The coexistence of conventional, organic and genetically modified (GM) crop systems is important for a number of reasons. This kind of system helps in exploiting market opportunities, upholding different cultural values, protecting biodiversity and coping with varying environmental conditions. But there is no easy solution, or widely accepted model, for putting coexistence into practice. Genetically modified crops can be separated from non-GM crops spatially or temporally, and labelling is increasingly seen as essential to protecting consumer choice. But coexistence strategies are still lacking in much of the developing world. Policymakers need to develop rules that are proportionate, efficient, cost-effective and specific to particular crop and farming systems.
What is coexistence and why does it matter?
'Coexistence' in agriculture refers to conventional, organic and genetically modified (GM or transgenic) crop systems all operating in proximity and with the fewest possible undesirable mutual effects.
There are clear advantages to coexistence — particularly considering the opportunities offered by different crops for coping with different production systems and varying environmental conditions, including climate change.
But successful coexistence is also needed to address widely held concerns about GM crops. Many people argue that this is an issue of economics and values. Farmers, seed developers, traders and food companies want to be able to cater for different niche markets, driven by consumer demand. Freedom of choice is important to large-scale growers — who want to be able to adopt different production systems as they wish — as well as family farmers and indigenous communities, who often choose crop varieties for their religious or cultural values, special flavour or cooking qualities.
Some people are more concerned about the safety of transgenic plants, especially those designed to produce pharmaceutical chemicals or industrial materials. For example, some food crops, such as maize, soybean, canola and rice, have been engineered to produce blood anticoagulants, vaccines for rabies or pig diarrhoea, or proteins for treating cystic fibrosis. Such crops could pose health risks if they get into human or animal food supplies.
Many people are also worried that GM crops could cause environmental harm by creating new or more problematic weeds, affecting pollination mechanisms or causing insect pests to develop resistance to biopesticides.
In addition, the exchange of genes (gene flow) between GM plants and wild relatives could affect biological diversity, although this is also a feature of general plant breeding. If a new variety — GM or not — is widely adopted in agriculture, a decrease in local crop diversity is likely.
This issue is particularly important in 'centres of diversity' — regions especially rich and abundant in varieties of a particular plant or crop. Some of these, also called 'centres of origin', represent the source of modern crops and are important for identifying useful traits to be used in plant breeding programmes.
Many of the world's centres of origin and diversity are found in the developing world. Asia is the centre of origin for rice. The uncontrolled spread of GM rice here would be particularly risky, as there are numerous wild species that cultivated rice can cross-fertilise.
Latin America alone is home to five of the world's 12 centres of origin of main food crops — maize and tomato originated in Mexico, cassava in Brazil and potato and beans in Peru. Successful coexistence strategies here are vital to ensuring the survival of these centres.
Gene flow mechanisms
There are several ways in which modified genes may find their way into non-GM crops or wild relatives.
Pollen drift, for example, occurs when GM pollen carried on the wind or by animals pollinates nearby non-GM crops or wild plants.
But the risk of gene flow through pollination differs across crops. Cross-pollination is affected by plants' reproductive structures, behaviour and geographical distribution. 'Open pollinated' crops such as maize — where one plant pollinates another — are more susceptible than 'self-pollinated' crops, such as soybean or rice. This has implications for centres of diversity, since the presence of parental species makes gene flow more likely to be a problem.
Like pollen, seeds may be carried by wind, water, animals and people and seed dispersion can lead to gene flow between GM and non-GM plants.
Seeds left in a field at the end of one growing season can also result in the appearance of 'volunteer' plants in the next season that 'contaminate' the crop.
Post-harvest, seeds can also be dispersed and intermingled by farm machinery and in storage bins, or get mixed during grain handling or transportation.
Managing the risk of gene flow
Managing gene flow risk is important to GM and non-GM growers alike. On the one hand, organic farmers want to ensure their crop does not contain GM material that might cost them their organic certification — and the price premium that goes with it. On the other hand, farmers growing high-value GM crops such as soybean that produces high oleic acid oil may also need to maintain a given level of purity, and equally cannot afford to be 'contaminated' by non-GM or organic breeds.
At the farm level, technical and management measures based on isolating crops need to be applied. These may include enforcing either spatial isolation measures — designating zones free of GM crops, planting GM and non-GM fields some distance from each other, or creating physical barriers against pollen drift — or temporal ones, such as crop rotations or time-lags between GM and non-GM plantings.
Studies have shown that spatial isolation and pollen barriers are the most effective methods for preventing cross-contamination. [1-3] This is especially true of maize — a staple crop for much of the developing world.
Cross-fertilisation also decreases as field size increases.  For many developing countries — particularly in Africa, where small-scale farms with diverse cropping systems dominate the agricultural sector — this poses particular challenges.
The use of GM food crops to produce pharmaceuticals or industrial chemicals requires more rigorous segregation management. Some scientists suggest that only non-food and non-feed crops be seriously considered for such 'biopharming'.  But discussions elsewhere are more liberal, envisaging production systems in high security, dedicated facilities or in areas away from any food crop. 
