http://www.agbioindia.org/newsletter.asp

14 February 2003

Sub: A scientific fairytale on Bt cotton

For the biotechnology industry, painting a successful picture of the introduction of the first genetically modified crop in India - Bt cotton - has tremendous significance. After all, unlike China, India is a democracy and public acceptance of a faulty technology will surely add to the profile of an industry faced with a credibility crisis.  This monumental task can only be accomplished with the help of scientists and economists.

News reports appearing on the basis of a paper that the Science journal published in its edition of Feb 7, 2003, claiming an increase of 70 to 80 per cent in the yields of cotton in India, appeared all over the world. Strange that the international media, which does not find anything newsworthy when the Indian Prime Minister visits the United States, gloats over a study that talks of cotton of all the things. Well, you guessed it right. The issue here is not what happens to cotton farmers in India but the media's renewed interest in promoting the commercial interest of the biotechnology industry.

"Yield Effects of Genetically Modified Crops in Developing Countries" (Science, Feb 7, 2003, Vol. 299) a paper jointly written by Matin Qaim of the University of California, Berkeley, USA and David Zilberman of the University of Bonn, Germany, is the latest in the series. We will perhaps have many more such papers being doled out by dozen, all thanking the seed company Mahyco (clearly avoiding to mention its partner - Monsanto) for its research support.  Nevertheless, what is interesting is that the paper begins with the wrong premise that Bt cotton increases crop yields.
---------------------------------------------

A Scientific Fairytale

Providing a Cover-Up to the Bt Cotton Fiasco in India

By Devinder Sharma

In the mid 1980s, a World Bank team was travelling through parts of the frontline agricultural State of Haryana in northwest India, assessing the impact of its 'dream' project - Training and Visit (T&V) System - of farm extension. That was the time when 'T&V' was the buzzword and the World Bank had doled out millions of dollars to promote the new farm technology dissemination system to take the latest technology from the agricultural laboratories to the land.

My newspaper, the Indian Express, amongst the largest selling dailies in India, deputed me to accompany the team. As the Agriculture Correspondent of the newspaper, I was obviously very keen to follow the outcome of the famed programme. The team travelled through some of the dry and semi-arid regions of the State, and was accompanied by the project director of the 'T&V' System along with his supporting staff.

At most places, farmers were collected to enable the World Bank team to interact with them. The Bank's team would ask the same question - whether the programme had benefited them  - to farmers wherever they went. And I remember vividly that at most places the farmers would say that the programme hasn't made any difference to their lives. But what they said was in Hindi, and the project staff translating it for the benefit of the Bank's team would invariably turn it around saying: "Sir, he says that the programme has changed his life for the better."

No wonder, the World Bank gave a favourable report. It is however another matter that the 'T&V' System now is all but forgotten.

The Genetic Engineering Approval Committee (GEAC), the apex body responsible for granting commercial approval to genetically modified crops, too had conducted a similar survey to assess the impact of the transgenic crop in its very first year of planting. A team comprising four experts, who were either part of the government's approval process or represented the State of Andhra Pradesh had toured Nalgonda, Karimnagar and Warangal districts. The team members, all wearing 'bollgard' caps and accompanied by company
officials from Mahyco-Monsanto, actually saw the standing crop in ten acres (out of the 9,300 acres sown with Bt cotton) and submitted their 'favourable ' report, as expected.

Another team comprising representatives from three NGOs -- Centre for Resource Education, Sarvadaya Youth Organisation and Greenpeace India, trailed the expert team They interviewed the same farmers who were earlier visited by the expert team, and their testimony before the video camera exposes the rot in scientific assessment and analysis. No wonder, the Minister for Environment and Forests, Mr T.R.Baalu, whose son has defaulted the nationalised banks to the tune of Rs 240 million, was quick to make a
statement in Parliament on December 16, 2002, stating that 'studies conducted by an expert team from his ministry has shown "satisfactory performance" of Bt Cotton in the first year of its planting.'

Times haven't changed, isn't it?

For someone who still believes in 'good science' despite global efforts being made to replace it with the industrial prescription of 'sound science' , Matin Qaim and David Zilberman's paper in the American journal Science actually ends up doing a great disservice to good science. The title of the paper -- Yield Effects of Genetically Modified Crops in Developing Countries -- itself is deceptive. The authors have merely tried to look into the yield performance of Bt cotton, which in no way is representative of the
entire range of genetically altered crops. But we can't blame the authors for a faulty and misrepresentative title. This is the prerogative of the editors.

