Access and control of agro-biotechnology: Bt cotton, ecological change and risk in China

Abstract

This article argues that if the introduction of genetically modified crops (GM crops) in developing countries is to be successful, we can and should not evade questions of access and control of technology. It implies probing into the experiences, perceptions and understanding of GM crops by the prime user: the farmer. Exactly in these respects the scholarly literature is remarkably silent. We know little about farmers' experiences and perceptions of GM crops' potential risks and benefits. This is evident when concentrating on a major GM crop – Bt cotton – and studying this in the context of China, its second largest producer in the world. Based on the results of a large survey, we demonstrate that Chinese farmers' awareness (‘having heard of’) and their understanding (‘being able to explain’) of Bt cotton is low. This may lead to ill-informed, distorted risk perceptions and a general inability to relate agricultural production problems to the specific nature of transgenic cotton cultivation. A great majority of the farmers find that the Chinese seed market was liberalised too early, in turn leading to a high incidence of ‘stealth transgenics’ or illegal seeds, the undermining of farmers' trust in private institutions, and a weakened biosafety regime. This finding points to the need for continued state intervention in the seed market, particularly in a developing context. Finally, we have discovered that farmers report a significantly lower reduction in pesticide use by Bt cotton than found in other studies. As suggested by recent research, we suspect that the higher pesticide use is necessary to control secondary pests – i.e. pests other than the cotton bollworm. We present empirical evidence that Chinese farmers perceive a substantive increase in secondary pests after Bt cotton was introduced.

Keywords:

• genetic modification
• secondary pest
• biosafety
• poverty
• pest management
• biotechnology

Notes

¹The global area of GM crops has increased from 1.7 million hectares in 1996 to 114.3 million hectares in 2007; over 10 percent of this is Bt cotton. Note that apart from Bt cotton, there are also other types of genetically modified cotton. Here we will only focus on Bt cotton. See International Service for the Acquisition of Agri-Biotech Applications (2007).

²See the literature review in this article, in the section entitled ‘Previous studies on Bt cotton’.

³For years, China accounted for the world's largest acreage in pest-resistant Bt cotton. For the first time since the introduction of Bt cotton, China was overtaken by India in 2006 (at the time the Chinese area of Bt cotton was 3.5 million ha, while India cultivated 3.8 million ha). China is the world's second largest grower of Bt cotton, and together with the number one, India, their area sown to Bt cotton accounts for 8.7 percent (respectively 3.8 and 6.2 million hectares) of the total area of GM crops in the world (International Service for the Acquisition of Agri-Biotech Applications 2007).

⁴For 2007 data on Bt cotton, see also footnote 3.

⁵In a recent survey conducted among 1,000 urban consumers, we found that few of them are aware that there is such a crop as Bt cotton. When asked to mention one or more GM crops, only a minor proportion mentioned cotton. Chinese consumers' low awareness of Bt cotton stands in stark contrast with the fact that China is the world's second largest producer of it. Of the 1,000 respondents, 32 percent could mention one GM crop, of which only five percent mentioned cotton (Ho and Vermeer 2006, 251).

⁶As is the case in the Comarca Logunera cotton-producing area in northern Mexico.

⁷In northern Mexico, the US, and Canada, Monsanto requires farmers to sign seed licensing contracts in return for the right to purchase and use Bollgard. Through the contracts farmers forfeit rights to replant their own harvested seed, while they are obligated to deliver the cotton to Monsanto-authorised gins. See Eaton et al. (2003). For instance, by buying Brazil's largest seed corporations, the previously domestic hybrid seed industry became 82 percent owned by Monsanto in one single year (Paarlberg 2001, 70).

⁸Monsanto/Delta Pine Land works through joint ventures in China: Andai company in Anhui and Jidai company in Hebei province.

⁹Of the interviewees only 25.3 percent save their own seed. The overall majority of the respondents stated that they do not keep Bt cotton seed, interestingly, not because the company requires this (in only 11 percent of the cases) but because of the decrease in seed quality (84.8 percent). Findings come from this survey.

¹⁰For a review of the critique, see Shantharam et al. (2008).

¹¹Jost et al. (2008) found that the use of transgenic cotton does not provide increased returns to the farmer. This is the conclusion of a 4-year study using different transgenic and non-transgenic cultivars of cotton grown under exactly similar conditions at two locations in the state of Georgia.

