Transgenic cotton and sterile insect releases synergize eradication of pink bollworm a century after it invaded the United States

Abstract

Invasive organisms pose a global threat and are exceptionally difficult to eradicate after they become abundant in their new habitats. We report a successful multitactic strategy for combating the pink bollworm (Pectinophora gossypiella), one of the world's most invasive pests. A coordinated program in the southwestern United States and northern Mexico included releases of billions of sterile pink bollworm moths from airplanes and planting of cotton engineered to produce insecticidal proteins from the bacterium Bacillus thuringiensis (Bt). An analysis of computer simulations and 21 y of field data from Arizona demonstrate that the transgenic Bt cotton and sterile insect releases interacted synergistically to reduce the pest's population size. In Arizona, the program started in 2006 and decreased the pest's estimated statewide population size from over 2 billion in 2005 to zero in 2013. Complementary regional efforts eradicated this pest throughout the cotton-growing areas of the continental United States and northern Mexico a century after it had invaded both countries. The removal of this pest saved farmers in the United States $192 million from 2014 to 2019. It also eliminated the environmental and safety hazards associated with insecticide sprays that had previously targeted the pink bollworm and facilitated an 82% reduction in insecticides used against all cotton pests in Arizona. The economic and social benefits achieved demonstrate the advantages of using agricultural biotechnology in concert with classical pest control tactics.

Keywords: Pectinophora gossypiella; eradication; genetically engineered crop; invasive species; sterile insect technique.

Fig. 1.

Management strategies. (A) The refuge strategy is the primary approach adopted worldwide to delay the evolution of pest resistance to Bt crops and was used in Arizona from 1996 to 2005. Refuges of non-Bt cotton planted near Bt cotton produce abundant susceptible moths (blue) to mate with the rare resistant moths (red) emerging from Bt cotton. If the inheritance of resistance to Bt cotton is recessive, as in pink bollworm, the heterozygous offspring from matings between resistant and susceptible moths die when they feed on Bt cotton bolls as larvae (24). (B) Bt cotton and sterile moth releases were used together in Arizona from 2006 to 2014 as part of a multitactic program to eradicate the pink bollworm. Susceptible sterile moths (brown) were released from airplanes to mate with the rare resistant moths emerging from Bt cotton. The few progeny produced by such matings (48) are expected to be heterozygous for resistance and to die when they feed on Bt cotton bolls as larvae.

Fig. 2.

The simulated effects of sterile moth releases, Bt cotton, and both tactics combined on the population dynamics of pink bollworm. (A) The simulations with realistic values based on empirical data for Arizona for all parameters: initial population size (N₀) = 200 million wild moths, proportion of cotton planted to Bt cotton (pBt) = 0.93, population growth rate per generation (R_o) = 1.6, proportion of moths emigrating out of the field from which they emerged (e) = 0.55, and the effective number of sterile moths released per generation (S_eff) = 3 million (SI Appendix, Table S1). (B) Conditions as in A, except R_o = 3.2 and S_eff = 120 million. Eradication is indicated by the lowest value for moths on the y axis (0.000001 × 1 million = 1 moth). The population stops growing when it reaches the carrying capacity of 200 billion moths.

Fig. 3.

The eradication program in Arizona reduced the pink bollworm population density and the costs associated with this pest. The arrows indicate the first year of the eradication program (2006). The log scale on the y axis has breaks to allow for the plotting of zero values. (A) Infestation of non-Bt cotton bolls by pink bollworm larvae (log [percent infested non-Bt cotton bolls]) decreased significantly during the eradication program from 2006 to 2009 (y = −0.70x + 1,407, R² = 0.99, df = 2, P = 0.0036) but not before the eradication program from 1998 to 2005 (y = −0.031x + 64.1, R² = 0.23, df = 6, P = 0.23). (B) Wild male pink bollworm moths trapped in cotton fields (log [males per trap per week]) decreased significantly from 2006 to 2012 (y = −0.88x + 1,757, R² = 0.94, df = 5, P = 0.0003) but not from 1998 to 2005 (y = 0.018x − 34.6, R² = 0.07, df = 6, P = 0.52). (C) The cost of insecticide treatments made against pink bollworm and the yield loss caused by the pest decreased to $0 from a mean of $33 per ha of cotton for 1998 to 2005. Boll and trap data from 1998 to 2009 were reported previously (25).

Fig. 4.

Pink bollworm population growth decreased as the ratio of sterile to wild moths increased. Linear regression of log-transformed data: log [y] = −1.0 × log [x] + 1.4, R² = 0.99, df = 4, P < 0.0001. Each point shows results from one year based on male pink bollworm moths captured in Arizona from 2006 to 2011. The ratio on the x axis is from yearly totals (SI Appendix, Table S2). Population growth is the mean number of males trapped per week for the last 5 wk of the season minus the first 5 wk of the season (SI Appendix, Table S5 and Fig. S7).