Dominance and fitness costs of insect resistance to genetically modified Bacillus thuringiensis crops.

Evolution of resistance to genetically modified Bacillus thuringiensis (Bt) crops in pest populations is a major threat to the sustainability of the technology. Incidents of field resistance that have led to control problems of Bt crops or significantly reduced susceptibility of individual Bt proteins in pyramided plants have increased dramatically across the world, especially in recent years. Analysis of globally published data showed that 61.5% and 60.0% of the cases of resistance with major alleles that allowed homozygous resistant genotypes to survival on Bt crops were functionally non-recessive and did not involve fitness costs, respectively. Dominance levels (DFLs) measured on Bt plants ranged from -0.02 to 1.56 with a mean (± sem) of 0.35 ± 0.13 for the 13 cases of single-gene resistance to Bt plants that have been evaluated. Among these, all six cases with field control problems were functionally non-recessive with a mean DFL of 0.63 ± 0.24, which was significantly greater than the DFL (0.11 ± 0.07) of the seven cases without field resistance. In addition, index of fitness costs (IFC) of major resistance was calculated for each case based on the fitness of resistant (R'R') and heterozygous (R'S') genotypes on non-Bt plants divided by the fitness of their susceptible (S'S') counterparts. The estimated IFCs for 15 cases of single-gene resistance were similar for R'R' and R'S', and for the cases with and without field resistance; and the values averaged 1.10 ± 0.12 for R'R' and 1.20 ± 0.18 for R'S'. Limited published data suggest that resistance of insects to dual/multiple-gene Bt crops is likely to be more recessive than the related single-gene resistance, but their IFCs are similar. The quantitative analysis of the global data documents that the prevalence of non-recessive resistance has played an essential role in the widespread evolution of resistance to Bt crops, while the lack of fitness costs is apparently not as critical as the non-recessive resistance. The results suggest that planting of 'high dose' traits is an effective method for Bt crop IRM and more comprehensive management strategies that are also effective for functionally non-recessive resistance should be deployed.