Pyramiding Breeding by Marker-Assisted Recurrent Selection in Upland Cotton: Selected Effects on Resistance to Helicoverpa armigera

The coincidence rates were more than 96% among the instar-weighted average of bioassaysin the lab, the percentage of resistance to Km in the field and the percentage of plantscontaining Bt gene. So, the performance of resistance to Km in the field can be used torepresent the transgenic Bt gene for selecting the resistance to bollworm. The instar-weighted averages were 30.585, 24.182, 16.615, 10.601, 10.123, 7.440 and 7.215 for theC0, P1, M1, M2, MP1, P2 and MP2 populations, respectively. The variance analysisindicated that the instar-weighted average in C0 was greatly significantly higher thanthat in all other populations, i.e., the performance of resistance to bollworm in C0 washighly significantly lower than all other populations. And the resistance in P1 wasgreatly lower than that of M1, M2, MP1, P2 and MP2, and M1 greatly lower than that of M2,MP1, P2 and MP2. There were no significant differences among M2, MP1, P2 and MP2. Withinthe populations of the first cycle selection, MP1 and M1 were greatly significantlyhigher than P1, and MP1 significantly higher than M1. The populations of the second cycleselection were significantly higher than their initial population M1, but no significantdifference among them. The boll size, seed index, the percent of the first harvest yield,fiber length, strength and elongation of the resistant plants to bollworm were significantlylower than that of sensitive plants to bollworm. And the yield of seed and lint cottonof the resistant plant to bollworm were lower than that of the sensitive to bollworm, butno significant difference between them. The boll numbers per plant, lint percent andmicronaire of the resistant plants to bollworm were significantly higher than that of thesensitive plant to bollworm.