1CABI Bioscience, UK Centre Egham, Bakeham Lane, Egham, Surrey, TW20 9TY, UK; 2Kawanda Agricultural Research Institute, P.O. Box 7065, Kampala, Uganda; 3National Agricultural Research Laboratories (KARI), P.O. Box 14733, Nairobi, Kenya; 4Faculty of Agriculture, Sokoine University of Agriculture, P.O. Box 3005, Morogoro, Tanzania

Background and objectives
Fusarium wilt (Panama disease), caused by Fusarium oxysporum f.sp. cubense (FOC), is a major constraint to the production of banana and plantain in many tropical and sub-tropical parts of the world. In East Africa, where bananas are produced primarily on smallholder farms and constitute an important staple food source, losses due to fusarium wilt are frequently reported. However, the development of effective strategies for managing the disease has been severely constrained by an inability to accurately diagnose and identify forms of the pathogen responsible for disease outbreaks. A number of approaches were developed and employed to investigate the importance of the pathogen within East Africa and to identify and characterise pathogenic populations.

Materials and methods
Molecular and biochemical techniques for the characterisation and diagnosis of strains of F. oxysporum causing fusarium wilt have been developed at CABI Bioscience, Egham [1]. Through the establishment of laboratory facilities and training of counterpart scientists, these technologies have been routinely used in Kenya and Uganda to characterise and monitor local populations of the pathogen. Detailed field surveys were undertaken in these two countries and in Tanzania to provide information on the distribution of the disease and to identify susceptible cultivars. Field trials, involving inoculation of host race differentials, were established to determine the pathogenic nature of strains and, if possible, to identify recognised pathogenic races.

Results and conclusions
Laboratory characterisation of F. oxysporum strains permitted the identification of a number of genetically distinct populations, including several which formed bridging complexes across vegetative compatibility groups (VCG) 0124, 0125, 0128, 01212, 01220 and 01222 [2] [3]. Diagnostic surveys showed that fusarium wilt is present in all of the key banana growing regions in the three countries and that a number of widely grown and economically important cultivars are affected. These include Gros Michel, Bluggoe, Ney Poovan and Pisang Awak [2]. Disease incidence was higher than 80% in some instances. Pathogenicity testing under field conditions successfully identified pathogenic variants of the fungus, including FOC race 1 (Kenya and Uganda) and FOC race 2 (Kenya only) [2]. The outputs of these combined activities are of major importance to the development and adoption of improved strategies for managing fusarium wilt, including quarantine measures and the development/deployment of suitable host resistance, which are appropriate to both local and regional needs.

1. Rutherford MA, Bridge PD, Paterson, RRM, Brayford, D, 1995. European Plant Protection Organisation (EPPO) Bulletin 25: pp.137-142.
2. Rutherford MA, 1998. Final Technical Report. UK DFID (NRI) project EMC X0345.
3. Lodwig EM, Bridge PD, Rutherford MA, Kung'u J, Jeffries P, 1998. Letters in Applied Microbiology (submitted).