Kaata is making our millet plants wilt
Scientists and West African farmers learn together how to manage a devastating weed—and how to integrate agricultural science into a treacherous farming environment.

By Tom van Mourik

Witchweed takes
its toll

There are about 41 species of witchweed on the African continent and parts of Asia. Africa is believed to be their region of origin. Two main species parasitize cereal crops and wild grasses, and a third parasitizes broadleaved plants and crops such as cowpea and tobacco. Yield losses from vast areas of fields planted with cereals and parasitized by witchweed makes it a serious plant pest.

Parasitic, partly

A “hemi-parasitic” plant such as witchweed draws most of its water, nutrients and carbohydrates from the host root, but it can do photosynthesis itself when it turns green after emergence.

However, it will always get half or more of its carbohydrates from its host. It cannot start its life cycle without being attached to a host. A germinated seed will die within three days if it has not attached successfully to a host root.

Posted December 14, 2006: Kaata or Itikira is the word in the Peuhl language for a mysterious plant with beautiful pink flowers that seems to cause wilting of millet plants wherever it appears in northern Mali.

One of the farmers dealing with its impact is Fatoumata Nouh Tamboura. She is one of about 150 farmers in the village of Gaye in northern Mali, a West African mostly drylands nation located north of Senegal. The farmers of Gaye are aware of the plant (Striga hermonthica or witchweed) and its damage, but often do not understand how it interferes with millet plant growth or how it reproduces.

Farmers in this community grow millet, cowpea, roselle (bissap), okra, and a bit of groundnut and sorghum on 1 to 2 hectares of land, spread throughout the village territory. Most experience a high level of witchweed infestation in their millet fields.

Fatoumata is a farmer-trainer in an experimental witchweed management project carried out at the multi-village level (see details in box below). She doesn’t let her illiteracy get in the way of learning and teaching new techniques. She often asks a teacher to make notes in her book for others to read. As one of the older women in her village, she enjoys a position of respect for what she reports.

Fatoumata brought home the insight that witchweed reproduces by seeds, giving the community a new way to understand the pest and reduce its negative impact.

De-mystifying the witchweed mystery

To be able to solve the witchweed problem, you first need to understand its biology and options for control. Botanically identified as Striga hermonthica (Del.) Benth, witchweed is a noxious, hemi-parasitic weed in semi-arid, sub-Sahara Africa. It causes substantial yield loss in main cereal crops such as sorghum, pearl millet, maize and upland rice.

Witchweed parasitizes the roots of its host plants. Through a specialized organ called a haustorium, it takes in nutrients and water from the host root. Due to its excessive transpiration (the stomata of witchweed cannot close), it “pulls” water, nutrients, and assimilates from the host. (See sidebar: “Parasitic, partly.”)

An estimated 26 million hectares of cereals (maize, sorghum and millet) are infested with witchweed in Sub-Saharan Africa leading to an estimated loss in production of about 10.7 million tons. West Africa alone accounts for about 17 million infested hectares and a yield loss of about 6.5 to 7 million tons—a full 3 percent of the pledged food aid for sub-Saharan Africa for 2005-2006, according to the world Food and Agriculture Organization (see sidebar: “Witchweed takes its toll”).

The most severe problems with witchweed—in terms of yield loss as well as persistence of the weed—seem to occur where soils are degraded, agriculture is extensive, fields are continuously cropped with a cereal host, and nutrient inputs (both organic and inorganic) are low. Continuous cropping of a preferred host like millet, sorghum or maize leads to a build-up of the witchweed seed bank. With more seeds in the soil, more of the parasite plants will attach to the host, leading to reduced growth and less grain in a downward cycle for the crops.

It’s heartbreaking to walk in a millet field infested with witchweed and see the crop plants wilting in areas where subsistence farmers already have to deal with seasons of drought, infertile soils and many other food-production constraints.

Integrated responses

There’s no simple solution to the witchweed problem, though research has developed ample methods and techniques for witchweed control. Only dedicated, labor-intensive manual weeding provides complete control of emerging witchweed plants. However, a combination of practices can lead to meaningful reductions in emerged plants and, ultimately, a reduction of long-term seed-bank densities.


Click image to enlarge

African farmers have adopted several control practices. In regions where agricultural management has intensified (in terms of labor, cropping and inputs), great advances were achieved in so-called Integrated Witchweed Management (IWM).

However, these systems are mostly corn/maize-based systems, and the control packages often involved maize hybrids. Particularly in the more extensive, low-input agricultural systems of the Sahel, farmer-acceptable measures are scarce, and technology transfer has thus not made much progress.

Farmers here reject control techniques that aim only at reducing witchweed. Technologies that serve additional goals such as direct-income increase through the introduction of potential cash (trap) crops has more chance for adoption by farmers in the drier Savannah zones.

The witchweed problem in the West African Sahel needs to be solved by interacting with farmers to find out what measures fit their situation and needs. Organic amendments, intercrops of cereals with non-host plants, host resistance, late weeding and rotations are options within reach of most farmers and easy to combine for effective witchweed control.