¹²The case of DDT is a classical example. Its insecticidal properties were discovered by the Swiss chemist Paul Hermann Müller in 1939, leading to its widespread use to control insect-born diseases (malaria and typhus) during the second world war. In 1948, Müller was awarded the Nobel Prize in Medicine for his research on DDT. Not until 1962, almost a quarter of a century after his discovery, when American biologist Rachel Carson published the book Silent Spring, did the realisation arise that DDT could lead to environmental damage. DDT was not banned in the US until 1972.

¹³Another ecological problem associated with GM crops is its impact on biodiversity through gene flow. This issue is less problematic for cotton as it is an annual plant and self-pollinator. However, in the case of maize, gene flow has already been verified (Quist and Chapela 2001).

¹⁴The laboratory findings apply to China, Australia and India (Downes et al. 2007, Gujar et al. 2007, Wu 2007). To avoid the development of resistance against Bt cotton, it is advised to plant 20 percent of Bt cotton acreage with non-transgenic varieties as a refuge area to avoid resistance development, because the population of bollworm that has become resistant to Bt cotton will then continuously cross with the population of non-resistant bollworm feeding on non-transgenic cotton (and thus has had no chance to develop resistance). But in developing countries with weak regulatory structures and small, fragmented agricultural plots, the enforcement of such refuge areas has been very problematic; China is no exception to this.

¹⁵As Men et al. (2004, 247) write: ‘Transgenic cotton … did not reduce numbers of predators considerably’.

²⁰Farmer interview, Nanshenzao Town, Jiangsu Province, 30 April 2004.

¹⁶More specifically, our study compares non-transgenic cotton with cotton of which the genome contains the Bt gene. As will be described in the section on the seed market, there are a wide assortment of Bt cotton seed varieties. Moreover, Chinese farmers generally experiment with different seed varieties at the same time. Therefore, it is methodologically impossible to compare farmers' views and experiences on different Bt cotton varieties through a sociological or an economic study. To be methodologically fully sound a controlled experiment should be carried out with various groups of farmers only cultivating one type of Bt cotton, and a control group only cultivating conventional cotton. However, it is not the purpose of our article to make a full-fledged comparison of different Bt cotton varieties as that could be better left to biologists. Instead, we intend to get a sense of farmers' personal experiences with GM crops. It should be noted that none of the previous studies on Bt cotton reviewed here in this article have properly accounted for the differences in Bt cotton seed.

¹⁷The village sites and farm households were randomly selected. For reasons of space, the full details of this research, in terms of persons interviewed and research sites, have not been included, but are available from the authors upon request.

¹⁸We had employed similar questions for a survey of urban consumers. For instance, ‘Is it false to say non-GM soybeans do not have genes?’ and ‘Is it false to say eating GM food may change one's genes?’. See also the survey methodology used in Ho and Vermeer (2006, 253).

¹⁹Of the respondents 10 percent had heard of Bt cotton and had some understanding about it; another 19 percent of the farmers knew the term, but were unaware of its linkage with human-made changes in the cotton plant (Zhang 2004, 27).

²¹See, for instance, Hebei Bureau for Agriculture (2003).

²²There is debate on farmers' ability to identify agricultural production problems in relation to the environment (Heisey 1990, Trutmann et al. 1996). Bentley (1992) posited that agricultural production problems that are relatively easy to observe and have a high perceived importance (e.g., weeds or termites) have a higher probability of being reflected in farmers' knowledge. Against this backdrop, it can be posited that Chinese farmers are likely to link the incidence of secondary pests (both easily detectable and accorded with a high importance) with the cultivation shift from conventional cotton to Bt cotton.

²³Through the pilot survey it was found that 67 percent of the sampled farmers think that Bt cotton does not impose any negative impacts on the environment, and 33 percent do not know. These results are not very surprising as the great majority of farmers have no understanding of genetic modification at all. See Zhang (2004, 27).

²⁴These categories have been formulated to falsify hypotheses by forcing the respondents away from affirmative answers if they were not fully sure. The hypotheses were designed on the basis of the most common answers during the in-depth interviews of the pilot study.

²⁵The change could be rated on a five-point scale from ‘strong’ or ‘slight’ decrease and ‘no change’ to ‘slight’ or ‘strong’ increase.

