Optimizing the timing of management interventions against fall armyworm in African smallholder maize: Modelling the pattern of larval population emergence and development

Since its invasion in late 2016, the fall armyworm has a widespread year-round distribution within Africa where it continues to threaten cereal production, particularly maize. Most recommended control advice emphasises the need for interventions against larvae early after colonization of a new maize crop by adults followed by a later intervention if an infestation persists. The current times for action are approximate action thresholds based on scouting which are difficult to implement as early development stages are cryptic and farmers, especially smallholders, have limited time for crop assessments. To improve the impact of controls, the modelling of early and late instar larval population emergence and development in relation to physiological time from planting was developed to enable times to action to be predicted and conveyed to farmers. The two larval population emergence models were built from field fall armyworm data from maize in Zambia and validated from similar data from multiple maize sites in Kenya. A component was included in the models to allow synchronization of maize emergence with larval development. Physiological time, in degree-days, was estimated using Earth Observation land surface temperature data sets. As precise information on action thresholds is lacking for Africa, recent published data on thresholds based on economic injury levels from Colombia were used as a guide but the models can be updated when new information becomes available for Africa. The practical implementation of the models in Africa is discussed including the outcome of some recent preliminary trials with maize farmers in Kenya.

Automated summary

This study developed early and late instar larval population emergence and development models of fall armyworm, aiming to improve control effectiveness by predicting the timing of actions. Preliminary trials in Kenya showed the potential of these models in practical applications.