Tillmann Elliesen (TE; 1) used the ballgame of Bt cotton in India to support his theory that ‘the critics of genetic engineering are increasingly becoming an entrenched fundamental opposition--and thus hurting their credibility’.   Had he also used the misinformation campaign and blind opposition of the critics of genetic engineering (GE) to GE mustard and Golden Rice in India, his purpose would have been served admirably.

The provocation for the present writing is that TE seems to provide succour to the paper published in SCIENCE by Quim and Zilberman (Q & Z; 2) on the performance of Bt cotton in India.

Q & Z (2) used Bt cotton in India to support their theory that ‘currently existing GM crops can have significant yield effects that are most likely to occur in developing world, especially in the tropics and subtropics’. Nothing wrong with the theory, but only with the use of one year’s field trial data on a crop reeling in teething problems, to support it.   The data used do not alter their other observation that ‘the limited experience with GM crops so far is insufficient to make broad generalizations about their impacts’, as far as the developing countries are concerned.   The encouraging performance of Bt cotton in the US, Africa, China and Indonesia and the annual increase acreage under Bt cotton in more and more countries is an adequate testimony of the benefits derived from Bt cotton in the other parts of the world.

Although there seems to be nothing wrong with Q & Z’s (2) basic claims of the performance of Bt cotton in India, the problem is with the kind of data they used to make sweeping generalisations.   The conclusions are not borne out by the data and we cannot use theories to convince the public and these are not needed to convince the scientific people who already know that Bt technology has a high promise, given the right conditions.   It should be clear that all of who were critical of Q & Z’s paper are not against GE technology.   Many of us, who are supporters of technology, emphasise the need to convince the public of its safety and demonstrate cost-benefits, and that such support of technology should be based on unassailable scientific grounds.  Q & Z’s paper left broad sides open for severe criticism, as they based their conclusions on just one season’s (that of 2001) field trial data and not data from commercial cultivation in the season of 2002.   The purpose of the field trial data is to convince the regulatory authority of the benefits of the technology to accord permission for commercial cultivation, after which the field trial data belong only to the archives.  

Qaim (3) accepted that the cotton season of 2001, on which their analysis was based, was a high pest pressure year and so the differences between Bt and non-Bt would be high.   On the other hand, the cotton season of 2002 is a low pest pressure year and so the yield differences between the two are quire low.   The critics of Bt technology are using the low differences of this year’s commercial cultivation and argue that the projections of Q & Z (2) are exaggerated.   Added to this, the season of 2002 is a draught year where not only cotton but also most of agriculture in India was affected to the extent that the Central Government sanctioned huge draught relief grants to different States, many of which constitute the major cotton cultivating areas.

The GEAC permitted commercial cultivation of Bt cotton for three years.   Why cannot the supporters and opponents of Bt technology wait till the end of this period to proclaim Bt cotton in India, as a success or a failure?   In the meanwhile they can certainly monitor the situation, in order to analyse the factors responsible for the outcome each year and to have an idea of what is likely to come up at the end of the three-year period.

TE (1) missed almost all of this.   I am surprised that TE did not mention anything written by Dr Shantharam, who had actually kick started the debate on Q & Z’s paper (2).   As Qaim (3) mentions, most of Dr Shantharam’s criticism relates to the sources and the quality of the data.   Had this not been the case, and had the data related to results of commercial cultivation, Dr Shantharam and I would have celebrated Q & Z’s paper.

TE (1) added to the controversy by saying that “Mahyco-Monsanto neither supplied the data nor funded the research”.

I do not consider data as suspect simply because they come from the company that is marketing a product.   In the face of TE’s emphasis that Mahyco-Monsanto did not supply the data used by Q & Z, consider the following:  a) at Reference 30 (2), Q & Z thanked Mahyco “for making the field-trial records available”, and b) Qaim (3) wrote that “We have used the company field-trial records about the pest infestation levels, such as larval counts per plant.   These were collected during weekly trial visits by local company agronomists, and we received the complete bundle of handwritten field records, not just aggregated summary statistics.”   Qaim was anxious to impress that the data were collected basing on “our own survey form”, that the data collection was “initiated by us”, that “we entered the data from the questionnaires into the computer ourselves” and “conducted the analyses completely independently from Mahyco or Monsanto”.   I do not put any value on this issue, as it does not matter who did what clerical job. 

The criticism of Q & Z’s paper (2) was on the type of data, that they were only field trial records.   Even if Mahyco-Monsanto supplied the analysed data in the final format, many would have accepted without suspicion, may be with a pinch of salt, provided they related to commercial cultivation.