²⁶Interview with official of the Agricultural Service Centre of Zougou Town, Anhui Province, 27 April 2004.

²⁷Bt cultivation started in 1998 in the provinces of Jiangsu, Anhui, and Shandong. The only exceptions are the GM-free zone, Xinjiang, where the cultivation of Bt cotton started illegally in 2000, and Hebei province, where Bt cotton was introduced on an experimental basis for the first time in China in 1996.

²⁸In Jiangsu an increase in only one type of pest, the lygus bug, was reported by 31 percent of the farmers.

²⁹For Anhui, 46 percent stated they saw an increase in spider mites, 33 percent in the beet army worm, and 57 percent in cotton leaf worm; for Xinjiang the figures for spider mites and aphids are 27 percent and 59 percent, respectively.

³⁰The other percentages for Shandong are 18 percent for spider mites and 15 percent for aphids.

³¹The percentages of Hebei farmers seeing an increase in spider mites and aphids are respectively 22 percent and 25 percent.

³²However, in Xinjiang where the incidence of the bollworm is less serious due to the climatic conditions, we see that more farmers – namely 56 percent – indicate ‘higher yields’ as the main reason for choosing Bt cotton. In Xinjiang pest-resistance as a reason for Bt cotton cultivation scored lower: 34 percent.

³³For instance, a study of Zulu smallholders at the Makhathine Flats in South Africa by Bennett et al. (2003) revealed that there had been a reduction in the average number of pesticide sprays per season (from 11.2 to 3.8 sprays) for farmers who adopted Bt cotton. As a result, there were cost savings in the form of lower inputs for pesticide and labour (Bennett et al. 2003, 128). See also Thirtle et al. (2003). Pray et al. (2001) have conducted similar research in China. Their results concur with the reports on Bt cotton in India and South Africa. They report that the reduction in pesticide sprayings ranged from 12 to 3 times (Bennett et al. 2003, 814).

³⁴In July 2006, a research group headed by Per-Pinstrup Andersen presented similar findings at the annual American Agricultural Economics Association. However, these findings have not been published yet. See Lang (2006).

³⁵The only exception was Hebei province, where news through neighbours accounted for 44 percent and news through the agricultural technology and extension station for 32 percent.

³⁶Interview with farmers during a focus group session in Xifeng township, Anhui province, 27 April 2004.

³⁷The cheapest cooking oil in China is cotton oil, which sold for around 6,000–6,200 RMB per kg at the wholesale market in 2004. By comparison, sesame oil, the most expensive cooking oil, sold for 24,000 RMB per kg in the same year.

³⁸In some regions it is even said that cotton oil has certain medicinal properties and is a remedy against greying hair.

³⁹As Kuiper et al. (2001, 503) concluded in a review: ‘the digestibility of [the Bt, PH] protein preparations under simulated conditions is of limited value, as questions can be raised as to whether these assays do mimic the physiological state of human beings … more extensive toxicity testing with the pure protein at exaggerated doses may be required’.

⁴²Interview with official of the Agricultural Service Centre of Zougou Town, Anhui Province, 27 April 2004.

⁴³Interview with Wang Guangming, CEO of the Guangming Grain and Oil Industry Limited Company of Anhui Province, Wuwei County, Jiangba Township, 22 September 2004.

⁴⁰Shandong, Anhui, and Jiangsu provinces have high proportions of farmers who refuse to consume oil from Bt cotton seed (respectively 76.1, 76.4, and 100 percent). Jiangsu is an exception with only two percent. We discovered that at the turn of the century, Jiangsu province was shaken by a food scandal involving contaminated (non-Bt) cotton oil. This food scandal started in Qidong County. The contamination was most likely caused by phenol in cotton oil, the result of improper processing during oil extraction. After the incident few Jiangsu farmers continue to use cotton oil for cooking. This is also reflected by our data: 92 percent of the respondents stated that they would not eat oil extracted from Bt cotton because they consume rapeseed oil. Although the percentages of farmers who indicated that they thought Bt cotton oil was toxic or unhealthy are high for Hebei and Xinjiang (75 percent in both cases), both provinces score substantially lower on the proportions of those who refuse to consume Bt cotton oil (respectively 18.2 and 10.6 percent). We could find no plausible explanation for this phenomenon. Therefore, only for Anhui and Shandong could a solid relation be established between farmers' perception of food safety and Bt cotton.

