1Natural Resources Institute, Chatham Maritime, Kent ME4 4TB, UK; 2University of Agricultural Sciences, Bangalore 560 065, lndia; 3International Rice Research Institute, PO Box 933, Manila, Philippines

Background and objectives
Farmers' knowledge of their local ecosystem is often extensive [1], however their perceptions of plant diseases may differ considerably from those of scientists, especially as disease pathogens are not easily observed [2]. If scientists are to work with farmers in improving management options for plant diseases, they need to understand how farmers perceive these diseases as this will affect their adoption of management practices. In this paper, results from two studies of farmers' perceptions of plant virus diseases are compared: tomato leaf curl virus (TLCV) in India and rice tungro virus in the Philippines. The effects of farmers' perceptions on actual practices and the factors affecting farmers' perceptions are discussed.

Materials and methods
Field work was conducted with tomato farmers in Karnataka, India and with rice farmers in the Philippines. In both studies, focused group interviews were held with farmers, followed by structured questionnaire surveys of randomly selected farmers. A total of 160 tomato farmers and 242 rice farmers were interviewed.

Results and conclusions
Both tomato and rice farmers had heard of the respective diseases and the majority had experienced them in their own fields. Tomato farmers ranked TLCV as the most important pest or disease problem in the summer season, whilst rice farmers ranked tungro as very important, even in areas where actual disease incidence was low. Farmers could describe the main symptoms but there were gaps in their understanding of the causes and mode of spread. Almost all tomato farmers were not aware of the vector of the disease, whitefly (Bemisia tabaci), but associated the disease with climatic factors. In group discussions, farmers said they took little notice of the insect because of its small size. The rice farmers had been subject to more extension messages over the years linking tungro with its main vector, green leafhopper (Nephotettix virescens), and over one-third of farmers were aware that tungro was spread by insects feeding on infected plants. However, many thought that there were other modes of spread in addition to leafhoppers. Farmers in endemic areas were aware of the differing susceptibility of rice varieties to tungro infection, but this led them to believe that tungro was seedborne. Farmers were not aware of the risks of leaving infected plants in the field as a source of disease inoculum.

Use of insecticides was the most common control method in both studies, even though farmers knew that they were not very effective and that both diseases were incurable. Varietal selection was used to a small extent by tomato farmers, as TLCV-resistant varieties had only recently been introduced, but was common practice amongst rice farmers. About one-third of farmers did not take any specific measures once the disease was in the field, as they did not think any would be effective.

The difference in farmers' perceptions of the two diseases was related to a number of factors. These included the nature and risk characteristics of the diseases; the importance of the crops in the whole farming system; information available to the farmers including extension messages and also from other sources such as pesticide and drug companies; feasibility of control measures; and farmers' beliefs about plant and human disease causality. An assessment of these factors assists in understanding farmers' perceptions of diseases, and why they adopt certain management strategies.

1. Thurston HD, 1990. Plant Disease 74, 96-102.
2. Bentley JW, 1992. In Gibson RW, Sweetmore A, eds, Proceedings of a Seminar on Crop Protection for Resource-poor Farmers. Natural Resources Institute/CTA, Chatham, UK/Wageningen, Netherlands, pp. 107-118.