Department of Agriculture, University of Reading, Earley Gate, PO Box 236, Reading, Berkshire RG6 6AT, UK

Commercial cultivation of bananas meets the demands of two contrasting market sectors: the 'international' markets of North America, Europe and Japan, and the 'local' markets largely within the countries of origin in South America, tropical Africa, India, South-East Asia and the Pacific. The international trade in bananas amounts to 10 m t/year and is almost exclusively based on a few sweet, dessert clones which are essentially minor variants of one genomic type, the Cavendish sub-group. The total fruit production for local markets is probably six times greater and is from a more genetically diverse group of cultivars which meet specific national or regional preferences.

Producers of dessert fruit for export are currently facing a crisis as the leaf-spot disease caused by Mycosphaerella fijiensis threatens productivity. The disease, also known as black Sigatoka, has spread to all tropical banana-growing regions since its discovery in 1964. Export bananas can be protected by a regime of fungicide application, but this option is not generally adopted by the producers of non-export fruit and in consequence productivity of bananas in some countries has declined. Fusarium oxysporum f. sp. cubense is a major pathogen, but is not a problem for the export producers who use the wilt-immune Cavendish clones, having had to replace the wilt-susceptible clone Gros Michel as the principal export type more than 50 years ago. Fusarium wilt is a cause of losses to non-export production where susceptible clones are grown. Plant-parasitic nematodes are another high-priority problem: Radopholus similis, the most damaging, occurs throughout the tropics.

Results and conclusions
Export producers can maintain the levels of production only by the regular application of nematicides. Leaf spots, wilt and nematodes can be managed by the exploitation of genetic resistance. Resistance to these diseases has been incorporated in male parent breeding lines used in banana improvement programmes and several hybrids are now under evaluation in multi-locational trials, but so far none has been perceived by the banana-exporting companies as being an adequate alternative to Cavendish [1]. Management of the wilt in non-export production can be achieved with resistant cultivars, but in this sector also, their adoption may be slow if the substitute varieties have agronomic or culinary defects. Transgenic approaches may eventually solve the problem of incorporating resistance in popular varieties. Other potential threats to banana production are viruses and bacterial wilt, for which quarantine is at present the only defence strategy.

1. Frisson EA, Horry J-P, De Waele D, 1996. New Frontiers in Resistance Breeding for Nematode, Fusarium and Sigatoka. INIBAP, Montpellier.