Résumé
Coffee Wilt Disease (CWD, Gibberella xylarioides) is a serious fungal disease that has caused losses of about $1billion in the Democratic Republic of the Congo, Ethiopia, Tanzania and Uganda since it re-emerged in the 1980s and 1990s. From 2000 to 2007, a regional programme, funded by the Common Fund for Commodities (CFC), European Union (EU), Department for International Development (DFID), and national contributions from affected countries, with scientific collaboration by CIRAD, UCL and CABI, studied many aspects of the disease including its distribution, spread, severity, taxonomy and control. Principal findings of the Regional Programme include: i) CWD is widespread in Uganda and Ethiopia, restricted in Tanzania, and spreading in DRC. It is found too in wild forest coffee, giving rise to concern that it may weaken the genetic base of both Robusta and Arabica genomes. Measures are needed to collect and conserve this material; ii) Two fungal strains exist, one infecting Robusta and the other Arabica. The strain of the current Robusta disease outbreak is very similar to a strain isolated from DRC in 1960. The Arabica strain does not infect Robusta coffee and vice versa. The Arabica and Robusta strains are likely to be co-evolved pathogens, evolving with their respective hosts, close to their respective centres of origin. iii) CWD transmission from infected wood to adjacent uninfected seedlings was confirmed; hence leaving infected wood near uninfected trees in the field is an infection pathway. Healthy seedlings become infected by CWD when potted in soil from around infected trees. Soil can be highly infective for at least three months; at least one year without coffee is advisable to avoid re-infection; iv) Many farmers weed by machete and hoe, which may spread CWD through frequent wounds to the base of the stem. Farmers can remove diseased coffee wood from plots and may sell them as fuel - a major route of transmission that must be stopped by increased training and quarantine inspections. The programme therefore intensified training and created awareness to help in stopping this other modes of transmission for CWD; v) No Arabica tested cultivars displayed mortality less than 20%, suggesting that Ethiopian Arabica CWD could present a serious threat to production in other countries if it spread, as has happened with Robusta CWD. On the other hand, a breeding programme in Uganda has screened thousands of Robusta plants for resistance to CWD. The initial screening produced over 1,500 lines potentially resistant to the disease. Further screening and agronomic trials have reduced this to seven final candidates for release to farmers, hopefully in 2010. Screening studies in Tanzania have resulted in identification of six Robusta clones with resistance to CWD. If CWD is to be effectively controlled, a distinct, long term and proactive strategy is needed to suppress it in the future. The paper will cover a number of suggestions as to what needs to happen next. (Résumé d'auteur)