Computational biogeographic distribution of the fall armyworm (Spodoptera frugiperda J.E. Smith) moth in eastern Africa

• DOI: 10.1016/j.heliyon.2023.e16144
• Publication Date: 2023-05-16T06:54:35Z
• Authors: Elfatih M. Abdel-...; Emily Kimathi; Bester Tawona Mud...; Henri E.Z. Tonnang; Raphael Mongare; Saliou Niassy; Sevgan Subramanian
• Keywords: Fall armyworm; Entomology

The fall armyworm (FAW), Spodoptera frugiperda J.E. Smith, has caused massive maize losses since its attack on the African continent in 2016, particularly east Africa. In this study, we predicted spatial distribution (established habitat) of FAW five countries viz., Kenya, Tanzania, Rwanda, Uganda, and Ethiopia. We used occurrence observations for three years i.e., 2018, 2019, 2020, maximum entropy (MaxEnt) model, bioclimatic, land surface temperature (LST), solar radiation, wind speed, elevation, landscape structure data (i.e., use cover harvested area) as explanatory variables. variables were inputs into a variable selection experiment to select least correlated ones that then predict establishment, suitability areas (very low - very high suitability). shared socio-economic pathways, SSP2-4.5 SSP5-8.5 2030 2050 effect future climate scenarios establishment. results demonstrated establishment eastern Africa based model strength true performance (area under curve: AUC = 0.87), but not randomly. Moreover, ∼27% is currently at risk Predicted are expected increase ∼29% (using each scenarios) year 2030, ∼38% SSP2-4.5) ∼35% SSP5-8.5) scenarios. LULC, croplands area, together with precipitation bioclimatic provided highest permutation importance determining pest Specifically, study revealed was sensitive isothermality (Bio3) rather than being single value year. preference ranges temperature, precipitation, area observed, implying once exotic critical production regions It recommended studies should thus embed present study's modeling dynamic platform provides near-real-time predictions farm scale.