Transgenic cotton: High hopes and farming reality

To determine the potential of any promising tool, its performance in practice must always be considered. Two recent articles reach different conclusions on one important benefit of Bacillus thuringiensis cotton management: the potential to reduce pesticide sprays.

By emphasizing the merits of genetically modified (GM) crops as a technology, we can sometimes forget that agriculture, like other work, involves social mediation — the plants do not grow themselves. In India, the labour required for spraying, weeding and plucking makes cotton an especially toxic plant. Most of this labour is done by hand, often by poor and lower-caste people without land of their own, by young people, and by women. A recent paper by Veettil et al.1 shows that the use of transgenic cotton with genes for insecticidal proteins from Bacillus thuringiensis (Bt cotton) has slowed pesticide exposure by limiting the overall need for pesticide sprays when compared with non-Bt cotton. This conclusion is in line with previous analyses2,3 of the changes in Indian cotton production since Bt cotton's commercialization in 2002. However, a recent article by Venkata et al.4 argues that this analysis misses the point. Even as exposure is dropping, Bt cotton is still sprayed with pesticides, presenting risks for farmers and the wage labourers who weed and pluck the cotton. Use of broad-spectrum pesticides is a major issue for the health of agricultural workers and farm ecosystems.

While the merits and risks of GM crops remain contested, much of the debate has focused on crop yields and the overall potential for increased production or profits. Noting that pesticide reduction is the raison d'etre of Bt cotton, Veettil et al. looked at the impact of Bt cotton on pesticide toxicity. They built a model to quantify Bt cotton's environmental impact during the 2002–2008 period when GM cotton was adopted in India, from which they argue that farmers who plant Bt cotton reduce their pesticide use generally, and their use of highly toxic pesticides specifically, when compared with farmers planting non-Bt cotton. This is an important finding with respect to environmental and public health, as cotton is one of the most heavily sprayed crops and potent insecticides have the potential to persist in soil and on skin throughout the long cotton season. By accounting for the variable toxicity of pesticides in their analysis, Veettil et al. provide a nuanced explanation for Bt cotton's role in encouraging farmers to use pesticides that are less harmful to human health and less likely to cause damage throughout the farm ecosystem.

These are positive results but not enough, suggest Venkata et al.4 They focused narrowly on the DNA damage, enzyme structure and haematological profiles of active cotton workers exposed to pesticides through their labour. This research asks not how farmers are changing their pest management practices as they move toward near-universal Bt cotton planting, but how Bt cotton is managed in practice. Venkata et al. found that workers reported significantly greater qualitative health problems, ranging from fatigue to hair loss, and experienced significantly greater DNA and chromosomal damage, as assessed by a micronucleus frequency test of blood and buccal cheek samples, than a control group of non-agricultural workers. These results align with other studies of pesticide exposure among farm workers5,6. However, this new study combines molecular and qualitative health analyses and focuses on the risks encountered by a sample composed of young female cotton workers, the population who provide the primary manual labour in Indian cotton fields. Despite reductions in highly toxic pesticide use since 2002, cotton agriculture remains a risky profession for farm workers.

The differing conclusions of these articles and their ramifications for Indian Bt cotton farmers can be explained in part by their methodologies. Veettil et al. draw on a large, national sample of 341–380 farmers in 58 villages across 4 states, surveyed every 2 years between 2002 and 2008. While this approach allows the authors to comment on a broad swath of Indian farmer experiences, their respondents represent relatively few people in any given village, and their panels do not assess the risks of pesticide use among the farm labourers, the group at highest risk of exposure. Furthermore, they report that 99% of their respondents were planting Bt cotton by the final year of the survey, and so their non-Bt control groups are not representative of normal cotton management practices. As Indian farmers have overwhelmingly adopted Bt cotton seeds, the most accurate assessment of the toxicological risk to public health and the farm ecosystem would come not from a comparison of Bt and virtually non-existent non-Bt farmers, but from an analysis of workers exposed to pesticides in Bt cotton fields.

This is the approach that Venkata et al. use. It is regionally limited when compared with Veettil and colleague's national scope but it presents compelling evidence for the continued risks of cotton agriculture, even when Bt cotton adoption is virtually universal. By sampling a group of active cotton workers, Venkata et al. were able to capture field conditions that influence risk factors, including negligible use of protective garments, exposure to mixtures of pesticides, and the long duration of Indian cotton farming. By quantifying the environmental impact as a function of different pesticides applied, Veettil et al. provide an important step in measuring the aggregate effects of Bt cotton on pesticide use. However, by measuring molecular damage among cotton workers, Venkata et al. illustrate that reduction alone is not enough.