To effectively manage gene flow risk, isolation during crop production should be accompanied by segregation in seed production and grain handling. Segregation is a well-established issue in seed production because seed companies often need to guarantee a certain level of purity for buyers. Similarly, organic food systems and non-food grain production — for animal feed or oils for example — have dealt with the need for segregation for many years.
But the arrival of GM crops has increased the industry's focus on the problem. For example, the US-based seed company Pioneer responded to the introduction of GM crops by increasing its isolation standard for seed maize production from 200 metres to 3.2 kilometres. The company also requires a gap of three to four weeks between the planting of a seed crop and other maize fields as potential sources of contamination. Such steps have yet to be widely adopted by companies in developing countries.
Specifications about growing conditions, segregation measures and purity standards are typically defined in the contract between grower and buyer.
But national or regional authorities can also lay down purity standards. For example, the European Union (EU) has proposed a new standard for non-GM seed maize, requiring no more than 0.01 per cent to 0.3 per cent GM seed mixture, down from five per cent.  Maintaining such purity as the acreage of GM maize increases, and with it the amount of GM maize pollen in the environment, presents a challenging step for the European seed industry.
Effective segregation of GM produce during transport and handling is equally important. Distribution systems for grains in most exporting countries, including Argentina, Brazil and China, were originally designed to handle bulk loads of undifferentiated commodities in which several different varieties of the same crop might be mixed together. These systems need adapting to include separate containers, conveyors and handling machinery for GM and non-GM produce.
In Brazil, all soybean consignments arriving at Paranaguá port are tested for GM content. Soybeans destined for specific organic or non-GM markets are then stored in separate grain elevators.
Keeping markets separate
It is widely recognised that keeping GM and non-GM ingredients entirely separate is virtually impossible. So many countries require GM food products to be labelled as such, in order to protect consumers' rights to choose.
To build a solid labelling regime, countries first need to establish appropriate thresholds of GM material that define a GM 'product'. Current standards vary. The EU requires food products containing more than 0.9 per cent GM material to be labelled as such; Australia, Brazil and New Zealand have thresholds of one per cent, while Japan has a threshold of five per cent. Most developing countries have yet to implement labelling regimes, although some, like China and Malaysia, have plans under way.
Thresholds for organic food are often much stricter. Some food companies or organic certification bodies — like the United Kingdom's Soil Association — insist that organic food be completely free of GM material.
Without strict regulation, problems can arise. For example, in 2001, unauthorised GM maize found growing in Mexico contaminated wild varieties (see Mexico confirms GM maize contamination), raising concerns about the genetic purity of maize in its centre of origin.
Such contamination could lead to the recall of hundreds of food products around the world, import bans in key international markets and billions of dollars in losses to the food industry and farmers.
The desire to regulate the spread of GM products has led to the adoption of international rules under the World Trade Organization agreements, the Codex Alimentarius and the Cartagena Protocol on Biosafety. The Cartagena Protocol requires exporters of GM products destined for food or processing to provide importing countries with full documentation concerning the GM plant and any transgenes contained in each shipment.
Brazil: a case in point
Brazil has a large and diverse set of agricultural systems — agribusiness contributes more than 30 per cent of the country's gross domestic product.
GM labelling regulations already exist, and the country is signed up to the Cartagena Protocol. It is a big exporter of both GM and non-GM commodities such as soybean and cotton. And it is also a centre of diversity for crops such as cassava, rice, beans and potato.
Coexistence rules are urgently needed to ensure the survival of natural species as well as both GM and non-GM crop industries.
Some measures to ensure segregation have already been adopted. Buffer zones, where planting of GM crops is prohibited, have been established around all protected areas and national parks.
Some individual crops have also come under regulatory frameworks. For example, the National Technical Biosafety Commission has defined specific rules for gene flow containment in GM maize field experiments, including spatial isolation zones of 400 metres plus pollen drift barriers consisting of ten rows of non-GM maize. Time lags of at least 40 days from the flowering times of adjacent maize fields are also used in some areas. In regions where communities plant local maize varieties near field experiments, spatial and temporal isolation measures must be applied simultaneously.
GM cotton is subject to similarly strict regulations (see Box 1).
A broader initiative to promote coexistence between GM and non-GM crops has
yet to be taken, however. Such an initiative will require considerable planning
and coordination, as well as new infrastructure. But it is unlikely to come
soon. Heated scientific and political debate over the commercial cultivation of
GM maize is draining government resources. Neither is it clear in the country's
biosafety framework which regulatory body should take the lead on establishing a
Box 1: Ensuring the biosafety of Brazilian cotton
Brazil is a big cotton producer, consumer and exporter. Domestic growers are predicted to produce about 1.3 million tons of cotton in 2007, of which 430 million tons will be exported.
The country is also home to many sexually compatible cotton varieties that can form fertile hybrids. Traditional breeding techniques have long been used to improve the cultivated cotton species for better fibre quality and stress resistance.
Brazilian experts agree that protection of this genetic diversity is extremely important.
