Ensuring plant breeders can take women into account

M9 is the code name of a new banana variety developed for the smallholder banana farmers of Uganda, where banana is a very important staple. Yields can be 60% higher and, because the variety is resistant to pests and diseases, it does not need to be replanted as often. When it came to trials, however, men and women differed in their appraisal of M9. Men rejected it, largely because the bigger bunches failed to get a higher price at market. Women, however, who cook bananas every day for their families, appreciated the new variety’s food value. As a result, M9 was the only one of 18 new varieties to be released, women’s preferences being the deciding factor.

M9 is noteworthy precisely because women’s preferences do not often carry much weight. Plant breeders sometimes assess gender differences in attitudes to new varieties, but not in any great depth. Many would like to have the tools to carry out a more profound analysis, and a new approach from the CGIAR Gender and Breeding Initiative (GBI) offers just that.

The implications of gender relations for modern approaches to crop improvement and plant breeding, by Jacqueline A. Ashby and Vivian Polar, is a chapter in “Gender, Agriculture and Agrarian Transformations,” a new book from Earthscan. In it, Ashby and Polar address a crucial step in the process of plant breeding and offer ways for plant breeders to ensure that the needs of women are fully taken into account.

Gender-based preferences

Polar is a Gender, Monitoring and Evaluation Specialist for the CGIAR Research Program on Roots, Tubers and Bananas (RTB) and GBI. She says that in the past, public-sector breeding, to benefit the poor, paid little attention to gender differences in deciding what traits to incorporate in new varieties. “That,” Polar says, “is changing.”

Driving the change, at least in part, is evidence gathered by a CGIAR research looking at gender-based trait preferences. That investigation found only 39 studies that document the preferences of women and men, or women only, for specific traits and offered reasons for the differences. While there are differences across crops and localities, some commonalities emerge.

In some cases, only women may, for example, value ease of threshing, or the storage life of the produce, or the time needed to cook it. In other cases, only men may talk about pest resistance or yield per hectare. Other traits can be mentioned by women and men alike, but there are still many that represent the preferences of either women or men.

A better measure of successful breeding

Plant breeding goes through several stages at which stringent selection is applied to ensure that only varieties likely to be successful pass to the next stage. Success, in this case, is judged by how closely the candidates match something breeders draw up called the Product Profile, a set of heritable traits that is biologically feasible and that will meet demand from a well-identified set of customers.

“That’s where we need breeders to consider women,” says Polar.

Women’s preferences often clearly reflect the work they do. Ease of threshing may be of little interest to men because when they are not responsible for threshing the harvest. But there are also important differences between different crops and places. In lowland rice production in West Africa, for example, women are responsible for weeding, and unsurprisingly prefer varieties that are better able to suppress weeds. With upland rice in Ghana, men do the weeding, and it is they who value weed suppression. Even this difference is not static, as men migrate for off-farm labor and women take on new tasks.

To cope with these complexities, which their chapter examines in detail, Ashby and Polar suggest a framework that breeders, working with social scientists, apply to each trait in the proposed Product Profile.

Do no harm

“The first objective is to minimize the risk of releasing a variety that could exacerbate existing gender inequities,” explains Ashby, who has worked in CGIAR senior management on the application of social and gender analysis to plant breeding, most recently as Senior Adviser for Gender in Research to the CGIAR System Office and to GBI. This she calls “do no harm” analysis. A second analysis asks whether a trait is uniquely advantageous to women, which could give that trait a higher positive rating. Of course, traits can be gender-neutral, with no general differential between women and men.

“A final product profile may end up with all gender-neutral traits,” says Ashby. “Gender screening should ensure that the profile has not overlooked the possibility of a harmful outcome for women. It also lets a breeding project discuss how important the traits beneficial for women should be in the final set of priorities.”

The sort of analysis proposed could help to avoid some of the problems that have beset recently released varieties. In Ethiopia, for example, women objected to modern short-strawed sorghum varieties because they lost income from the sale of the stalks as cooking fuel, while in East Africa women objected to productive maize varieties that had harder kernels, making grinding — women’s work — more time consuming and more difficult.

Sometimes, counterintuitively, women may prefer varieties that need more processing. In central Malawi, poorer women, unlike men, prefer bitter types of cassava, even though they need more post-harvest processing. Bitter cassava is less likely to be stolen.

“Niche traits,” such as specific qualities required for food processing, can be a real benefit for women by relieving their drudgery or giving them the chance to improve their income. Women in Nigeria, who process small amounts of cassava when they need extra income, prefer varieties that can be left in the ground until needed.

Closing the adoption gap

The framework for gender screening, set out in more detail in the book chapter, has been reviewed by breeders and social scientists from the Gender and Breeding Initiative. The analysis generated an informative discussion of gender implications of individual traits that will now inform breeding work in RTB and beyond.

Vivian Polar says “we know that women producers often have lower adoption rates of modern varieties than men, but plant breeding has had difficulty taking gender differences into account, in part because of the complexity of the picture. Gender screening makes it more likely that breeding will include important traits that are better aligned with women’s needs and priorities.”