A computational model to predict the population dynamics of Spodoptera frugiperda

Abstract

Among lepidopteran insects, the fall armyworm, Spodoptera frugiperda, deserves special attention because of its agricultural importance. Different computational approaches have been proposed to clarify the dynamics of fall armyworm populations, but most of them have not been tested in the field and do not include one of the most important variables that influence insect development: the temperature. In this study, we developed a computational model that is able to represent the spatio-temporal dynamics of fall armyworms in agricultural landscapes composed of Bt and non-Bt areas, allowing the user to define different input variables, such as the crop area, thermal requirements of S. frugiperda, migration rate, rate of larval movement, and insect resistance to transgenic crops. In order to determine the efficiency of the proposed model, we fitted it using a 4-year (2012–2015) FAW monitoring data for an area located in northern Florida, USA. Simulations were run to predict the number of adults in 2016 and examine possible scenarios involving climate change. The model satisfactorily described the main outbreaks of fall armyworms, estimating values for parameters associated with insect dynamics, i.e., resistance-allele frequency (0.15), migration rate (0.48) and rate of larval movement (0.04). A posterior sensitivity analysis indicated that the frequency of the resistance allele most influenced the model, followed by the migration rate. Our simulations indicated that an increase of 1 °C in weekly mean temperatures could almost double the levels of fall armyworm populations, drawing attention to the possible consequences of temperature rises for pest dynamics.