1Kawanda Agricultural Research Institute, PO Box 7065, Kampala, Uganda; 2Ecological Phytopathology Group, Agricultural University, PO Box 8025, 6700 EE Wageningen, The Netherlands

Background and objectives
Plastic tunnels have been used in agriculture for tomato and other vegetable production [1] or for raising seedlings under irrigation systems. Irrigation systems and large plastic tunnels are usually expensive to install and are thus not affordable by small-scale subsistence farmers in Uganda and other developing countries. In Uganda, these farmers rely on fungicides for control of late blight but quite often find them expensive, adulterated, expired or lacking efficacy. Research on polythene shelters and sanitation was initiated in 1995, aimed at developing alternative control of late blight in an integrated manner [2]. The use of transparent polythene shelters was investigated as it may provide a physical barrier against late blight (Phytophthora infestans) spores in the aerial environment. The sanitation component of the integrated control was introduced as a means of eradicating any foci of inoculum that could have established in the tunnels.

Materials and methods
The experiment included a study on the timing and duration of sheltering tomato plants with polythene tunnels. The shelters were evaluated against treatments with the fungicide mancozeb, sanitation alone, and control treatments in a randomized complete block experimental design conducted twice. Polytunnels 4 m long, 1.2 m wide and 1.5 m high were constructed to allow rain water to percolate and reach the roots of the tomato plants. Experimental units were plots of 5x4 m with or without three polytunnels. Each experimental unit had four artificially inoculated spreader tomato plants at every corner, 2 m distant. Data on disease development were obtained by assessing the number of healthy and diseased leaves, disease severity on the 5th and 9th leaves on the main branch, and disease incidence at regular intervals during the growing season. The effect of late blight on tomato production was assessed using the number of healthy flowers and healthy fruits and the weight of mature fruits as indicators.

Results and conclusions
The results showed that early sheltering and sanitation were generally as good as the use of mancozeb. Epidemiologically, mancozeb did not stop late blight epidemics but rather delayed them for ca 2 weeks, although it still provided better control than sanitation alone. Sanitation alone had an adverse effect on tomato growth and production, largely because of the number of leaves per plant that were removed during the epidemic as a consequence of high rates of blight development. The research indicated that shelters could help to alleviate the adverse effects of sanitation alone; hence combining the two measures could be used to replace mancozeb for the control of late blight, especially when shelters were used early in the growing season.

1. May D, 1991. American Vegetable Grower 39, 32-38.
2. Chee KH et al., 1988. Research Reports of the Rural Development Administration, Horticulture 30, 31-37.