Potential roles of selected forage grasses inmanagement of fall armyworm (Spodoptera frugiperda)through companion cropping

Production of maize,Zea maysL. (Poaceae), in sub-Saharan Africa is threatened by a new invasivepest, fall armyworm (FAW),Spodoptera frugiperda(JE Smith) (Lepidoptera: Noctuidae). To mitigatethis threat, push–pull companion cropping, a system originally developed for management of lepi-dopteran stemborers, may be used to control FAW. The original system involved trap crops thatfunctioned as a ‘pull’ component to attract moths away from the main crop. How grass species canbe used as trap crops in a push–pull system to control FAW is a question that remains to be answered,because maize is already a highly preferred host plant. Therefore, we tested oviposition preference ofFAW female moths in no-choice and two-choice experiments and larval performance on six selectedgrasses (Poaceae) to assess their roles as trap crop ‘pull’ plants in the system. In no-choice tests, num-bers of eggs deposited onBrachiaria brizantha(Hochst. ex A. Rich.) R. Webster cv. ‘Piata’, cv.‘Mulato II’, and cv. ‘Xaraes’, and Napier grass (Pennisetum purpureumK. Schumach) cv. ‘SouthAfrica’ were not statistically different from those deposited on maize. In two-choice tests betweengrasses and maize, there were no significant differences in number of eggs laid when the plants wereof the same size. However, in two-choice tests with maize plants half of the size of the grasses, signifi-cantly more eggs were laid onB. brizanthacv. Xaraes andP. purpureumcv. South Africa than onmaize, suggesting that crop phenology could make a difference. Numbers of larvae arrested on grassleaf cuts were considerably lower than those on maize leaf cuts after 48 h. In two-choice tests withmaize, molasses grass (Melinis minutifloraP. Beauv.) was the only grass that was significantly pre-ferred to maize for larval settlement after 24 h. After 48 h in the two-choice test, it was the only grassthat retained larvae, although the larval count was significantly lower than on maize. Our data showthat none of the grasses tested were strongly preferred to maize, but the results indicate plants attrac-tive to FAW adults and larvae that could be utilized in a multiple trap crop approach to target variousstages of the pest. Furthermore, results indicate the importance of planting these companion plantsearlier than maize