Feeding behaviour of Helicoverpa armigera larvae on insect-resistant transgenic cotton and non-transgenic cotton

Abstract: Feeding behaviour of Helicoverpa armigera Hübner (Lep.; Noctuidae) larvae on non‐transgenic Bacillus thuringiensis (Bt) cotton (Gossypium hirsutum L.), Zhong 30, and transgenic cowpea trypsin inhibitor (CpTI)‐Bt cotton, SGK 321, and non‐transgenic cotton, Shiyuan 321, was investigated in both choice tests and no‐choice tests. The results of choice tests suggested that neonates have the ability to detect and avoid transgenic cotton. In the choice tests of neonates with both transgenic and non‐transgenic cotton leaves, a significantly greater proportion of larvae and higher consumption were observed on non‐transgenic cotton than on the transgenic Bt or CpTI‐Bt cotton. In the choice tests with leaves of two transgenic cotton lines, the proportion of neonates on leaf discs of the two lines was not significantly different, but there was significantly higher consumption on CpTI‐Bt transgenic cotton than that on Bt transgenic cotton. In addition, significantly more neonates were found away from the leaf discs, lower consumption and higher mortality were achieved in the choice test with two transgenic cotton leaves than in the choice tests containing non‐transgenic cotton leaves. Leaves and buds were examined in choice tests of fourth instars. It appeared that fourth instars were found in equal numbers on transgenic and non‐transgenic cotton, except when larvae were exposed to leaves for 3 h. However, the total consumption on transgenic cotton was lower than that of the non‐transgenic cotton, so fourth instars may still have the capacity to detect transgenic cotton and reduce feeding on it, although they showed no preference on either transgenic or non‐transgenic cotton. More larvae were found off diet in the treatments with leaves than that of buds, and the number of injured leaf discs by per fourth instar was significantly higher than that of buds in choice tests, suggesting that leaf is a less preferred organ for H. armigera larvae, elicited more larval movements. Similarly, in no‐choice tests of fifth instars, significantly fewer feeding time and more moving time occurred on leaf than that of bud, boll and petal. When cotton line was considered, compared with non‐transgenic cotton, significantly lower feeding time and higher resting time occurred on the two transgenic cottons. Overall, H. armigera larvae have the ability to detect the transgenic Bt and CpTI‐Bt cottons or the less preferred organs and selectively feed more on the non‐transgenic cotton or the preferred organs, especially the neonates, which have a high capacity for avoiding transgenic cotton.