economics at Jawaharlal Nehru University, Delhi, and chairman of the High-Level Committee on India's Long-Term Grain Policy, says the data has such high standard errors that Qaim and Zilberman cannot really claim that Bt yield differentials are any higher here than elsewhere in the world.  "They have simply not been able to establish the main case that yields are 80% higher."

If the thesis is so full of holes, what was the rush to go to press? A possible clue might lie in an American news agency report published last month. It said  that Monsanto is trying to shake off a year-long profit slide (the firm lost $1.75 billion or $6.67 per share, compared to a profit of $399 million or $1.51 per share a year ago, forcing CEO Hendrik Verfaillie to resign) that was in part due to deeper concerns worldwide over biotechnology and a drought at home. Thus, the company needs to sell aggressively, say analysts. There may be more than a seed of truth in this.'

thanks to AgBioIndia for these:
*Yields of discontent
*Bt Cotton's new controversy

***

Yields of discontent

By Latha Jishnu
Business World

Articles in reputed scientific journals like Science or Nature usually do not stir up much heat and dust in India. But the 7 February issue of these two magazines has created quite a storm in environmental and agricultural circles in the country.

What provoked the outrage? It began with Science publishing a paper by Matin Qaim of the University of Bonn's Centre for Development Research and David Zilberman, professor at the University of California in Berkeley, extolling the yield potential of American seed giant Monsanto's biotech Bt cottonseed. The basis on which these claims were made - using the data provided Mahyco-Monsanto - prompted some to term the findings a scientific fairytale. And not all those who are critical are aligned with the green brigade that opposes biotechnology.

Shanthu Shantharam, for instance, who finds their analysis outlandish, is a scientist who has worked as a regulator with the US Department of Agriculture and is an authority on problems such as 'pest-resistant genes in managed ecosystems'. Recently with Syngenta and the Washington-based International Food Policy Research Institute, Shantharam has no doubt that biotechnology in any crop when deployed as part of an integrated pest management programme, will outperform ordinary crops. But to suggest that Bt cotton has out-yielded non-Bt cotton by more than 80% and linking it directly to a single Bt gene is, he feels, a preposterous idea. "This kind of shoddy publication based on meagre and questionable data in reputed journals do more harm to science and technology development", and will perhaps set GM technology back, he warns.

Devinder Sharma, who heads the Delhi-based Forum for Biotechnology and Food Security, is not surprised by this strong reaction. To start with, the researchers have neatly interchanged the terms 'yield increases' and 'reducing crop losses' to justify the wrong usage of the term yield increase. "Quantum jumps in production were achieved earlier by pesticides too, but does that mean pesticides increase yields?" he queries.

The crux of the issue is that authors have relied on field trial data supplied by Mahyco-Monsanto - data that is still not in the public domain. Most field agronomists and plant breeders know that carefully controlled field tests and the performance in a normal field provide widely divergent results. A retired scientist with the Indian Agriculture Research Institute says the results of the study would have been more credible if the researchers had waited for a couple of years to collect statistically meaningful data. Abhijit Sen, professor of economics at Jawaharlal Nehru University, Delhi, and chairman of the High-Level Committee on India's Long-Term Grain Policy, says the data has such high standard errors that Qaim and Zilberman cannot really claim that Bt yield differentials are any higher here than elsewhere in the world.  "They have simply not been able to establish the main case that yields are 80% higher."

If the thesis is so full of holes, what was the rush to go to press? A possible clue might lie in an American news agency report published last month. It said  that Monsanto is trying to shake off a year-long profit slide (the firm lost $1.75 billion or $6.67 per share, compared to a profit of $399 million or $1.51 per share a year ago, forcing CEO Hendrik Verfaille to resign) that was in part due to deeper concerns worldwide over biotechnology and a drought at home. Thus, the company needs to sell aggressively, say analysts. There may be more than a seed of truth in this.

***

Bt Cotton's new controversy

A study claims  the yield from Bt cotton  is  80% higher. AP says it is below normal
Latha Jishnu & G. S. Radhakrishna

WITH the harvest of India's first biotech crop, the gene-spliced Bt cotton hybrid from US seed giant Monsanto, coming in from the six states where commercial production was initiated last year, the yield is proving highly contentious. A global debate has erupted over a controversial study that claims Bt cotton yields are up a hefty 80-87% while the Andhra Pradesh government says the crop has been a disappointment.
 
The findings of agriculture economist Matin Qaim of the University of Bonn's Centre for Development Research and David Zilberman, professor at the University of California in Berkeley, has triggered a searing debate worldwide on a question that is central to developing economies. Can genetically modified crops like Bt Cotton raise yields as significantly as these two researchers claim it does and can their findings be extrapolated to genetically modified (GM) food crops in general?

Although the researchers conceded that the limited experience with GM crops is insufficient to make broad generalisations, "we use the example of Bt cotton in India to suggest that existing GM crops can have significant yield effects in the developing world, especially in the tropics and subtropics," they say.  Their analysis is stunning because Monsanto's Bollgard Bt cotton, derived from the soil-borne bacteria Bacillus thuringiensis, is best known for its ability to kill the American bollworm, which is a major destroyer of the cotton crop but not for raising yields to a noticeable degree. In the Bt cotton growing areas of the world, notably the US and China, the yield has been up a mere 5-10% in the four years since it was introduced.

This is of critical interest to India, which has the world's largest acreage of cotton (25% at nine million hectares) but accounts for just a little over 12% of the production. Our yields are notoriously low: just 300 kg per hectare against over 1,000 kg in China and a world average of 580 kg. Given that the crop is prone to heavy pest attack, Bt cotton should be the answer to the Indian farmer's cotton cultivation. But is it?

