The development and status of Bt rice in China

Summary

Multiple lines of transgenic rice expressing insecticidal genes from the bacterium Bacillus thuringiensis (Bt) have been developed in China, posing the prospect of increases in production with decreased application of pesticides. We explore the issues facing adoption of Bt rice for commercial production in China. A body of safety assessment work on Bt rice has shown that Bt rice poses a negligible risk to the environment and that Bt rice products are as safe as non‐Bt control rice products as food. China has a relatively well‐developed regulatory system for risk assessment and management of genetically modified (GM) plants; however, decision‐making regarding approval of commercial production has become politicized, and two Bt rice lines that otherwise were ready have not been allowed to enter the Chinese agricultural system. We predict that Chinese farmers would value the prospect of increased yield with decreased use of pesticide and would readily adopt production of Bt rice. That Bt rice lines may not be commercialized in the near future we attribute to social pressures, largely due to the low level of understanding and acceptance of GM crops by Chinese consumers. Hence, enhancing communication of GM crop science‐related issues to the public is an important, unmet need. While the dynamics of each issue are particular to China, they typify those in many countries where adoption of GM crops has been not been rapid; hence, the assessment of these dynamics might inform resolution of these issues in other countries.

Introduction

A rapidly developing country with relatively limited arable land such as China has devoted great effort into developing genetically modified (GM) crops to boost agricultural productivity. In particular, China is playing a leading role in development of insect‐resistant GM rice Oryza sativa lines (Li et al., 2014a). Many rice lines expressing insecticidal genes from the bacterium Bacillus thuringiensis (Bt) have been developed, posing increases in rice production with decreased application of pesticides (Chen et al., 2011). Developers of two Bt rice lines – Bt Shanyou 63 and Huahui 1 – have obtained biosafety certificates, but the lines have not been approved for agricultural production. Against this background, we note that several issues face commercial adoption of Bt rice in China, including ecological risk assessment, food safety assessment, regulatory issues, adoption by farmers and public acceptance. We explore each of these topics, assessing progress in each area. While the dynamics of each issue are particular to China, they typify those in many countries where adoption of GM crops has not been rapid; hence, the assessment of these dynamics might inform resolution of these issues in other countries.

Issues facing rice production in China

Rice is one of the most important crops worldwide and is a primary food source for more than half of the world's population. China is the largest rice producer and consumer in the world; approximately 20% of arable China's land is grown with rice, a larger share than those for other agricultural crops (Chen et al., 2011). With the increase of China's population, Chen et al. (2011) predicted that rice yield has to be increased to 7850 kg/ha by 2030, resulting in a total of 200 million kg of rice per year to feed the Chinese people. Among multiple constraints to rice production, insect pests are an important factor (Wang et al., 2014a,b). In Chinese rice ecosystems, four major lepidopteran pests – rice striped stem borer Chilo suppressalis (Family Crambidae), yellow stem borer Scirpophaga incertulas (Family Pyralidae), pink stem borer Sesamia inferens (Family Noctuidae) and rice leaf roller Cnaphalocrocis medinalis (Family Pyralidae) – cause serious damage to rice, resulting in severe yield losses in every crop season (Chen et al., 2011). The stem borers are the most serious pests, and in China, about US$19 million is lost every year due to stem borer damage (Sheng et al., 2003). Large amounts of chemical insecticides are applied each year as the main measure for controlling these pests. The application of chemical insecticides has resulted in a series of problems, such as air, water and soil pollution, food contamination and the development of pest resistance to the insecticides (Huang et al., 2001).

Genetic engineering is one of the most powerful 21st‐century technologies, and its use is driving the ‘new green revolution’ in agriculture (Li et al., 2014a). Worldwide, the use of GM crops has increased rapidly since the first commercialization in the United States in 1996. In 2013, GM crops were grown on 175 million ha, of which 43% expressed an insect‐resistant trait (James, 2013). The experience with planting of Bt maize and Bt cotton suggests that the use of insect‐resistant GM (IRGM) crops can significantly reduce pesticide use and, as a result, decrease environmental impacts associated with application of chemicals to these crops (Brookes and Barfoot, 2013). Therefore, development of IRGM rice expressing Bt insecticidal proteins is an attractive alternative for pest control (Chen et al., 2011).