In 2005, scientists at Embrapa — a public institution linked to the Ministry of Agriculture — held a two-day risk assessment workshop for scientists, regulators and government representatives to consider the implications of releasing transgenic, insect-resistant 'Bt' cotton for conserving native cotton varieties and farmers' locally adapted cultivars.
The workshop led to a plan of action to protect the native cotton germplasm. This involves mapping wild cotton populations, establishing isolation zones around areas of native cotton, collecting and preserving germplasm in seed banks and researching the reproductive biology and phenology of the species. [7-8]
The plan is implemented by state agricultural surveillance services and regulated by the Ministry of Agriculture. Although there is no official report on the project's success to date, a monitoring scheme, based on sampling cultivated and native cottonseeds and analysing them for the presence of transgenes, is under development.
Comprehensive coexistence strategies for GM and non-GM crops are urgently needed across the developing world. These could be informed by developed country experiences.
A number of studies, particularly looking at maize, have been carried out in Europe. In addition to recommending technical measures such as isolating crops and establishing pollen barriers, as discussed above, these studies have also called for insurance and liability schemes to be explored. Such schemes might include monitoring and dispute-resolution boards convened and facilitated by a credible authority, or government funds set up to cover economic losses arising from cross-contamination. [9-10]
It is widely recognized by scientists, regulators and the biotechnology industry at large that coexistence strategies should be science-based. Because countries — particularly those in the developing world — have a variety of farm and field sizes, production systems, cropping patterns and environmental conditions, research will be needed on a case-by-case basis to determine optimal approaches.
Investigations are especially needed on pollen flow — to determine appropriate buffer zones — and testing methods for determining the presence of GM material in grain shipments. Genetic use-restriction technologies that prevent seeds from germinating after harvest need to be explored, and each country will need cost-benefit analyses of GM crops.
Appropriate coexistence measures will then need developing and implementing as close to the farm as possible. This means that regional governments will need to play a major role in putting policy into practice and enforcing regulations.
But they will only succeed if the farmers themselves are fully informed and willing to comply. In Brazil, most soybean, maize and cotton growers belong to cooperatives or farmers' associations that could facilitate dialogue and education, especially if government officials and extension workers lend assistance and leadership from. Similarly, Ethiopia has a strong network of farmer field schools, where farmers are brought together to learn best practice from extension workers, that could provide the basis for a coexistence education programme.
Any measures taken would have to address the rights and interests of all farmers and farming cultures — ensuring equal access to technology, legislation, education and funding to enhance their economic stability.
Perhaps most importantly, all initiatives should start with a consensus-building effort involving all affected stakeholders — farmers, agribusiness, the food industry and government. It is important that all agricultural sectors — GM, non-GM and organic alike — embrace the concept of coexistence and work together to accommodate each other.
Peaceful coexistence between GM, non-GM and organic agriculture may not be simple, but it is possible. The goal should be to implement rules that are proportionate, efficient, cost-effective, crop- and farming system-specific, and effective in conserving biodiversity.
Eliana Fontes is a project leader at Embrapa — the Brazilian Agricultural Research Corporation — in Brasilia, Brazil.
 Devos, Y., Reheul, D. and de Schrijver, A. The co-existence between transgenic and non-transgenic maize in the European Union: a focus on pollen flow and cross fertilization. Environment Biosafety Research 4, 71-87 (2005)
 Messeguer, J., Peñas, J., Bas M. et al. Pollen mediated gene flow in maize in real situations of coexistence. Plant Biotechnology Journal 4, 633-645 (2006)
 Weber, W. E., Bringezu, T., Broer, I. et al. Coexistence between GM and non-GM maize crops – tested in 2004 at the field scale level. Journal of Agronomy & Crop Science 193, 79-92 (2007)
 Klein, E. K., Lavigne, C., Picalt, H. et al. Pollen dispersal of oilseed rape: estimation of the dispersal function and effects of field dimension. Journal of Applied Ecology 43, 141-51 (2006)
 Andow, D., Daniell, H., Gepts, P. et al. A Growing Concern: Protecting
the Food Supply in an Era of Pharmaceutical and Industrial Crops. www.ucsusa.org/food_and_environment/genetic_engineering/
 Ma, J. K-C, Drake, P.M.W, Christou, P. Recombinant pharmaceutical proteins in plants. Nature Reviews Genetics 4, 295-305 (2003)
 Barroso, P.A.V, Freire, E.C., Amaral, J.A.B. et al. Zonas de Exclusão de Algodoeiros Transgênicos para Preservação de Espécies e Gossypium Nativas ou Naturalizadas. Campina Grande: Embrapa Algodão (2005)
 CTNBio Commercial Use of Resistant Insect Cotton. Previous conclusive technical opinion no. 513.2005. http://www.ctnbio.gov.br/index.php/content/view/3663.html (2005)
 Byrne, P.F. and Fromherz, S. Can GM and non-GM crops coexist? Setting a precedent in Boulder County, Colorado, USA. Journal of Food, Agriculture & Environment 1, 258-261 (2003)
 Jank, B., Rath, J., Gaugitsch, H. Co-existence of agricultural production systems. Trends in Biotechnology 24, 198-200 (2006)