Epidemiology of cassava mosaic disease in Coast province, Kenya

Abstract

Studies on the epidemiology of cassava mosaic disease were conducted at the Coast Agricultural Research Station (CARS>, Mtwapa between the period 1976 to 1979. The results of the investigations indicated that the development of the mosaic in the field followed a .linear relationship with time and was greatly influenced by the direction of the prevailing winds. The apparent rates of infection in the field on different cassava varieties were generally low (0.01 to 0.05 per unit per day) depending on the varietal resistance to the causal agent of the disease. Although the incidence of the disease was observed to vary during the crop cycle (10 - 12 months) prominent peaks of the disease incidence occurred during the long and short rains. The whitefly vector was present on cassava all the year round and there were marked fluctuations in population build-up within a crop cycle. Peaks in the whitefly population were observed subsequent to rains and the development of adult whitefly population was found to be highly correlated (r = 0.697) with atmospheric temperature {xiii} and relative humidity. Rainfall was found to have an indirect effect on the population build-up of the vector as cassava produced new flush of leaves on which the whiteflies preferred to feed and rest. There was a very high correlation (r - 0.912) between the whitefly population and the incidence of the cassava mosaic in the field. This explained the coincidence of high population levels of the vector and high incidence of the disease during the period of the long and short rains. Whitefly transmission of cassava mosaic virus in the screen-house indicated that transmission of the virus could result from feeding of a single infective whitefly (11.0 per cent transmission rate) but the number of successful transmissions increased with the increase of the number of the viruliferous whiteflies per plant (15 - 20 insects per plant caused 56.08 per cent transmission). It was also shown that there was linear correlation between cassava mosaic disease on all the varieties tested with dosage response of the virus. This relationship was highly significant (p L 0.01). Out of the five cassava varieties evaluated for their resistance to mosaic in the field and in the screen-house, variety 53~8/34 was found to be highly resistant, 46106/27 moderately resistant, (xiv) 37244E susceptible and Aipin Valenca and N Mex 55 highly susceptible to cassava mosaic disease. The difference in the level of resistance in the varieties appeared to be inherent and quantitative in nature. The method of vector transmission in the screen-house was found to distinctly \ separate the cassava varieties tested into different resistance groups. This method could prove useful to plant breeders for quick screening of cassava material for mosaic resistance. The average crop loss due to cassava mosaic disease in the resistant cassava varieties evaluated was 36.3 per cent. The difference in yields between mosaic-free and mosaic-infected treatments was highly significant (p L 0.01).