China has devoted great efforts to developing GM rice lines. Since the first Bt rice plant was developed in China in 1989, so far over a dozen Bt rice lines have been developed (Table 1). Among these lines, some contain single Bt insecticidal genes such as cry1Ab in the Kemingdao line, cry1C* in T1C‐19 and Zhonghua 11, and the cry1Ab/Ac fusion gene in Bt Shanyou 63, TT9‐3 and TT9‐4. Some lines contain dual insecticidal genes such as CpTI (cowpea trypsin inhibitor) and cry1Ab in the MSA or MSB lines and cry1Ab/Vip3H in the G6H line (Akhtar and Ye, 2012; Chen et al., 2006). In addition, some Bt rice lines have been stacked with other types of transgenes, such as the bar gene for herbicide tolerance (Yao et al., 2002) and xa21 for disease resistance (Wang et al., 2002a). Most of the Bt rice lines have been shown to express high resistance to target pests in laboratory and field trials (Cohen et al., 2008; Han et al., 2011; Sui et al., 2011; Tang et al., 2006; Tu et al., 2000; Wang et al., 2014a,b; Zhang et al., 2011a; Zheng et al., 2011). In 2009, China's Ministry of Agriculture issued biosafety certificates for commercial production of the Huahui 1 and Bt Shanyou 63 Bt rice lines in Hubei Province, which raised expectation that China would use GM technology in commercial rice production (http://www.rsc.org/chemistryworld/China/Issues/2010/April-June/GMHotlyDebatedInChina.asp). The biosafety certificates were renewed on 11 December 2014; however, the Bt rice lines have not been commercially planted to date (http://news.xinhuanet.com/tech/2015-01/07/c_127364741.htm).

Table 1.

Insect‐resistant Bt rice events developed in China

No.	Event	Gene transformed	Promoter; method of transformation	Recipient cultivar	Trait (target insect)	References
1	–	Bt aizawai 7‐29a	CaMV 35S; Polyethylene glycol method	Taibei 309 (japonica)	IR (L)	(Yang et al., 1989)
2	–	Bt aizawai 7‐29a	CaMV 35S; Pollen‐tube pathway	Zhonghua 11 (japonica)	IR (L)	(Xie et al., 1991)
3	KMD1/KMD2	cry1Ab	Ubiquitin; Agrobacterium‐mediated	Xiushui 11 (japonica)	IR (L)	(Shu et al., 2000)
4	–	cry1Ac	Ubiquitin; Agrobacterium‐mediated	Xiushui 11 (japonica)	IR (L)	(Xiang et al., 1999)
5	–	Cry1Ac	Ubiquitin; Gene gun‐mediated	Minghui 81 (indica)	IR (L)	(Zeng et al., 2002)
6	–	Cry1Ac	Ubiquitin; Agrobacterium‐mediated	Minghui 63 (indica)	IR (L)	(Chen et al., 2008)
7	T1C‐19	cry1C	Ubiquitin; Agrobacterium‐mediated	Minghui 63 (indica)	IR (L)	(Tang et al., 2006)
8	RJ‐5	cry1C	Rice rbcS promoter; Agrobacterium‐mediated	Zhonghua 11 (japonica)	IR (L)	(Ye et al., 2009)
9	T2A‐1	cry2A	Ubiquitin; Agrobacterium‐mediated	Minghui 63 (indica)	IR (L)	(Chen et al., 2005)
10	–	cry9C	Ubiquitin; Agrobacterium‐mediated	Minghui 63 (indica)	IR (L)	(Chen et al., 2008)
11	Huahui1	cry1Ab/1Ac fusion	Actin1; Gene gun‐mediated	Minghui 63 (indica)	IR (L)	(Tu et al., 2000)
12	TT9‐3/TT9–4	cry1Ab/1Ac fusion	Actin1; Gene gun‐mediated	IR72 (indica)	IR (L)	(Ye et al., 2001)
13	mfb‐MH86	cry1Ab	Ubiquitin; Agrobacterium‐mediated	Minghui 86 (indica)	IR (L)	(Wang et al., 2014a)
14	MSA, MSB and MSA4	Cry1Ac/sck	Ubiquitin; Agrobacterium‐mediated	Minghui 86 (indica)	IR (L)	(Liu et al., 2006)
15	–	cry1Aa+pta	CaMV 35S, Ubiquitin; Agrobacterium‐mediated	Jijing81, Jijing88 and Tong887 (japonica)	IR (L&H)	(Lin et al., 2006)
16	Kefeng6	cry1Ac+CpTI	Actin1; Gene gun‐mediated	Minghui 86 (indica)	IR (L)	(Zhao et al., 2004)
17	–	cry1Ab/Ac+Xa21	–	Minghui63, Zhenshan97A and Maxie A (indica)	IR (L), DR	(Jiang et al., 2004a)
18	–	cry1Ab/Ac+bar+Xa21	Actin1; Gene gun‐mediated	Zhongguo 91 (japonica)	IR (L), HR, and DR	(Wang et al., 2002b)
19	G6H1, G6H2, G6H3, G6H4, G6H5, and G6H6	cry1Ab/vip3H+epsps	Ubiquitin; Agrobacterium‐mediated	Xiushui 110 (japonica)	IR (L), HR	(Fang, 2008)

IR, insect resistant; HR, herbicide resistant; DR, disease resistant; L, lepidoptera; H, hemiptera; ‘−’ denotes being unclear.

^aNo information was found to specify the Bt genes using the current nomenclature.

Issues facing adoption of Bt rice in China

There are several main issues that directly relate to the commercialization of a GM crop, including ecological risk, food safety, biosafety regulation, adoption by farmers, and ultimately public acceptance. We discuss these issues with the objective of analysing the main problems preventing commercialization of GM rice in China, with the intent of sparking critical discussion of these issues both in China and globally.

