Abstract. 1. Onset and spread of the severe outbreaks of African armyworm, Spodoptera exempta (Walk.), over Ethiopa, Kenya, Tanzania and Burundi in late 1984 have been analysed in relation to trap and weather records.2. Although first outbreaks in each country appeared progressively further south, winds on the best estimated dates of moth concentration and egg laying blew from directions precluding moth movements between them.3. Such independently developing outbreaks are called ‘primary’ to distinguish them from any ‘secondary’ outbreaks that may develop from them.4. Parent moths giving rise to all seven areas of primary outbreaks were derived from low‐density populations that had persisted through the dry season, probably near the coast of East Africa.5. Primary outbreaks developed on the south‐eastern sides of individual rainstorms over the highlands, most likely because windborne moths were concentrated there by convergent winds.6. Of the secondary outbreaks within Kenya, Tanzania and Burundi, most were derived from only one area of primary outbreaks in Kenya, which were therefore critical for effective control of spread; the paucity of other secondaries was presumably due to a lack of timely rainstorms or other concentrating mechanisms.

Outbreaks of the African armyworm, Spodoptera exempta (Walker), in East Africa follow a seasonal pattern related to the movement of the inter-tropical convergence zone. The season typically begins in central Tanzania with primary outbreaks that coincide with the arrival of the rains. An analysis of 27 years' light trap records shows a high negative correlation between the total number of moths caught by light traps during the season over the whole country and the number of rain days during November in central Tanzania. Out of the 27 seasons, 22 could have been correctly predicted as 'light' or 'severe', using the November rainfall at one station, Dodoma, in time to give two months' warning before the peak of the season.

No abstract

No abstract