Data on the millet yield saved by the application of Metarhizium were collected at the end of the 1997 trial in Niger. However, the distribution of millet throughout the treatment plots was highly uneven, and the data could not be given a thorough statistical analysis. Nevertheless, the exercise was useful in developing analytical techniques based on multiple linear regression analysis which were applied in Madagacascar and in the 1998 trial.

The 1998 trial in Niger was designed so that each plot contained half millet and half pasture (see section 4.A.3 for experimental details). However, grasshopper incidence was very low in 1998, and again, data may be compromised (De Groote et al., 1999 (4: 99/8).

The costs of the mycopesticide were analysed in the Cotonou production plant. Previously published data indicated a production cost figure of around $20 per ha for a standard dose of 100g/ha (De Groote in LUBILOSA, 1997). Field trials in 1998 gave satisfactory control at 50g/ha, reducing costs to $10/ha. Formulation, shipping and application cost an additional $1.4, $3.3 and $0.7 respectively for aerial application. Ground application further increases application costs to about $7 per ha.

The medico-economic study carried out by V. Houndekoun has been re-analysed and presented at a meeting of the American Agricultural Economics meeting, and will be published. The costs to human health were estimated at $0.15/ha, livestock losses $1.5/ha and destruction of obsolete stocks at $0.10/ha. This gives a total cost of externalities of chemical pesticide treatment of $1.75/ha, a small but significant amount (Houndekon et al., (9 and 4: 99/3)).

Research on willingness to pay was carried out in Mali and Niger in 1997. In Mali, farmers indicated a willingness to pay about 5000 CFA/ha ($7), about 10% of the value of the crop. In Niger, the inhabitants of the treatment zone (sedentarized Fulani) do not have an area measurement unit and cannot therefore express yield except in terms of total per family. This came to an average of CFA 240,000, which also comes out to about 10% of the crop value at $7 per ha (Attignon et al., 1998 (4: 98/3)).

Farmer participatory trials in southern Benin, progressed through an initial topical PRA in Hontoui. Then participatory technology development trials were conducted in 5 villages nearby. Farmers compared dose rates, sprayers and application strategies. The dose rate selected was 20g per ha. ULVA-plus sprayer was selected over the Pompi - although more expensive, the higher work rate was appreciated. The selected application strategy is to treat aggregations of young nymphs. This area of Benin is intensively cropped, and these aggregations are accessible. However, in other areas of coastal West Africa, these aggregations may be in impenetrable bush (Muller et al., 1999 (4: 99/2).

Conclusions were that sales of Green Muscle 609 to farmers' groups would be viable, although the total annual demand has not yet been estimated.

An attempt to develop a cheap 'artisanal' production system was abortive owing to the high degree of contamination. An interesting spin-off to this attempt, however, is the development of robust Beauveria isolates, growing on oil palm waste. This is not being pursued by LUBILOSA, but by a student project at IITA (I. Godonou PhD thesis IN PREP).

Data on the market size for a Zonocerus-specific mycopesticide in Ghana and Côte d'Ivoire have been collected, but not yet analysed; much of the data obtained from government offices is too highly aggregated to be useful.

For Sahelian grasshoppers, the market will be through NGOs and PVs. See Section 7.

For Desert locust, most sales are likely to take place through FAO procurement, and a series of meetings with the FAO EMPRES programme have taken place recently.

Other potential locust markets include Brown locust in south Africa, Red locust in Mozambique, Tree locust in Kenya and Sudan, Elegant grasshopper in Kenya and Moroccan locust in Morocco, Europe and the ‘Stans. Data are being collected on crop areas at risk and the value of their production.

The interest in establishing such markets is two-fold. Firstly, the existence of these markets allows commercial companies to invest in production plant, from which favourable purchasing terms may then be agreed to farmers unable to afford the full cost of production. Secondly, demonstration of feasibility for some markets demonstrates that as the cash economies of currently poor farming systems develops, markets for Metarhizium will emerge (De Groote et al., 1999 (4: 99/7)).

The LUBILOSA gender policy statement has been widely distributed, and is accepted by LUBILOSA's commercial partners.

A student project on gender and pest control was carried out in Benin, and confirmed suspicions that women have less access to pest control technology than men. In particular, they sometimes apply pesticides by carrying a bucket on their heads and using a brush - inevitably, this leads to increased exposure to pesticide (Gerard, 1999 (5); De Groote et al., 1999 (4: 99/5)).

A workshop on 'Gender issues in Crop Protection' was organised by DFPV. A network has been developed, in which LUBILOSA participates.