A study of the Indian Council of Agricultural Research (ICAR) (4), that monitored the Bt cotton field trial prior to GEAC’s approval, also used the field data of 2001 and projected increase in yield ranging from 60 to 92 per cent.   This study was cited by the ISAAA(5).   I do not understand why Q & Z did not use the field trial records of all the five years (seasons of 1997 to 2001) and that why they sho chose only 157 locations (on what basis?) out of 395 locations of the same year? (Ref. 12, Science paper), which makes it less than 42 per cent of the total trial locations.   This is not even substantial, yet Qaim stated that the ‘data are far from meagre’ (3).   When Qaim himself observed that the gains in field trials could not simply be transferred to commercial agriculture (3), what purpose analyses of field data of 2001 serve?

The performance of a particular Bt cotton variety should be compared only with that of its isogenic variety.   Q & Z’ (2) did this but also made comparisons with ‘popular check’ without mentioning the check varieties.   When five varieties were under field testing, three of which the GEAC has approved for commercial cultivation, it is difficult to read the relative performance of the different Bt varieties chosen in the study, when the data from all varieties of Bt cotton are pooled (Table 1, in 2).   The relative performance of different Bt cotton varieties is the not the same.   Actually, Q & Z (2) did not mention any specific varieties of Bt cotton.   Since it is now known that Bt varieties like MECH 162 are not superior to several local varieties, the yields of the local popular check varieties cannot be lower than those of the isogenics (Table 1 in 2) in all the cases.

Qaim (3) wrote that “having isogenic hybrids with and without the Bt gene growing next to each other, and controlling for other inputs, ‘it is an appropriate inference that most of the yield effects is due to the Bt gene’”.   This position is scientifically untenable, as they have overlooked the vagaries of conditions of cultivation.   There is no mention of negative control and there was no randomised design to attest to the statistical significance of the results.   Yield increase in Bt cotton is due to realisation through protection from loss and not a real increase on account of any gene of the Bt technology.   The function of the Bt gene to afford control of the bollworm and nothing more.

Qaim (3) accepted that the “Science paper is not a substitute for a careful analysis of broader Bt cotton impacts in India in commercial agriculture”, and yet chose to extrapolate the analysis into broad generalisations.   He also wrote that (Bt strategy) “should be part of an integrated pest management strategy that assesses technologies according to their real impacts”.   This is not a new finding, as Bt is already a part of IPM strategy in India.   On both the counts, there is no need for this paper.  

Qaim (3) wrote, “Some of the broader conclusions go beyond the statistical analyses, but they area based on theory and lessons from the crop-protection literature”.   May be so, but stretched extrapolation damages the cause.

Another point made by Q & Z (2) and emphasised by TE (1) is about the source of funding.   I believe that the MNCs that develop and market GE products also have a responsibility to support surveys on product performance and programmes on public perception, awareness and education on technology issues.   Had Mahyco-Monsanto supported the study of Q & Z’s study, it would be alright with rational people.   There is no one need to be apologetic to accept such funding nor be anxious to convince that no such funding was received.   I do not understand how the value and credibility of a particular study becomes elevated simply because it was supported by a public finding agency.   Here is an unwarranted commentary on the integrity of the scientific community.   We should rise above this kind of implicit linking of our credibility to the kind of funding we receive for our study, and get into a position where people would trust us, no mater who funds our study.   I wonder, how can (and why should) a public institution spend money on a project related to a product generated and commercialised by a private enterprise?

The veracity of a study and the value of its conclusions depend upon the integrity of the investigators, the quality of the data and the rigour of the analysis, and not on who funded the project.   I have been a witness for four decades or so, as to how much public research funding goes down the drain in India because of poor quality of science and lack of integrity on the part of the researchers.   Few black sheep should not be the reason for undermining the credibility of the whole scientific community.  

References:

1. Elliesen, T.  The futile dispute over genetic engineering.  D + C, 30: 207-209, 2003.5
2. Qaim, M., and Zilbermann, D.  Yield of genetically modified crops in developing countries.  Science, 299: 900-902, 2003.
3. Qaim, M., in AgBioView, in response to comments on Q & Z’ paper (2), February 21, 2003.
4. Indian Council of Agricultural Research (ICAR).  Report on 2001 IPM trial cost benefit analysis.  2002, New Delhi.
5. Bt Cotton in India.   ISAAA Monograph, March 2003.

Professor C Kameswara Rao
Foundation for Biotechnology Awareness and Education
Bangalore, India (krao@vsnl.com)