Biosafety regulation for agricultural biotechnology

Having targeted agricultural biotechnology as an area of strategic scientific investment, the Chinese government also has devoted considerable attention to the development of public policy for regulatory oversight of agricultural biotechnology. In 1993, the Chinese Ministry of Science and Technology (MST) issued the first biosafety regulation for agricultural biotechnology, which stated general principles, safety issues, the need for risk assessments, application and approval procedures, safety control measures and legal responsibilities (Li et al., 2014a). The regulation was followed by the Implementation Guidelines issued by the Ministry of Agriculture in 1996. However, the early Chinese biosafety regulations were not very detailed. The State Council of China (2001) subsequently decreed a new and general policy for administrative regulation of GMO biosafety, the Regulation on Safety Administration of Agricultural GMOs (RSAGMOs) to replace the early regulations. The RSAGMO regulation provides detailed prescriptions for laboratory research, biosafety testing, production, processing, marketing and import and export activities with respect to agricultural GMOs. Following the State Council Regulations, the MOA in 2002 announced a series of three implementing regulations (Li et al., 2014a). In addition, the Ministry of Public Health also promulgated the regulation on GMO food hygiene. These regulations collectively covered all aspects associated with development of GMOs, including laboratory research, food and environmental safety testing, labelling, production, processing, marketing and import and export of GMOs and GMO‐derived products, thereby providing the basis for promoting the sustainability of agricultural biotechnology in China.

In addition, China also has developed a comprehensive regulatory framework for regulating GMO biosafety. To incorporate representation of stakeholders from different ministries, in 2002, the State Council established an Allied Ministerial Meeting, comprised of the leaders from the relevant departments, which is responsible for decision‐making and coordination of major issues regarding the safety regulation of agricultural GMOs (Li et al., 2014a). In the State Council Regulations, the MOA is designated as the leading and primary institution responsible for nationwide supervision and inspection of agricultural GMO biosafety (State Council, 2001). Under the MOA, an Office of Agricultural Genetic Engineering Biosafety Administration (OGEBA) was created in 1996 for dealing with the routine work and daily operations with respect to GMO safety. In addition, the National Agricultural GMO Biosafety Committee (BC) was formed by the MOA in 1997 with responsibility for the management of agricultural GMO research and risk assessment (Li et al., 2014a). Other than the MOA, other ministries also are involved in routine regulation of GMOs and related products. The National Health and Family Planning Commission of China is responsible for technical issues of GMO‐related food safety, and the General Administration of Quality Supervision, Inspection and Quarantine of China, for issues related to import and export of GM agricultural products (Li et al., 2014a).

With over 20 years of nationwide promotion of agricultural biotechnology, a relatively well‐developed regulatory system for risk assessment and management of GM plants has been developed to establish a firm basis for safe use of GM crops. From 2002 to 2012, the MOA approved 2775 applications for pilot‐scale testing of GM plants, 459 for environmental release for field testing and 317 for preproduction testing and has issued 1830 biosafety certificates. However, only seven GM crops involving 10 integration events have been approved by the MOA for commercial planting in China (Li et al., 2014a). Our analysis comparing current biosafety regulation and commercial use of GM crops in China with those in other countries (Li et al., 2014a) suggested that China might streamline the regulatory process for agricultural biotechnology.

Safety assessment of Bt rice

According to Chinese biosafety regulations, the potential risks to the environment and human health have to be fully assessed prior to commercialization. Since 1998, a large number of laboratory and field tests have been conducted for assessing the potential effects of Bt rice on the environment, respectively, focusing on effects on nontarget arthropods, soil ecosystems, potential for gene flow and effects on human health associated with Bt rice lines (Chen et al., 2006, 2011; Cohen et al., 2008; Lu and Snow, 2005; Yu et al., 2011).

Nontarget effects of Bt rice

One of the risks associated with the growing of insect‐resistant GM rice plants is their potential to adversely affect nontarget organisms. The assessment of such effects is thus part of the environmental risk assessment that is conducted prior to commercialization of any novel GM variety, especially for IRGM plants. Natural enemies of pest arthropods are of particular interest, as they impose demographic regulation of herbivores, an important ecological function, and thereby contribute to a sustainable agro‐ecosystem (Li et al., 2013, 2014b,c; Romeis et al., 2006; Yu et al., 2011; Zhang et al., 2014). Assessment of IRGM crops on nontarget organisms follows a tiered framework that is conceptually similar to that used to assess the environmental impact of conventional chemical plant‐protection products (Garcia‐Alonso et al., 2006; Romeis et al., 2008; Rose, 2007). Within the tiered framework, assessments typically start with laboratory experiments under worst case and confined conditions (Garcia‐Alonso et al., 2006; Li et al., 2014d; Romeis et al., 2008, 2011). If no effect is observed in the laboratory studies, semifield and open‐field studies normally are not needed (Romeis et al., 2008; Yu et al., 2011). However, according to China's GMO risk assessment guideline, field studies for nontarget effects of IRGM crops are required.