Where possible, there will be additional gains in suppressing witchweed pressure through increased use of improved varieties, increased use of fertiliser/organic amendments, higher crop density and more labor input per unit area.

Farmers, and how they are learning to impact witchweed

In 2004, the Mopti region of northern Mali was plagued by locusts that destroyed most of the food harvest. The situation worsened in 2005 when the farmers faced intense drought and yet another year of serious crop losses.

These events led to a serious food security (hunger) problem as 83 of 108 communes (towns) were officially declared in a state of food crisis. Malian government and several voluntary aid groups like Catholic Relief Services (CRS) started food distribution and subsidized grain sales. CRS intends to aid in longer-term recovery and increase the resilience of these agro-pastoral communities to produce food across a variety of conditions.

Providing technical assistance to start farmer field schools was staff from the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). This is a non-profit, apolitical, international organization for science-based agricultural development. An experienced technician from the Institute d’Economie Rurale (IER), a national agricultural research institute, helped with the organisation and participatory planning. Teams from CRS, ICRISAT and IER talked with farmers of the Mopti region and learned that witchweed was one of the top three problems limiting agricultural production.

In June 2006, project teams started farmer field schools in six villages with about 150 farmers participating.


Researcher from the IER (Institute
Economie Rurale) discussing the
problem of weeds and timing of weeding
with the farmer-trainers at the time of
establishment of the crops.

Each village chooses five representatives that become farmer trainers. Thirty farmer-trainers from the six villages come together every week for “train the trainer” instruction. Here they learn to set up and work an experimental field, do observations within the experiments and discuss subjects that concern the cultivation of millet and cowpea.

The training focused on Integrated Witchweed Management (IWM) to teach decision-making that weighs the costs and benefits of using multiple witchweed suppressive techniques—including cropping changes—with potential yields of host and non-host crops for witchweed.

This approach is designed to bring about: (1) more yield and net returns from labor, organic and/or mineral soil amendments and seeds; (2) reduction in the number of emerged weeds and the witchweed seed bank; (3) increase in soil organic matter and fertility; and (4) more sustainable cropping systems in general.

Learning tied to crops

Class teaching content was formed around what was happening in the crop fields at that time of the classes. Using botanical extracts (neem tree leaves, chili pepper or tobacco) to manage insect pests on cowpea was the topic during the flowering phase, whereas the underground stages of witchweed are discussed at the weed pest’s first emergence (at the moment of millet tillering).

Experimental fields have two plots. One compares standard farmer practice with the IWM package of strategies. The other looks at the effects of single components of this package.

During preparatory meetings, program staff worked with the farmer-trainers to design the integrated management package. It included:

  • Sowing at the first big rain (as compared to local farmer custom of dry sowing before the rains).
  • Intercropping millet and cowpea in rows and at higher plant densities than is usually done locally.
  • Using organic amendments (at around 2 tons/hectare).
  • Application of low doses of mineral fertiliser (N-P-K, and urea).
  • Millet plant thinning to two plants per planting hole (as compared to allowing five or more plants).
  • Pulling witchweed plants one or two weeks after the first flowers were observed in the integrated practices demo plot.

By harvest time in November, some things had already become clear to the scientists and the farmers.

  1. Witchweed density was lower in the integrated plot, but pulling the flowering witchweed plants manually was still necessary to prevent unacceptably high levels of seed production.
  2. Although cowpea bean production was about the same in the two practices, cowpea fodder production was much higher in the integrated plot.
  3. Millet grain production is much higher in the integrated plot, despite a later sowing date.

From the agronomic point of view, the IWM package worked out well. However, a final economical evaluation together with the farmers will have to show whether the package is also economically viable. In this evaluation, the costs of a practice (fertilizer, manure, labor) will be balanced against the economic yield value of the fodder and grain, calculated by weight. Only then do we know whether the package is also viable economically and socially. Soil organic matter changes will also be noted.

If the IWM package turns out to be profitable compared to the customary farmer practices, the project leaders will ask the farmers to evaluate whether they feel the techniques used are practical and whether the farmers are willing to take the risk of investment of using them next year.


Proud farmer trainers showing their own
version of millet growth and the
witchweed life cycle, which was explained
by a scientist at the train the
trainer class the week before.

Because it’s so hard to predict whether next year’s growing conditions will be similar to 2006, deciding to adopt or adapt new approaches based on a single year’s experience will be a significant challenge.

If it turns out that the IWM package isn’t as profitable as common farmer practice, the discussion will be to decide if particular improved practices would be worth using, adapting or combining.

In this overall training process, the Malian farmers learned about managing weeds and crops, but also gained the human capital skills of setting up and analyzing on-farm experiments. They built upon the strong social capital (adherence to rules, relationships of trust, mutuality of interest) already important for them, and this may help them to gain financial capital if they choose to form a marketing group. The farmers hope to continue their cooperation with CRS, ICRISAT and IER in the season ahead.