| Summary: | Number of plants per experimental unit and number of replications for the efficient and precise assessment of black sigatoka leaf spot disease caused by Mycosphaerella fijiensis in East African Highland bananas were determined. Two representative cultivars were used. Host response to black sigatoka infection was measured by recording the youngest leaf with necrotic spots. The number of plants per experimental unit was determined, using the methods of maximum curvature and comparison of variances, while the number of replications was estimated by Hatheway's method. The optimum experimental plot size was 3 plants (18 m2) for the beer banana cultivar 'Igitsiri', and 30 plants (180 m2) for the cooking banana cultivar 'Igisahira Gisanzwe', using the comparison of variances method. However, the optimum plot size was 15 plants (90 m2) for both cultivars using the method of maximum curvature. The latter statistical method was preferred because of the low precision of the estimates in the former method. Unreplicated trials with plots of 15 plants could be adequate to assess black sigatoka response in East African bananas if uniform disease pressure exists.
|
|---|