Recently, several dietary exposure tests have been conducted specifically for assessing the potential toxicity of Bt rice‐produced insecticidal proteins on nontarget arthropods occurring in rice ecosystems (Li et al., 2013, 2014c,d, 2015; Wang et al., 2012, 2015; Zhang et al., 2014). The potential toxicity of Cry proteins such as Cry1Ab, Cry1Ac and Cry2A that are expressed in multiple Bt rice lines have been intensively evaluated on a number of beneficial insect species, such as lacewings, ladybirds and bees (reviewed by Li et al., 2014d; Wang et al., 2015). Some bitrophic studies have been conducted for two biological control agents, green lacewing Chrysoperla sinica and ladybird Propylea japonica, which were fed Bt or non‐Bt control rice pollen (Bai et al., 2005a,b; Li et al., 2014c, 2015; Wang et al., 2012; Zhang et al., 2014). Results from the laboratory feeding studies indicated that adults of C. sinica and P. japonica are not affected by Cry2A‐, Cry1C‐ and Cry1Ab‐expressing rice pollen. In addition, multiple laboratory studies using tritrophic systems were conducted for assessing the potential effects of Bt rice on insect predators and parasitoids (Bai et al., 2006a; Chen et al., 2007, 2009; Jiang et al., 2004b, 2005; Li et al., 2013; Liu et al., 2011a; Tian et al., 2012). In general, no significant negative effects were found on predators such as P. japonica, Cyrtorhinus lividipennis, Ummeliata insecticeps, Pardosa pseudoannulata and C. sinica in prey‐mediated studies, where nontarget insects were used as prey (Bai et al., 2006a; Chen et al., 2007; Li et al., 2013; Tian et al., 2012). However, when target pests were used as prey/host in such studies, negative effects were normally detected on predators such as Pirata subpiraticus and C. sinica (Chen et al., 2009; Li et al., 2013; Liu et al., 2011a) or parasitoids such as Apanteles chilonis (Jiang et al., 2004b, 2005). These data demonstrate that the detrimental effects observed in the tritrophic studies using target pests as prey items can be attributed to decreased prey quality due to the sensitivity of the prey to Cry proteins, rather than direct toxicity of the Cry protein expressed in Bt rice to the predator (Li et al., 2013; Romeis et al., 2006). For example, larval development time of C. sinica larvae was significantly prolonged when fed C. suppressalis larvae that had been reared on an artificial diet containing Cry2A (100 μg/g diet) compared those in control treatment (Li et al., 2013). C. sinica larvae were not negatively affected when directly fed an artificial diet containing Cry2A (200 μg/g diet) (Li et al., 2014c). However, ELISA measurements showed that the mean Cry2A concentration in C. sinica larvae was 0.09 μg/g fresh weight when fed Cry2A‐containing C. suppressalis larvae, but was 0.37 μg/g fresh weight when directly fed a Cry2A‐containing diet (Li et al., 2014c). This outcome demonstrates that the detrimental effects on C. sinica larvae observed were mediated by C. suppressalis larvae as prey, but not attributable to the toxicity of Cry2A to the predator (Li et al., 2014c). More details regarding the prey quality‐mediated effects of Bt proteins on predators can be found in a recent review article by Romeis et al. (2014). Similarly, results of field surveys indicated no significant difference between non‐Bt rice fields and Cry1Ab/Ac‐ Cry1C‐ or Cry2A‐expressing Bt rice fields on population dynamics of predatory arthropods (Akhtar et al., 2013a; Lu et al., 2014a; Xu et al., 2011). However, for parasitoid species, multiple field surveys did not find consistent differences between Bt and non‐Bt rice fields in terms of species richness, diversity, evenness and dominance indices for parasitoid communities (Chen et al., 2003; Li et al., 2007a; Liu et al., 2003, 2006; Tian et al., 2008), although populations of the parasitoid wasps of target insect pests may be adversely affected due to the reduction of host density (Romeis et al., 2006; Tian et al., 2008).

In addition to pest natural enemies, other potential effects of Bt rice on nontarget herbivores also have been tested, focusing on planthoppers, leafhoppers and thrips. Laboratory studies showed that feeding on Bt rice producing Cry1Ac, Cry1Ab, Cry1C, Cry2A and vip3H proteins had no negative effect on survival and development of the planthoppers Nilaparvata lugens and Sogatella fucifera (Chen et al., 2011; Fu et al., 2003; Lu et al., 2014b,c, 2015; Mannakkara et al., 2013; Zhang et al., 2011b). In accord with the results of laboratory studies, field trials did not find consistent differences in the population dynamics of planthoppers and thrips between Bt rice and nonrice fields (Akhtar et al., 2013a,b; Chen et al., 2011, 2012; Jiao et al., 2006; Li et al., 2013; Liu et al., 2007a,b; Romeis et al., 2006;)