It also makes a tactical error in treating savings in crop losses as yield increase. Interestingly, the Indian Council of Agricultural Research (ICAR) - the umbrella organisation governing the world's second biggest agricultural research infrastructure in India - had earlier objected to Mahyco-Monsanto's repeated claims that Bt cotton increases yields. Bt cotton for all practical purposes acts like a pesticide and pesticides do not increase yields. They merely reduce crop losses. But then, for an industry under tremendous pressure for public acceptance of its risky technology, playing the yield card was a simple way to hoodwink the masses. In fact, the reality is that none of the genetically modified crops have broken the yield barrier that was established by the high-yielding varieties, which ushered in the famed green revolution.

It is true that the potential of GM crops in developing countries is limited without a substantial yield effect, especially in regions with strong population growth. After all, despite the hype that economists like Per Pinstrup-Andersen have created about the Bt cotton success, the fact remains that there is a negligible increase in yield differences vis a vis the non-Bt cotton hybrids in China. Showing a quantum jump in 'yield' (not in reducing the crop losses) was therefore what Mahyco-Monsanto wanted to establish for India. And it is here that the authors fell in a well laid out trap. The data that Mahyco-Monsanto supplied for this paper (and which has been acknowledged by the authors) is based on field trials carried out by the company on 395 farms in seven states. The authors say "in addition to regular trial records, more comprehensive information was collected for 157 farms on agronomic aspects and farm and household characteristics." They
conclude that a cross check of summary statistics showed that these 157 sites are fairly representative of the total 395 trial locations.

The media gloated over the research findings. The international media, which was worried at recent news reports of Bt cotton failure coming from various parts of the country, actually mistook this 'research' to be for the year 2002. News reports give an impression as if the study is for the crop season that has just ended. In reality, the analysis is based on the data that Mahyco-Monsanto had collected in the final year of field-testing in the year 2001, a year before the crop was commercialised. This was the data that the
company had presented before the GEAC. This was the data that still remains hidden from the public gaze in India. And this is the data, which has no relevance to the crop harvest in 2002-03.

Matin Qaim and David Zilberman would have done a yeomen service to the  biotechnology industry if they had also incorporated the results of the field trials conducted a year earlier in 2000. That was the year when Bt cotton crop was sown two months late and still the company claimed that it gave a yield advantage of 50 per cent. This was essentially because the bollworm attack is the heaviest in the first two months of crop sowing and by sowing late the crop had escaped the insect attack. When asked that why wasn't the government advising the farmers to sow the crop two months late
if the yields can go up so dramatically, the GEAC had remained silent. The fact remains that such a claim was completely incorrect and cannot be substantiated in repeat trials. In any case, the outcome of the research trials was certainly known given the fact that the company had paid all expenses for the field trials.

The trials were conducted in plots not exceeding 25x25 metres. The output data was extrapolated to one hectare. And prior to this, the trials were conducted in still smaller size plots of 10x10 meters. It is understandable that you can count the number of plants and then multiply to arrive at the harvest the farmer will reap from one hectare. This is the way agricultural scientists work out the potential yield of crop plants. And this is the
reason why the potential yield is so unrealistic. In case of rice and wheat, the potential of high-yielding varieties has been estimated at eight to nine tonnes a hectare. What the progressive farmers harvest on an average is not more than five to six tonnes. The gap in yield is blamed on the farmers' inability to manage the farm. What the scientists refuse to divulge is that the small plot size is not the scientific way to ascertain yields. Punjab
farmers, for instance, say that they can work out a miracle yield in a plot of 25x25 metres but that does not mean it can be replicated over 10 hectares.

Bt cotton trials and the entire process of monitoring, evaluation and approval has remained shrouded in mystery. The data has been kept classified as if it is country's nuclear deterrent ability that is not to be disclosed.
 
