The control of coffee rust in Kenya by fungicides

Abstract

SUMMARYThe results are reported of a series of four factorial field trials carried out at Ruiru, Kenya, to determine the effect on the control of coffee rust (Hemileia vastatrix B. et Br.) of the main variables associated with the application of copper fungicides, i.e. concentration, volume per acre, leaf surface to which applied, time of application, and types of fungicide and application used.Rust control increased with log concentration, there being evidence for a linear relationship. As neither asymptotic nor optimal values occur the optimum for commercial practice must be determined by economics and possible phytotoxic effects.Control also increased with volume per acre, but adjusting the amount of spray residue via concentration was more effective and less wasteful of fungicide than via volume. Upper‐surface application was as effective as lower in one trial at all concentrations, and in another at the higher concentrations. It is considered that liberation of fungicide from upper surface residues by rain, and the presence of thick aggregations of these residues may be of importance for rust control.Differences in rust control following the use of three spraying machines were correlated with the differences in amount of spray residue on the leaves, there being a linear relation to log residue. For the same rust control, a mist‐blower with swirl‐plate atomisation required the greatest expenditure of fungicide per acre and a pneumatic hand‐sprayer, closely approached by a mist‐blower with spinning‐cage atomisation, the least. However, to obtain as heavy residues and as good rust control as with the hand‐sprayer, higher concentrations were necessary. At the same copper content of spray fluid no differences in rust control by Burgundy mixture, cuprous oxide (as ‘Perenox’) and copper oxychloride (as ‘Blitox’) were found. Some observations on the fungicides fermate, cuprous oxide in oil and captan were made. Captan did not increase foliage density. Evidence that this is directly influenced by copper sprays is presented.For reducing the amount (but not the percentage) of rust attack March and April sprays were most effective, January least, and October was intermediate. The earlier sprays reduced the effect of later ones. Reduction of inoculum in January to March did not reduce the July‐August rust peak. The implications are discussed.