The silkworm, Bombyx mori (Lepidoptera: Bombycidae), an economically important insect in China, has the potential to be exposed to Cry protein by feeding on Bt rice pollen‐contaminated mulberry leaves if Cry proteins are expressed in rice pollen. Thus, silk worms may be subject to indirect risk posed by production of stem borer‐resistant Bt rice, especially because silkworms belong to the same order as the stem borers (Romeis et al., 2008, 2013). Therefore, the potential effects of Bt rice on silkworms have raised much concern (Wang et al., 2001, 2002a; Yang et al., 2014; Yao et al., 2006, 2008; Yuan et al., 2006). Previous studies showed significant differences in toxicity of pollen from different Bt rice lines to silkworms (Wang et al., 2001, 2012; Yang et al., 2014; Yao et al., 2006, 2008; Yuan et al., 2006). The differing results may be attributed to (i) different types of Bt proteins expressed in pollen; (ii) different expression levels of Bt proteins contained in pollen; and (iii) different levels at which silkworm larvae were exposed to Bt protein in the respective feeding studies (Yang et al., 2014; Yao et al., 2008). In addition, different silkworm varieties showed significantly different sensitivities to the same Bt protein (Yuan et al., 2006). Although Bt rice pollen from some transgenic rice lines poses toxicity to B. mori larvae under worst‐case conditions, considering that the deposition of rice pollen on mulberry leaves is very limited under field conditions, a general conclusion was made that growing of Bt rice may pose a negligible risk to silkworm (Chen et al., 2011; Fan et al., 2003; Yang et al., 2014; Yao et al., 2008). For example, recent laboratory experiments were conducted to assess the potential effects of Cry1C‐ or Cry2A‐producing transgenic rice (T1C‐19, T2A‐1) pollen on B. mori fitness. B. mori larvae were not negatively affected when fed mulberry leaves covered with pollen from Bt rice lines for 3 days, even when Bt rice pollen density was at 1800 pollen grains/cm² mulberry leaf, which is much higher than the mean natural density of rice pollen on leaves of mulberry trees (93 grains/cm² leaf) near paddy fields (Yang et al., 2014). In the long‐term assay, the larvae were fed Bt and control pollen for their entire larval stage (approximately 27 days), which was significantly longer than the rice pollen shedding duration, B. mori larvae were negatively affected at Bt pollen densities exceeding 150 grains/cm² leaf (Yang et al., 2014). All available data suggest that Bt rice lines probably represent a negligible risk to B. mori larvae because of the limited exposure of the larvae. However, to further guarantee the safety of the silk industry, Yang et al. (2014) suggested that mulberry leaves on trees that are near paddy fields planted with Bt rice lines should not be used to feed B. mori larvae or should be washed before they are fed to B. mori larvae.

Impact on soil ecosystems

One important aspect of environmental risk assessment of Bt plants is the possible accumulation and persistence of plant‐produced Cry proteins in agricultural ecosystems, especially in soils where Bt crops are grown and residues of the crop plants are incorporated by tillage or as litter, because this pathway could affect sensitive nontarget organisms or interfere with biological processes (Giovannetti et al., 2005; Head et al., 2002; Li et al., 2007b). Wang et al. (2007) reported that Cry1Ab protein from Bt rice plants quickly degraded in paddy soils under aerobic conditions, with a half‐life ranging from 19.6 to 41.3 days, while the degradation was significantly prolonged to 45.9–141 days under flooded conditions. In addition, Wang et al. (2007) found that the degradation of Cry1Ab in soil was mostly biotic and not related to any specific soil property. Similar results were found by Li et al. (2007b), reporting that the degradation of Cry1Ac protein from Bt rice was significantly faster in nonsterile water than in sterile water. The results imply that soil microbial organisms may be exposed to Cry proteins for longer periods in flooded Bt rice fields than in a dry Bt cotton or Bt corn field (Chen et al., 2011). Several studies investigated the degradation and persistence of Cry proteins contained in rice plant stubble after harvest. Cry proteins in stalks, roots or leaves quickly degrade at the initial stage after entering soil and then enter a relatively stable phase with low concentrations for a long term (Li et al., 2007b; Wang et al., 2007; Zhang et al., 2011b). Although Bt proteins can be released to and persist in soil when growing or after harvest of Bt rice, soil physical and chemical characteristics may not be changed significantly (Song et al., 2011; Wei et al., 2012; Wu et al., 2003).

Many laboratory and field studies were conducted to assess the potential effects of Bt rice on soil organisms, with focus on bacterial communities, fungal communities and collembolan species. Populations of culturable micro‐organisms in the soils amended with Bt rice straw and non‐Bt rice straw were investigated under laboratory conditions, and the results indicated that Cry1Ab released from Bt rice straws has no detrimental effect on culturable soil micro‐organisms including bacteria, actinomycetes and fungi (Wu et al., 2003). In contrast, Xu et al. (2004) reported that Bt rice straw significantly increased the number of hydrolytic‐fermentative and anaerobic nitrogen‐fixing bacteria and decreased denitrifying and methanogenic bacteria in flooded paddy soil under laboratory conditions. In addition, Ren et al. (2004) observed statistical differences in the numbers of anaerobic micro‐organisms and bacterial community between flooded paddy soils amended with Bt or non‐Bt rice straw during the early stages of incubation, but no significant difference was detected 11 weeks after incubation. However, such effects were not confirmed by recent field trials. For example, a 2‐year field study did not detect any measurable difference on the key microbial processes or microbial community between rhizosphere soil samples from transgenic rice or nontransgenic rice fields during the rice‐growing season (Liu et al., 2008; Wei et al., 2012). Likewise, Song et al. (2012) reported no significant effect of growing cry1Ac/cpti transgenic rice on community composition and abundance of ammonia‐oxidizing bacteria in paddy soil.

