A physiologically-based model predicting the potential establishment and impacts of the fall armyworm (Spodoptera frugiperda) in Europe

The fall armyworm (Spodoptera frugiperda Smith) is an invasive pest originated from the tropical and subtropical regions of south and north America. The species feeds on more than 300 host plants, including economic valuable crops such as corn, rice and sorghum. It represents one of the major threats for agricultural productivity and food security, especially in countries with poor economies and/or lacking of funds to be allocated to the control of the pest. The high migratory capacity of the species and the risks linked to the introduction of infested products are causing concerns related to the introduction of the pest in Europe. Furthermore, transient populations are able to survive in areas outside their establishment range, causing impacts on the crops. In this work, we present the first physiologically-based model for the assessment of the risks of establishment of S. frugiperda in Europe, and for estimating the potential impacts caused by both established and transient populations. The model simulates the stage-specific temperature-dependent physiological responses and accounts for the potential role of density-dependent factors on species mortality and fertility. Parameters related to the density-dependent control factors were estimated through a calibration procedure on 65 locations in central and north America, covering areas where the species is established and areas characterised the seasonal presence of migratory populations. From the calibration procedure we have estimated the uncertainty distributions of model parameters. Combinations of the 10th, 50th, and 90th quantiles of model’s parameters were used for generating best, worst and medium scenarios of potential establishment in Europe. The same combination of model’s parameters were also used for simulating the impacts of the introduction of migratory population at different times of the favourable seasons (1st of April, 1st of June and 1st of August) in five locations (Agrigento, Terracina and Ghedi in Italy, Ouarville in France and Engelsberg in Germany). The results of the model shows that the coastal areas of southern Europe are at risk of establishment of S. frugiperda. In the worst-case scenario, the species’ potential establishment is observed also in limited inland areas of southern Spain and southern Italy. Furthermore, model’s outputs showed that the species was able to generate transient populations outside the area of establishment when the introduction occurred at the 1st of June and at the 1st of July, with population abundance highly dependent on local climatic conditions. Only in Agrigento and in Terracina the species was able to generate transient populations also if the introduction occurred at the 1st of April. The model represents a suitable tool for the identification of the areas at major risks of establishment of S. frugiperda in Europe. The outputs presented might be used for guiding surveillance and prevention plans towards the reduction of the risks linked to the establishment of the species. The model can be also used at the local level for investigating the potential impacts of populations, thus allowing planning and implementing monitoring and control actions for the local management of the species.