The Qaim-Zilberman study published in the 7 February issue of Science has had activists, the scientific community, economists and farm policy analysts up in arms. This is primarily because of the questionable basis of their study (see 'Yields of Discontent'). The two academics used data collected from field trials undertaken by Monsanto's Indian partner, the Maharashtra Hybrid Seed Company (Mahyco), in 2001 and not from commercial planting, which as the authors themselves concede, normally show lower yields. And the yields have been disappointingly low in Andhra Pradesh, according to the state government. In a statement to the legislative assembly, agriculture minister V. Shobanadeeswar Rao said farmers have complained of smaller boll size and lesser staple length for Bt cotton compared to other hybrids. This had resulted in lower market value. Rao told Businessworld: "AP farmers had been pinning their hopes on Bt cottonseed but they are disappointed because not only is yield low, the quality of the cotton is also not up to the mark. The market price is Rs 300-400 less per quintal compared with other hybrid varieties." Agriculture commissioner S. Bhattacharya confirms the poor showing. According to the department's survey, the average Bt cotton yield was 2-8 quintal per acre against 10-15 quintals for other varieties. The numbers of farmers who tried Bt cotton though, are small: it was grown on 9,000-odd acres, mainly in Warangal, Mahboobnagar and Kurnool districts, and accounted for less than one per cent of the state's one-million hectares of cotton crops.

The minister's statement has sparked off a spate of calls from non-governmental and farmers' organisations for compensation, although Rao emphasises that the government has so far not received any complaints of the cotton crop having failed. But clearly, there has been a communication gap on what Bt cotton is intended for and what farmers might have been led to expect. "It is just a solution for fighting the American bollworm pest," he points out.

True, says Fauzdar Singh, general manager (R&D) with Mahyco Monsanto Biotech (India) (MMB), the joint venture between Mahyco and Monsanto. "Bollgard does not make any claims of additional yield but only protection against three varieties of bollworms - American, pink and spotted bollworms." In any case, the seed plays a minimal role in the yield. In the weightage given by the company for successful farming, "seed would account for 30% while crop management would get around 50%. The rest would be taken up by integration of both" in the harvest.

The frustration of the farmers can be partly explained by the fact they shelled out nearly three times the cost of other hybrids on Bt cotton in a year that agriculture officials say has been free of pest load. The major selling point of this seed is a 70% reduction in the use of pesticide resulting in average savings of Rs 750 per acre, although in AP, where pesticide use is high, it could reduce costs by over Rs 1,000 per acre. This year, however, with hardly any incidence of bollworm, the cost advantage of Bt cotton has almost been neutralised.

Singh says another reason for the disappointment is that MMB had promoted mainly Bollgard's Mech 162, a medium staple length cotton. "We sold 10,000 packets containing 450 grams of Bt cotton seed along with 150 grams of hybrid variety for Rs 1,650," says Singh. This could have led to wrong expectations of a longer staple crop. And how did the company address this issue? Each packet carried an audio cassette, two brochures and an open letter. The cassette explains what 'total cotton management' means. Monsanto India officials, too, feel the present campaign against the poor yield of Bt cotton is by vested interests, including commercial rivals and a range of activists. Indeed, the environmental lobby has been in overdrive with Suman Sahai, president of Gene Campaign, asking union agriculture minister Ajit Singh to make sure that Mahyco-Monsanto paid adequate compensation for crop losses.

However, neither MMB, nor any seed firm is liable for damages for poor yield, feels Singh. According to the memorandum of understanding (MoU) that seed companies (around 40 in AP) sign with state governments, the farmer is only protected against failure of germination and lack of  genetic purity. In the case of the former, companies make good the cost of seeds and part cost of germination, while for failure of genetic purity they pay the cost of yield loss. What do illiterate farmers know of genetic purity and failure of germination, demands agriculture scientist M.S. Chari, advisor to the Centre for World Solidarity, a lobby group on sustainable farming.

Even if the AP experience is an aberration, does GM cotton, specially the Bt gene offer hope for small farmers in poor countries? Decidedly yes, says Ranjana Smetacek, director (public affairs) with Monsanto India. According to her, the nearly 55,000 cotton farmers who harvested the insect-protected Bollgard have reported a 30% increase in yields that has seen an average rise in income of over Rs 7,000 per acre. Based on data compiled by MMB from a 'large sampling of farmers', who have completed their final pickings, the firm says Bt cotton has led to a 65-70% reduction in the use of pesticides. The data for Maharashtra, Madhya Pradesh, Gujarat and  AP (the figures forTamil Nadu are not in) shows an average increase of 3.24 quintals per acre, the highest being in Madhya Pradesh, which recorded an increase of 4.33 quintals per acre and a jump in farm incomes of Rs 9,600 per acre.

Many remain sceptical of such claims. A. Padmaraju, director of Research  at the Acharya N.G. Ranga Agricultural University, maintains that Bt cotton is not a high-yielding  variety. "Apart from its characteristic of  being resistant to the American bollworm, it has no special features," says Padmaraju, whose report on the performance  of Bt cotton was the basis for the debate in the AP assembly  last week. Although there are several other hybrids which are better yielding than the Monsanto brand, "we are yet to arrive at a conclusion because the report we gave the government is based on just 20% samples", he clarifies. A correct assessment of yield cannot be made unless it's tagged with pest resistance, and this year there was no pest load at all, he clarifies.

But in October last year there were reports of bollworm attack near Warora, Maharashtra,  in the three-acre farm in social activist Baba Amte's Anandwan complex. There have been other reports of bollworm attack from Karnataka and Madhya Pradesh.

It's believed that there is a strong chance of the American bollworm developing resistance to the Bt gene. In Australia, China and the US, farmers have been told to increase the number of sprays over the years. Will the worm turn in India? No one is certain, but, for now, the question of yield remains the core of the problem.