As soil‐dwelling detrimental species, such as collembola, play an important role in rice ecosystems, much attention has been paid to collembolan species. A 2‐year field survey found no significant difference for population levels of three collembolan families (Scatopsidae, Sminthuridae and Tomocericdae) between Bt and non‐Bt rice fields (Liu et al., 2003). Population densities of the collembolan species Entomobrya griseoolivata, Bourletiella christianseni, Hypogastrura matura and Isotoma monochaeta did not differ statistically between Bt and non‐Bt rice fields (Bai et al., 2006b). Yan et al. (2009) reported no significant effect on E. griseoolivata by growing Bt rice. Yuan et al. (2011) reported significantly lower reproduction of the collembolan Folsomia candida when fed Bt rice plant tissue than for those fed non‐GM near‐isogenic rice, although the authors claimed that the effects could have been caused by the differences in plant composition related to variety of rice, rather than on Cry proteins produced by the Bt rice lines.

Gene flow

As cultivated rice is primarily self‐pollinating, there is little crop‐to‐crop cross‐pollination among different rice cultivars (Lu and Snow, 2005). Thus, exogenous gene‐flow frequencies between transgenic Bt rice to nontransgenic rice are very low, with the reported maximum average frequency of 0.875%; with the increase of distance between transgenic and nontransgenic rice, the average transgene flow declines dramatically, reaching 0% when the distance was 7 m (Rong et al., 2004, 2005; Rong et al., 2006; Rong et al., 2007; Zhang et al., 2012). Therefore, it is widely accepted that an isolation distance of 100 m is effective for minimizing gene flow from GM to non‐GM rice cultivars in China (Li et al., 2012; Rong et al., 2007; Tang, 2012).

Common wild rice (Oryza rufipogon Griff.) is the putative ancestor of Asian cultivated rice (Oryza sativa L.) and the most important genetic resource for rice improvement in terms of its accessibility from gene transfer through sexual means (Lu and Snow, 2005). Regulatory policy requires characterization of exogenous gene flow from transgenic rice to wild rice prior to its commercialization, especially in China (Chen et al., 2004; Song et al., 2003), a centre of origin for rice. In contrast to crop‐to‐crop gene flow, previous studies suggested that wild ancestor O. rufipogon has high compatibility with O. sativa (Song et al., 2002; Suh et al., 1997). The gene‐flow frequency from cultivated rice cultivars to their wild relatives (crop‐to‐wild species) ranges from 1.21 to 2.19% under field conditions (Chen et al., 2004). While Song et al. (2003) reported that the maximum frequency of gene flow between rice cultivars and O. rufipogon was up to 3%, and the maximum distance over which gene flow was observed was 43.2 m in natural habitats at Hunan Province, China. A large‐scale (1.3–2.4 ha) rice gene‐flow study reported that the highest frequency of transgene flow from transgenic rice plants to O. rufipogon was up to 11.24% in Sanya (Hainan, China, at 1 m) and 18.00% in Guangzhou (Guangdong, China, at 0 m), and the maximum distance of transgene flow was 50 m (0.076%) and 250 m (0.008) at Guangzhou and Sanya, respectively, with the detection limit of 0.01% (Jia et al., 2014). Some short‐term studies showed that once exogenous genes escape from transgenic rice to wild relatives, they can be stably inherited and expressed, suggesting the exogenous genes might persist in wild rice populations (Song et al., 2004; Su et al., 2012). However, a recent 12‐year systemic study suggested that F₁ hybrids of transgenic rice/O. rufipogon gradually disappeared within 3–5 years and that the transgene was not detectable in the mixed population, suggesting the O. rufipogon may possess a strong mechanism of reproductive isolation for self‐protection (Jia et al., 2014). Therefore, Jia et al. (2014) inferred that introgression and persistence of transgenes in common wild rice population are unlikely, at least under the conditions of southern China where O. rufipogon is widely distributed.

