Coffee Research Foundation, P 0 Box 4, RUIRU, Kenya

Background and objectives
The varieties of coffee (Coffea arabica) grown in Kenya over the last 15 or more years are susceptible to three major diseases: coffee berry disease (Colletotrichum kahawae); coffee leaf rust (Hemileia vastatrix) and bacterial blight of coffee (Pseudomonas syringae pv. garcae). In areas where bacterial blight of coffee (BBC) is endemic, coffee trees are also affected by coffee berry disease (CBD). The latter is characterized by dark, sunken lesions on young expanding berries resulting in berry shedding or mummification while BBC symptoms include dark, water-soaked necrotic lesions on leaves, tips and nodes of vegetative and cropping branches culminating in a die-back. Two to three-week sprays of copper-based formulations are recommended at 7.0 kg/ha for the 50% WP and Bordeaux mixture (25% WP) at 10.0 kg/ha for dual control of BBC and CBD over a period of 8 months in a year. One or two copper sprays in a year give a tonic effect on coffee trees (Griffiths, 1971; Aduayi, 1972). Regular copper sprays resulted in substantial accumulation of copper in coffee foliage and top soil but without adverse effect on the trees (Aduayi, 1971 a; Dickinson and Lepp, 1985). The phytotoxicity of copper was demonstrated using coffee seedlings grown in nutrient solution (Aduayi, 1971 b). However, experimental evidence to relate field application of copper sprays to potential deleterious effects on growth and cropping of coffee trees has not been available. Against this background, cumulative effect of copper sprays on coffee trees were studied over a long period in order to relate efficacy against the two diseases to growth and cropping behaviour.

Materials and methods
Sprays of 50% WP copper formulations, namely, Kocide 101 (cupric hydroxide); Nordox (cuprous oxide); cobox (copper oxychloride) were applied each at the rate of 0.4% and 0.7% and Bordeaux mixture at 0.5% and 1.0% on 10-tree plots replicated 12 times in randomised complete blocks. A spray interval of 2 to 3 weeks was maintained from mid-March to November of each year. The incidence of BBC was recorded by tagging newly infected shoots on three labelled branches per tree. Incidence of CBD was determined by counts of infected berries on 5 to 10 cropping nodes of each of three labelled branches per tree. Tree growth was determined on four random trees per plot by: counts of new shoots on one labelled primary branch; measurement of extension growth of one secondary branch; and measurement of area of 2 fourth pair leaves from each tree. Crop was picked regularly from each plot, weighed and yield expressed as weight of clean coffee per hectare.

Results and conclusions
Data obtained from the same site over a period of eleven years showed that sprays of the specified copper formulations maintained a significant (P=0.05) control of BBC, CBD and resulted in higher yields ranging from 2 to 15 times compared to the unsprayed plots. The growth of sprayed coffee trees was not impaired by the treatments. On the epidemiology of BBC, the work confirmed that repeated use of copper sprays has the potential of selecting for strains of P. s. garcae resistant to copper.

1. Aduayi EA, 1971 a. Kenya Coffee 36,13-15.
2. Aduayi EA, 1971 b. Turrialba 21, 53-57.
3. Aduayi EA, 1972. Soil Science and Plant Analysis 3, 323-328.
4. Dickinson NM, Lepp NW, 1985. Agriculture, Ecosystems and Environment 14,15-23.
5. Griffiths E, 1971. Tropical Science 14,79-89.