Both studies contribute to the complicated scholarship of transgenic crops by asking readers to consider the synergistic, emergent effects of a GM technology. Bt cotton modified to express insecticidal cry genes capable of defending cotton from lepidopteran pests has reduced the required pesticide load for US cotton farmers by as much as 28% (ref. 7) since their adoption in 1996. These issues have high stakes in India, which boasts the most land under cotton cultivation, the largest population of small farmers growing GM crops, and more than 90% Bt cotton adoption by farmers. Initially, it seemed that the Bt cotton reduced pesticide sprays in India, an improvement for both farmer budgets and exposure to toxicity3,8,9. Yet by 2010, when Bt cotton was ubiquitous in India, total insecticide applications had largely returned to their pre-GM levels8 (Fig. 1). One reason for this is the rise in spray applications for non-lepidopteran sucking pests capitalizing on a new ecological niche. Additionally, Indian farmers rarely plant refuge areas of non-Bt cotton designed to slow the evolutionary selective pressure toward Bt resistance.

Figure 1: Metric tonnes of insecticides used in India between 1995–2011 against major insect pests.

Data taken from ref. 8.

Just as important are the social reasons that drive farmers to continue to spray their cotton with pesticides: the need to be seen caring for one's fields, the uncertainty that drives farmers to spray when it is not agronomically necessary, the ineffectiveness of expired or improperly applied pesticides, or even the impulse to spray because one's neighbour is spraying. Venkata et al. observe that farm workers that are described in toxicology literature as wearing protective clothing and having access to washing stations experience a much lower risk of contamination, a dynamic that speaks to the need for pesticide protection designed for poor workers in tropical and sub-tropical climates. On small farms in rural India, farmers also encounter pesticides in another form10. A cotton field, prime real estate for farmers, often has gaps where a small percentage of the cotton fails to germinate. Not wanting to waste fertilized or irrigated land, farmers often plant vegetables in these gaps to bolster local food security (Fig. 2). When the cotton is sprayed, these food crops are sprayed as well.

Figure 2: Roselle (Hibiscus cannabinus L.), which is eaten raw as a chutney, growing in gaps between Bt cotton plants in Telangana, India.

Image courtesy of Andrew Flachs.

Veettil et al. never imply that GM cotton is a panacea to pesticide overuse, and they conclude both that transgenic seeds must be an element within a broader agricultural development strategy and that many Bt cotton farmers continue to overuse chemical pesticides. However, their comparisons between Bt and non-Bt cotton farmers miss important social dynamics driving farmer pesticide use: namely, the consequences of continued pesticide spraying even when Bt cotton obviates that need, and the risk that this poses to the largely young, poor, and female farm labouring population. Bt cotton is here to stay in India (and globally), and its impacts should be measured in practice, not as a comparison to the few, unrepresentative non-Bt planting holdouts. There is every reason to believe both that Bt cotton has helped reduce the quantity and overall toxicity of cotton agriculture, and that this profession, in practice, remains a precarious one.


  1. 1

    Veettil, P. C., Krishna, V. V. & Qaim, M. Aust. J. Agr. Resour. Econ. (2016).

  2. 2

    Qaim, M. New Biotechnol. 27, 552–557 (2010).

    CAS  Article  Google Scholar 

  3. 3

    Kouser, S. & Qaim, M. Ecol. Econ. 70, 2105–2113 (2011).

    Article  Google Scholar 

  4. 4

    Venkata, R. P. et al. Biomarkers (2016).

  5. 5

    Damalas, C. A. & Eleftherohorinos, I. G. Int. J. Environ. Res. Publ. Health 8, 1402–1419 (2011).

    CAS  Article  Google Scholar 

  6. 6

    Dawson, A. H. et al. PLoS Med. 7, e1000357 (2010).

    Article  Google Scholar 

  7. 7

    Benbrook, C. M. Environ. Sci. Eur. 24, 24 (2012).

    Article  Google Scholar 

  8. 8

    Kranthi, K. R. Bt Cotton Q&A (Indian Society for Cotton Improvement, 2012).

    Google Scholar 

  9. 9

    Qaim, M. Annu. Rev. Resour. Econ. 1, 665–694 (2009).

    Article  Google Scholar 

  10. 10

    Flachs, A. J. Ethnobiol. 35, 406–426 (2015).

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Andrew Flachs.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Flachs, A. Transgenic cotton: High hopes and farming reality. Nature Plants 3, 16212 (2017).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing