Coppicing experiments on the spread and control of cocoa swollen shoot virus in Ghana

Abstract

A coppicing technique was used to assess the incidence of latent and missed infection around 76 outbreaks of cocoa swollen shoot disease found during routine surveys of cocoa farms in Ghana. When all apparently healthy trees were coppiced within at least 27 m of trees found with symptoms, some failed to regenerate, with overall mortalities of 15% and 19% in progenies of Upper Amazon and West African Amelonado parentage, respectively. Many of the regenerating shoots developed stem and/or leaf symptoms of swollen shoot disease within 3 years of coppicing. There were great differences between sites, with overall incidences of infection of 21% and 16% in the Amazon and Amelonado plantings, respectively. The infections revealed by coppicing were not usually grouped closely around trees originally found with symptoms. Gradients of infection were seldom evident and the overall incidence of infection within 3 m of the initial outbreak areas (17%) was little greater than that beyond (14%). The original inspections of the standing trees detected only c. 12% and 18% of all the infected trees present at the Amazon and Amelonado sites, respectively. These results emphasize the limitations of routine inspections in determining the true incidence of infection. They have an important bearing on the conduct and effectiveness of the eradication measures adopted in Ghana and are discussed in relation to the future policy for limiting spread.