Let us now look at the pest and the pesticides equation. The authors say "under Indian conditions, bollworm have a high destructive capacity that is not well controlled in conventional cotton." This is a strange observation. The authors should tell us in which country they find the bollworms to be less destructive? Is the insect less destructive in America, China or Australia where Bt cotton is being cultivated on a large scale? In any case, American bollworm is a polyphagus insect and feeds on over 90 crops and has a life cycle, which sustains on numerous crops. The insect is perhaps the world's most destructive pest. Behind the references quoted to establish the point that the insect pest damage is substantially lower in China, remains hidden the dubious fact that pesticides use and abuse has created a much bigger crisis on the farm front in China on account of exposure and the resulting health impact. In the early 1990s, China Daily had reported that in one year alone, over 10,000 farmers and farm workers had died from pesticides poisoning. I wonder what they were doing with pesticides if the soil and environment conditions were not so conducive for pest attack.
 
Indian farmers are often indebted and credit constrained and do not have access to chemicals at the right point in time. True, but if the growing indebtedness and credit constraint were the factors there seems to be no justification for pricing the genetically modified seed so high. On an average, the cotton grower spends between Rs 2,500 to Rs 4000 on pesticides sprays depending upon the region and the intensity of insect attack.
Mahyco-Monsanto had provided 450 grams of Bt seed (along with 100 grams of non-Bt seed for refuge plantings) priced at Rs 1600. Farmers normally use one kilo of seed per acre and that means the actual seed cost comes to Rs 3,500. The high seed cost outweighs the advantage from less pesticide sprays that the authors have talked about. Numerous studies have exposed the miscalculation done by the ICAR and the Department of Biotechnology, which had earlier worked out a profit of Rs 10,000 from a hectare. The profit calculations have gone awry in the very first year of planting, with Bt cotton farmers protesting at numerous places in central and south India.
 
Pesticides are not the only external input that is important for cotton. Water consumption is another issue, which is being deliberately ignored in the debate on Bt cotton's economics. Cotton hybrids require more water than the traditional varieties but what is little known is that the water requirement for Bt cotton is much higher than the non-Bt hybrids.

While the Bt cotton fraud is all too apparent, let us take a look at another crisis in the making. The GEAC had only approved Bt cotton hybrids for the central and southern states. For Punjab, Haryana and Rajasthan in northwest, which together have a third of the country's cotton area under cultivation, field trials have been held in the previous crop season. Interestingly, ICAR had asked for three years of research trials before any recommendation can be made. But the Agriculture Minister, Mr Ajit Singh, was so keen that he directed ICAR to forgo the scientific regulations and increase the number of trials so that the approval can be granted on the basis of just one year's data. The ICAR accepted the directive and the field trial data has just been compiled. A year earlier, some adaptive research trials were also conducted, which came in handy to justify the final outcome.

The Punjab Agricultural University conducted field trials in 2002 using Mahyco-Monsanto's Bt hybrids - Mech 12, Mech 162 and Mech 9. In addition, field trials were also performed using Bt cotton varieties (produced by Rasi Seeds) Rch 132 and Rch 138. It is reliably learnt that the best results have been given by the local non-Bt cotton with yield levels of 24 quintals per hectare. Mahyco's Mech 915 yielded 21 quintals per hectare. Rasi's Bt chbrids were higher yielding than Mahyco's. Also, what has been observed is that Bt cotton has less fibre length as a result of which the market is not
very excited. Farmers are therefore getting low price, an estimated Rs 300-400 less on every quintal (100 kilos). In addition, boll shedding is more in Bt hybrids and the insect resistance remains for about 90 days after which the total pest attack multiplies.

And yet, it is an open secret that PAU will be recommending to the GEAC the Bt cotton varieties for approval. The ultimate green signal has to come from the GEAC.

Meanwhile, it is astonishing that no agricultural scientist and economists is excited about the results achieved by some cotton growers in the thick of the cotton belt in Mansa in Punjab. Some farmers, visibly disgusted with excessive use of chemicals, decided to go organic. This year they have harvested cotton to the tune of 28 quintals per hectare and that too without applying pesticides and fertiliser. No economist will like to analyse the
economics and sustainability of the Mansa cotton experiment. The ICAR and the department of biotechnology are also not very enthused with the results. Agricultural scientists must keep their eyes closed, after all resource-starved PAU is looking for joint collaboration with Mahyco-Monsanto. It is the farmer who must continue to pay the actual cost of all these unwanted experimentation, more often than not taking the fatal
route to escape the growing indebtedness from cotton failure over the years.