Science: Witchweed casts a new curse over the savannah: Martyn Kelly on a parasite that sleeps in the soil as it waits for a host

TRADITIONALLY at the mercy of drought, locusts and a myriad of other scourges, farmers in the African savannah are facing a new threat. It comes from a weed long found in the region but whose prevalence has increased dramatically in recent years. Scientists at the International Institute for Tropical Agriculture (IITA), based in Ibadan, Nigeria, estimate that in parts of Africa it infests more than 40 per cent of fields, causing annual yield losses worth about dollars 7bn.

The weed is called striga, or witchweed, and is unusual in a number of respects, not least in that it is a parasite. Its seedlings grow towards host plants - usually valuable cereal crops - to which they attach themselves and then absorb their nutrients.

In the past, the weed was kept under control by lengthy fallow periods. However, as pressure on the land increases and new, high-yielding varieties of maize supplant native cereals such as sorghum and millet, it has become impossible to continue having fallow periods. Maize - a crop not indigenous to the region - lacks the natural tolerance of these native cereals (or of legumes such as cowpea), acquired as they co-evolved with striga in the African savannah.

Control of the weed is a problem that has taxed agriculturalists. Each plant produces thousands of tiny seeds that can lie dormant in the soil for many years. Whereas most plants' seeds germinate when the environment becomes favourable, and are soon capable of independent growth, striga seeds are programmed to germinate only when they detect a suitable host plant growing nearby. Unless they find a host, the seedlings die within a short time of germinating.

But how does a seed know when a host plant is in the vicinity? Scientists have known for some time that the seeds do not germinate if soaked in water, unless that water has previously been used to grow a host plant such as sorghum. This suggested that the seeds were responding to a chemical produced by the root. It was only in 1986, however, that this chemical was identified as a relative of quinone.

It is not only sorghum and other economically important hosts that stimulate germination: a number of other plants produce compounds that have the same effect. Previously, the number of striga seeds in the soil was kept at a naturally low level by growing these (expendable) plants during the long fallow periods. Now that fallows are no more, alternative approaches need to be sought.

S N C Okonkwo, of the University of Nigeria at Nsukka, has made a long study of the weed's germination, and believes that inducing 'suicidal' germination is one of the most promising methods. However, to date there is little evidence that planting a 'trap' crop, letting the striga germinate and then ploughing it under before it sets seeds, is an effective control. A more promising control technique, in Professor Okonkwo's view, would be to apply artificially synthesised germination stimulants directly to the fields.

A more surprising finding, which particularly excited Professor Okonkwo and the IITA scientists, was the discovery in the early Eighties that the weed's germination is inhibited by certain types of nitrogen fertiliser. Agriculturalists have known for more than 50 years that striga infestation is reduced in the presence of fertiliser, but as it was impossible to grow the parasite away from its host, researchers were unsure exactly why. Was the nitrogen simply strengthening the host plant, or did it act directly to suppress striga?

It was only with the discovery of the germination stimulants that this was resolved. As striga could be grown away from the host plant, Arnold Pieterse, at the Royal Tropical Institute in Amsterdam, was able to demonstrate in the laboratory that urea in particular inhibits not only germination but also the subsequent growth of the seedling towards the host. This is particularly interesting, as most plants show the opposite response: increased growth in response to fertilisers such as nitrogen.

Outwardly, this is an extremely attractive control option: add some nitrogen fertiliser to your crops and it will not only control the weed but increase yield into the bargain. This is fine on paper, but small-scale farmers in sub-Saharan Africa often cannot afford even basic agrochemicals.

This is the crux of the problem. Hugh Doggett, now retired, was an agronomist with the Colonial Service. He does not doubt that anyone with cash can control the weed, and points to examples such as the Gezira irrigation scheme in Sudan where showcase agricultural projects are free from striga while the fields of neighbouring poor farmers are seriously infested. What has yet to be found, he says, is a method of control that is both cheap and effective.

Cereal varieties being bred for resistance by the IITA and others will provide at best a partial solution. Research by IITA plant breeders indicates that tolerant maize varieties only really perform in well- fertilised soils - beyond the means of many farmers.

Maize still needs to be carefully shoehorned into traditional farming practices. IITA scientists are investigating whether changes in farming practices could help to reduce infestation at a lower cost than nitrogen fertilisers. Studies on crop rotations, for example, may give insights into factors that control seed numbers in the soil. 'Know your enemy,' Professor Okonkwo advises in the most recent IITA report, 'before you decide on your strategy.'

(Photograph omitted)