In addition, gene flow between cultivated rice and weedy rice (crop‐to‐weed gene flow) also has been detected, with quite low levels in many different studies (Lu and Snow, 2005). For example, Chen et al. (2004) reported that gene‐flow frequency from herbicide‐tolerant GM rice to weedy rice ranged from approximately 0.011–0.046%, which was significantly lower than that (1.21–2.19%) occurring between cultivated rice and common wild rice. Cao et al. (2009) compared the performance of three weedy rice strains, and their F₁ hybrids with two IRGM rice lines (CpTI and Bt+CpTI) were compared under field conditions, and the results indicated enhanced relative performance of the crop–weed hybrids, with taller plants, more tillers, panicles and spikelets per plant, as well as higher 1000‐seed weight. Subsequent studies suggested that the introgression of insecticidal genes into co‐occurring weedy rice populations may significantly increase the fitness of weedy rice relative to the weedy parents under high target pest pressure, but not under low pest pressure (Yang et al., 2011, 2012). The authors concluded that insect‐resistant transgenes pose limited fitness advantages to hybrid progeny resulting from crop–weed transgene flow owing to the significantly reduced ambient target insect pressure when an IRGM crop is widely grown.

With current results, it can be concluded that there is gene flow between rice cultivars and wild or weedy rice, while the gene‐flow patterns and the gene‐flow frequencies differ significantly among different rice types and climatic conditions. The ecological risk of transgene flow from GM rice to wild and weedy rice populations seems to be limited and can be controlled by flowering time isolation and spatial isolation (Jia et al., 2014).

Food safety

Rice is one of the main staple foods for over half of the Chinese people; hence, the safety of Bt rice as food has drawn much concern from the public. In China, food safety assessments for GM crops follow the principle of substantial equivalence, as suggested by the Organization for Economic Cooperation and Development and World Health Organization. In the case of the Bt rice lines, Huahui 1 and Bt Shanyou 63, comprehensive assessments were conducted before they received biosafety certificates for commercial production in China (Xiao et al., 2012). The main issues for food safety assessments were thermal stability, digestibility, toxicity and allergenicity of Cry proteins produced by Bt rice, as well as comparisons of nutrient composition and antinutritional factors between Bt and non‐Bt control rice products. In addition, the safety of the whole Bt rice products as food was assessed (He et al., 2008; Xiao et al., 2012; Zheng et al., 2013). The results suggested that both Bt rice products are as safe as non‐Bt control rice products as food (Fan et al., 2014; Xiao et al., 2012; Zheng et al., 2013).

In recent years, especially after the two Bt rice lines were issued biosafety certificates, Chinese people became concerned about the potential chronic effects of GM foods. To address such issues, long‐term rat feeding studies were conducted. A 106‐week chronic test using Sprague–Dawley rats was conducted to evaluate the potential toxicity and carcinogenicity of Huahui 1 rice (Fan et al., 2014). No statistical difference was detected between Bt and non‐Bt treatments for multiple endpoints, including clinical symptoms, body weight, food utilization efficiency, haematological and biochemical index, organ weight and index, gross anatomy and histopathological observations (Fan et al., 2014).

Adoption by farmers

Whether farmers choose to adopt a new technology in crop production depends on whether the technology can bring benefits to them. From 2002 to 2004, Huang et al. (2005) conducted a 3‐year survey to evaluate the potential effects of planting Bt rice on insecticide use, rice yield and farmers' income (Huang et al., 2005). Results of the survey of 330 households from 17 villages in eight counties in Hubei and Fujian provinces showed that compared to non‐Bt rice, planting of Bt rice resulted in the reduction of insecticide use by over 60% and significantly increased rice yield. Meanwhile, incidences of insecticide poisoning were significantly decreased. Results of a 2‐year field study with Bt Shanyou 63 in Wuhan (Wang et al., 2010) suggested that planting of Bt rice can reduce pesticide spraying by 50–60% compared to non‐Bt rice, and that Bt rice could increase rice yield by 60–65% compared with non‐Bt rice when no insecticide was applied. A recent study of farmers' physical examination was conducted to estimate the invisible health impact of pesticide reduction through the adoption of Bt rice on farmers' health, and the results suggested that Bt rice significantly reduced the pesticide use and, as a consequence, improved the health of farmers in China (Huang et al., 2015). Results of these studies indicated that use of transgenic Bt rice benefits the environment, increases farmers' income and improves their health. Thus, it can be expected that farmers should be willing to adopt Bt rice, much as Bt cotton was quickly adopted since its commercialization in 1997 in China (Li et al., 2014a). For example, in 2010, 463 rice farming households from different rice‐producing regions in Hubei Province were surveyed by professors from Huazhong Agricultural University regarding their attitude towards production of Bt rice (Liu and Liu, 2013). The results showed that although the farmers' cognition of the new technology was limited, most of them would adopt Bt rice due to its potential benefits.

