Opinion: Genetically modified crops need greater farmer inclusion

The recent U.N. Food System Summit brought renewed attention to the promise of genetically modified crops to address poverty and food insecurity in sub-Saharan Africa. The scientific committee to the summit reported that “a lot can be achieved by building research capacity and reducing institutional barriers” for GM crops and other agricultural technologies. Others went further, advocating for a “rapid adoption” of GM crops in Africa to address hunger, because they have the potential to “achieve the good [of modern agricultural systems] without achieving much of the bad.”

But a sober look at past performance shows that GM crops in sub-Saharan Africa haven’t yet achieved “the good.” GM crops have produced gains for many commercial and wealthy farmers. But for the dominant class of farmers on the continent — smallholders operating fewer than four hectares — gains have been elusive, and when achieved, have not been sustained.

Many claim that a new generation of GM crops tailored to smallholder farmers is more likely to boost food security and reduce poverty. These newer crops — most of which have yet to reach the hands of farmers — incorporate development-oriented traits, such as disease resistance and nutrient enhancement, into staple crops grown by smallholders, such as cowpea, cassava, and banana.

But will the programs developing next-generation GM crops avoid the missteps that befell their predecessors? Our new research suggests they are in danger of repeating the mistakes of the past.

Farming systems

A major reason why GM crops have underperformed is they are ill-suited to smallholder farming systems. One problem here is the lack of attention GM projects give to the supportive institutions needed for them to be widely adopted and perform as expected.

In 1999, an insect-resistant form of cotton, Bt cotton, was introduced to smallholder farmers in the Makhathini Flats in South Africa. Adopters of Bt cotton reported some limited reductions in pesticide use and modest yield increases.

But after an initial surge in adoptions, the supportive institutions, which enabled reliable access to GM seeds, fertilizers, and pesticides, dissolved causing most smallholder farmers to discontinue Bt cotton production.

Newer GM crop projects appear poised to make a similar mistake. One example is Water Efficient Maize for Africa, or WEMA, — a GM maize variety designed to perform well under drought conditions.

WEMA requires a “package” of high levels of fertilizers and increased and precise weeding in order to perform well. In the words of a WEMA maize program officer, “You will not have [the] benefit of that product [WEMA maize] without having that package.”

But without reliable access to credit in order to afford fertilizers and robust extension services, many farmers are unlikely to reap meaningful benefits from WEMA. In other words, those with the means to adopt the seeds and “package” — wealthier farmers — are poised to be the primary beneficiaries.

A related concern occurs when breeding GM crops. Critical decisions are made in the breeding process that essentially determine the types of farmers who are most likely to benefit from a GM crop.

Breeding issues have been a problem in the past, most notably in Burkina Faso — where at its peak in 2014 almost 144,000 households grew Bt cotton. Burkina Faso’s Bt cotton variety had lower cotton lint quality than conventional cotton varieties. Though these cotton quality issues were identified by researchers as early as 2006, Burkinabè cotton companies continued to distribute Bt cotton seeds with poor lint quality to farmers, leading to around $85 million in lost revenue. Finally, in 2016, those losses became too much to bear, and Bt cotton was discontinued.

Similar breeding issues are present in next-generation GM crops. One example is disease-resistant matooke banana, a GM cooking banana currently under development to mitigate the impacts of banana bacterial wilt disease, which causes over $500 million in damages in East and Central Africa annually.

The parent variety — or the crop variety into which the GM traits are inserted — known as M9, has traits that many banana farmers and consumers find unappealing. Farmers describe M9 as being too hard to mash and having a white color, which is undesirable for consumers. As one Ugandan farmer told us, “If it [is] white then it [is] not even matooke.”

Moreover, M9's recommended growing practices include more fertilizer, wider spacing, and more maintenance. Many matooke farmers do not currently employ such practices and may be reluctant to embrace them due to increased cost, time, or labor burdens. As such, a potential barrier to GM matooke adoption — and smallholder farmers reaping any benefits — are the undesirable characteristics of the parent variety.

Improving Outcomes

How can next-generation GM projects produce crops that work better for smallholder farmers?

One immediate change is greater transparency. Most information on GM crop research in Africa is either not available to the public or very hard to find. Opening up GM crop projects to greater public scrutiny can build trust — which has been persistently lacking in debates around GM crop benefits.

Greater transparency can also improve GM crop performance by identifying potential issues more quickly, and holding projects accountable. Greater transparency in Burkina Faso would have pressured leaders to find swift and appropriate solutions to Bt cotton quality issues. Improved transparency in current projects such as disease-resistant matooke in Uganda could incentivize breeders to integrate more popular parent varieties as a means to widen adoption and benefits.

Another area where GM crop programs can immediately improve is in crop evaluations — by testing GM crops under actual farming conditions. Current evaluations test trait efficacy and yield performance under ideal farming conditions. Extending this to include trials in diverse farming contexts could inform efforts to produce GM varieties that are more appropriate for smallholder farmers.

Regulations around GM crop experimentation make farming-context crop evaluations more difficult, but creative solutions exist. One way is to make better use of farmer field trials, where farmers, as opposed to research scientists, grow GM crops to test performance. Though most GM crop projects currently use farmer field trials, they often enroll “model” farmers who do not represent the diversity of farmers or farming contexts. Widening such farmer field trials to include a greater diversity of farmers and farming conditions can give a more realistic view of performance and better inform breeding decisions.

Farmer-context crop evaluations can also help identify the institutional elements needed to support wide adoption and optimal performance. Currently, GM crop projects focus on varietal development, but do not extend into building the supportive institutions necessary for enhanced and more equitably distributed benefits. Programs can strengthen credit institutions, agricultural extension, seed development and multiplication, and market access, among others, in order to better achieve these goals.

Greater farmer inclusion should not stop at farmer field trials. Rather, a diverse group of farmers should be included as partners in all aspects of crop development and experimentation. In some cases, it may be too late to achieve such participation across the project life cycle. But even current projects can include a diverse range of farmers in discussions regarding production constraints and how to best address them. Future endeavors should view farmers as valuable knowledge holders and partners in the co-creation of technologies that work for them.

If GM crop promoters are serious about achieving the goals of improving food security and reducing poverty, the crops they produce must resonate with a diversity of smallholder farmers. The possibility of achieving this elusive “good” must begin with more attention to farming context and a greater commitment to including a diversity of farmers as full participants in all project phases.