t is a hot July day in Sbiate, Safi Province, Morocco. An assortment of scientists, farmers and extension workers are standing in the sun, apparently looking at a modified lawn-mower. The lawn-mower is being towed by a donkey.
      The outsider might be forgiven for wondering what on earth this has to do with science. For the 60 farmers, five extension agent and others clustered around this device, however, it has a great deal to do with science, and with sustainable food production. And it works.
       The field day at Sbiate had been organized by one of ICARDA's most important collaborators, the Institut National de la Recherche Agronomique (INRA), Morocco's agricultural research body. INRA has been interested for a long time in medic species as a way of producing more feed. In 1985 the Ministry of Agriculture started a 15-year project to convert 150,000 ha of grazed fallow to annual legume pastures, mainly medic.  Farmers were positive about medic's fast regrowth after grazing, and reported better animal health and productivity. They also found that bread-wheat yields were better after medic; indeed, ICARDA originally planned to use medic species for rotation with cereals in the driest areas of West Asia and North Africa.
      But seed supply was a serious constraint. It was available, but at a price. As a result, farmer adoption was slow, and in the early 1990s INRA, ICARDA and the Ministry of Agriculture in Morocco started to test the ICARDA sweeper and thresher. Results on-station were good, and it was decided to begin on-farm trials.
       In 1995/96, six farmers were involved; the area of medic established was 29 ha. The trials continued in the following season.

      They were a success. Average pod yield was 1000 kg/ha; given that seed weight is roughly 30% of pod weight, this corresponds to 300 kg/ha. Farmers reported, however, that the sweeper was hard to manage on uneven ground; two people were needed, and rocks, clods and dry vegetation had to be removed.
So farmers suggested: why not pull the sweeper with a donkey? This was demonstrated at the Sbiate field day.Using animal traction required only  minor modifications to the sweeper. But it worked very well, and enabled the sweeper to pick up 500 kg of pods in a day.  This is enough to seed 2 ha of weedy fallow. The thresher proved more limited in its application, as it needs electricity (although it can, in fact, be adapted to run from diesel power). But it did prove very efficient, and there is a role for it--although it might have to be shared between

farmers, or perhaps run by a local cooperative.
       The President of the Rural Community at Sbiate said that he would be willing to help farmers establish 20 ha of medic pasture from which they can produce their own medic seeds. Indeed, he suggested a three-year community-based project. This could be extended to other forages. Given the positive attitude farmers have expressed towards medics, and the way in which the sweeper does seem to overcome the major constraint, we hope that this work will indeed go ahead.


Mazhar Mohamed is a Forage Economist, INRA, Settat, Morocco. Mustapha Bounejmate is Consultant to the Natural Resources Management Program (NRMP) at ICARDA.