Public acceptance

The Chinese government has invested great financial support into agricultural biotechnology research over the past 20 years. However, the pace of commercialization of GM crops has been very slow, largely due to low public acceptance (Chen et al., 2014). The granting of safety certificates for two GM varieties of rice and one of maize in 2009 caused great concern regarding potential effects on public health, food safety and the environment in different circles, and the anti‐GM voice was unprecedentedly high. In 2010, a survey was conducted with 4239 persons (farmers, workers, soldiers, students, scientists and others) from 30 provinces and cities of China (Qu et al., 2011); the results showed that over 55% of people surveyed believed that GM crops may pose risks to human health and the environment. The results of the survey suggested that Chinese people have very low cognition on GMOs; only 21% of people surveyed know the technology, and almost half of the people know little or cannot understand GMOs at all (Qu et al., 2011). The low cognition of Chinese people on GMOs may be attributed to: (i) the technical complexity and novelty of GM technology and the low scientific literacy of the public; (ii) different opinions on GMOs from different scientists, confusing the public's cognition on GMOs; (iii) some scientists overstating the benefits of GM technology; consequently, the GMO is mistrusted; and (iv) researchers have little interest in participating in public ou which was significantly treach on scientific issues (Liu et al., 2011b). In addition, it may also be attributable to most people having no awareness of pesticide contamination in food due to the overuse and misuse of chemical pesticides.

To change this condition, scientists, in particular mainstream scientists working with GM organisms, should be actively engaged in risk communication regarding GMOs. Meanwhile, the mainstream media also should play important roles in popularizing and communicating GMO science to the public, creating a positive environment for use of the new technology. Furthermore, it is appropriate to enhance the oversight of GMO safety and improve the ability of law enforcement agencies for supervision of the safety of GMOs, avoiding or preventing potential incidents regarding GMO safety, as any safety events related to GMOs will significantly decrease public confidence in the safety of GMOs (Li et al., 2014a).

Conclusion and synthesis: How can current impediments be addressed in support of innovation?

Environmentally friendly production of pest‐protected rice would be a boon to Chinese agriculture, and the Chinese government invested considerable resources in the development of Bt rice lines. With other lines also in development, two such lines, Bt Shanyou 63 and Huahui‐1, received biosafety certificates which subsequently were allowed to lapse, and the lines have not been approved for agricultural production. We reviewed a body of ecological risk assessment research on Bt rice that showed effective control of the target pest and negligible risk to the environment and human health; we conclude that ecological risks for Bt rice are less than those for conventional pesticide‐treated rice production. Studies of thermal stability, digestibility, toxicity, allergenicity, nutritional composition, and antinutrient concentration showed Bt rice products to be as safe as conventional rice products. Noting that Chinese farmers readily adopted use of Bt cotton and virus‐resistant papaya, we predict that farmers would value the prospect of increased yield and decreased use of pesticide and would readily adopt production of Bt rice. However, rigorous biosafety regulation and rapid adoption by farmers would not ensure the acceptance of GM rice by the public. In the face of public opposition to GM rice, the Chinese government has effectively stalled adoption of Bt rice. Against this background, we find that enhancing communication of GM crop science‐related issues to government officials and the general public should receive heightened attention in China – as in other countries.

Misinformation about GM crop‐derived products is widespread, with impacts on both domestic food production and international trade. For example, the GM content of imported grain has become an important international trade issue, with most such grain becoming incorporated into animal feeds. Imported grain is not incorporated into the human food supply because of the non‐science‐based issue of whether GM grain is safe for human consumption. Most Chinese people believe that in the United States – where over 90% of corn and soybean production is of GM lines – humans do not directly consume GM plant products. In fact, Americans do consume GM plant‐derived ingredients, especially in processed foods (Haspell, 2014); after billions of meals have been consumed, no negative effects have been associated with GM ingredients. The misconception on the part of the Chinese public seems to follow from food products in the United States not being labelled as containing GM content. More generally, a key explanation for lack of understanding of biotechnology‐related issues on the part of the Chinese public is how the public gets its information about biotechnology. The nongovernmental organization (NGO) sector in China is limited. Greenpeace East Asia (http://www.greenpeace.org/eastasia/) is active in China, regularly sends representatives to meetings concerning biotechnology and actively opposed approval of Bt rice (Greenpeace East Asia 2012). The biotechnology industry also has a presence in China through the international NGO, CropLife Asia (http://www.croplifeasia.org/). In contrast, in the United States, while the debate over GM crops contains a mix of scientific and nonscientific information, a number of different kinds of organizations are active. Among key players are the probiotechnology Biotechnology Industry Organization BIO (https://www.bio.org/) and the antibiotechnology Center for Food Safety (http://www.centerforfoodsafety.org/#;). Critically, there are independent sources of unbiased scientific information. Leading examples include Information Systems for Biotechnology (http://www.isb.vt.edu/), the Council for Agricultural Science and Technology (http://www.cast-science.org/) and the International Food Information Council (http://www.foodinsight.org/). The outreach activities of many individual scientists, especially at public universities, have notable impact upon the thinking of public officials, media reports and presumably upon public attitudes about agricultural biotechnology. The outcome of all this communication is an active, if not always fact‐based, discussion of issues pertaining to GM crops. Against this background, we encourage full discussion of issues pertaining to GM rice, all GM crops and agricultural biotechnology generally in China. We note that the International Food Information Council (http://www.foodinsight.org/) has recently developed a guide for improving understanding of the role of biotechnology in our food supply in Mandarin Chinese and several other languages, and we hope it finds wide readership in China and elsewhere. Our hope more broadly is to advance political acceptance of selected products found to benefit food security and environmental sustainability in China.