⁴¹Farmer interviews, Zougou Town, Anhui Province, 27 April 2004.

⁴⁴As a Monsanto representative stated: ‘These things [frustration of Monsanto's activities in China] go back to 1996 when Guo Sandui [of CAAS] started his research on Bt cotton. Monsanto started in 1998 in Hebei province, which made the ministry rather anxious. Officials wanted to ensure that Chinese Bt cotton would come on the market soon. Later, the Ministry of Agriculture said we missed certain data for the biosafety approval, but others told us this was not the real reason. I believe the Chinese Ministry of Agriculture has held back our biosafety approval in favour of local Bt cotton research to catch up’. William Deng, Technical Representative responsible for Bt cotton and former general manager of Andai Company, 4 May 2004, Beijing.

⁴⁵The sale of fake seeds was limited in Hebei (3.5 percent had direct experiences with fake seed and 3 percent knew neighbours or friends with that experience), while Anhui farmers encountered serious problems with an overwhelming 44 percent of the respondents having personal experiences with fake seed and 44 percent personally knowing people who had bought fake seed.

⁴⁶As a Monsanto representative maintains: ‘We estimate that there are over 100 different kind of illegitimate seed varieties, of which 80–90 percent are in fact 33B. However, they are sold on the market as “American Bt cotton No. 1” or something fancy like that. In this way, biosafety procedures are evaded’. William Deng, Technical Representative responsible for Bt cotton and former general manager of Andai Company, 4 May 2004, Beijing.

⁴⁷Local Chinese officials have reported the heterosis effect for hybridised Bt cotton. Under similar production conditions, it is estimated that Bt cotton has a productivity of around 400–500 kg per mu (1/15 ha), while hybrid Bt cotton can reach 500–600 kg/mu. Interviews with officials of the Bureau for Agriculture of Wangjiang and Dongzhi County, Anhui, 23–26 September 2004; Dafeng City, 28 April 2004, Jiangsu; and the head of the Agricultural Technology and Extension Station, Hua Jianhuang, Dazhong Town, Jiangsu, 28 April 2004.

⁴⁸The official Chinese standard for the germination rate has been set at 72 percent of the seed.

⁴⁹As the law stipulates: ‘If the seed user undergoes loss because of the seed quality, the seed supplier should compensate for it, and the compensation includes the money number used to buy the seeds, related expense and the available profit loss.’ (Article 41 of the 2000 Seed Law).

⁵⁰Only 27 percent obtained some kind of compensation either financially or in kind from the concerned seed company.

⁵¹Farmer interview, 30 April 2004, Nanshenzao Town, Jiangsu province. At the time one RMB had a value of 0.0994 Euro.

⁵²Of the respondents, 78.8 percent thought that the state could regulate the market, 16.9 percent had some or serious doubts about it, while 4.3 percent did not know.

⁵³Interview with official of the Agricultural Committee of Wuwei County, Anhui Province, 27 April 2004.

⁵⁴Biosafety regulations are also hard to enforce in India and Brazil. As Herring (2007, 130) noted: ‘the ability of seeds to move underground … undermines widely-assumed biosafety-regime capability’.

⁵⁶From an interview with respondent 29/82008MoA. For reasons of privacy the name and department of this official cannot be given here.

⁵⁵As stipulated in article one of the 2000 Seed Law.

⁵⁷This problem is also noted by Tripp (2002), although he regards the lack of information as a managerial problem rather than a rights problem. Yet, an important difference between GM and conventional crops is their level of public contestation, which may affect their marketing. In this context, farmers should have a ‘right to know’ the potential risks of GM crops.

⁵⁸Of the global area sown to GM crops, 43 percent (49.4 million hectares) were cultivated in developing countries where growth between 2006 and 2007 was ‘substantially higher (8.5 million hectares or 21 percent growth) than industrial countries (3.8 million hectares or 6 percent growth)’ (International Service for the Acquisition of Agri-Biotech Applications 2007).

Additional information

Notes on contributors

Peter Ho

We would like to thank Sjaak Swart for his ever-detailed and helpful comments, as well as the suggestions from the two anonymous referees of this journal. This research was supported by the Royal Dutch Academy of Sciences, Project 08CDP 010.