Pearl millet injury by five grasshopper species (Orthoptera: Acrididae) in Mali

Spike injury rates of pearl millet, Pennisetum glaucum (L.) R. Br. (nonbristled Souna variety), were compared for five grasshopper species that occur in the Sahel zone of Mali, West Africa. Immature females of the species Kraussaria angulifera (Krauss), Oedaleus senegalensis (Krauss), Hieroglyphus daganensis Krauss, Cataloipus cymbiferus (Krauss), and Kraussella amabile (Krauss) were confined in sleeve cages over spikes. Cages built around millet hills were used to test the influence of millet foliage and the weed Digitaria ciliaris (Retz.) Koel. on spike and foliar injury by K. angulifera. Feeding injury in sleeve cages (square centimeters of millet surface area per day per grasshopper), averaged over all spike stages, was higher for the larger C. cymbiferus (9.5 cm2/d) and K. angulifera (8.0 cm2/d) than for the smaller H. daganensis (3.4 cm2/d), K. amabile (2.5 cm2/d), and O. senegalensis (2.3 cm2/d). Mean injury to millet from all grasshopper species was higher for early milk, early flower, and late flower stages (7.9, 6.5, and 6.4 cm2/d) than for late milk and dough stages (3.3 and 1.9 cm2/d). The mean grain weight/grain surface area ratio was 0.26 +/- 0.01 g/cm2 (mean +/- SE). Dry weights of all grasshopper species were correlated with injury rates for each spike stage except the dough stage. The rate of milk stage spike injury by K. angulifera held in whole-plant cages with foliage and ground vegetation removed did not differ from that in sleeve cages. Spike injury was decreased by 55% when millet foliage was present, which indicated some preference for foliage during grain filling. Defoliation in large cages averaged 56 +/- 4.8%. Although the weed did not affect spike injury and was not readily eaten by the grasshopper, millet defoliation was reduced by 38%, indicating it had some influence on feeding behavior. Artificial (100%) and partial (mean, 47%) grasshopper-induced defoliation caused grain weight reductions of 54 and 25%, respectively, compared with controls. Therefore, both direct spike feeding and defoliation effects should be considered in crop loss assessments and in estimating economic thresholds.