FORUM Large-Scale Management of Insect Resistance to Transgenic Cotton in Arizona: Can Transgenic Insecticidal Crops be Sustained? YVES CARRIE ` RE, 1 TIMOTHY J. DENNEHY, 1 BRENT PEDERSEN, 1 SHIRLEY HALLER, 2 CHRISTA ELLERS-KIRK, 1 LARRY ANTILLA, 2 YONG-BIAO LIU, 1 ELIZABETH WILLOTT, 1 AND BRUCE E. TABASHNIK 1 J. Econ. Entomol. 94(2): 315Ð325 (2001) ABSTRACT A major challenge for agriculture is management of insect resistance to toxins from Bacillus thuringiensis (Bt) produced by transgenic crops. Here we describe how a large-scale program is being developed in Arizona for management of resistance to Bt cotton in the pink bollworm, Pectinophora gossypiella (Saunders) (Lepidoptera: Gelechiidae), and other insect pests of cotton. Financial support from growers makes this program possible. Collaboration between the Arizona Cotton Research and Protection Council, the University of Arizona, and government agencies has led to development of resistance management guidelines, a remedial action plan, and tools for monitoring compliance with the proposed guidelines. Direct participation in development of resistance management policies is a strong incentive for growers to invest in resistance manage- ment research. However, more research, regularly updated regulations, and increased collaboration between stakeholders are urgently needed to maintain efÞcacy of Bt toxins in transgenic crops. KEY WORDS Bacillus thuringiensis, Pectinophora gossypiella, dispersal, geographical information system, resistance management, transgenic cotton CROPS ENGINEERED TO produce toxins from Bacillus thu- ringiensis (Bt) can be a valuable asset for agriculture. To date, Bt crops have been effective in controlling several key pests, thereby eliminating the need for many insecticide treatments. In areas where elimina- tion of those treatments did not cause resurgence of pests, deployment of Bt crops reduced insecticide use (Smith 1998, USDA Economic Research Service 2000). This reduction is beneÞcial because the envi- ronmental effects of most synthetic insecticides are worse than those of Bt proteins. Moreover, reducing insecticide treatments often resulted in increased proÞts for growers, and it could foster greater success of integrated pest management (IPM) programs. The potential for rapid evolution of insect resis- tance to Bt crops threatens the aforementioned ben- eÞts (Tabashnik 1994, Gould 1998, Frutos et al. 1999). Many of the next generation of transgenic varieties engineered for insect control will produce a different Bt toxin in addition to the one currently produced. This pyramiding strategy is expected to have higher potential for delaying evolution of pest resistance, as long as there is no cross-resistance between toxins and the frequency of alleles conferring resistance to each toxin is low in populations of the targeted pests (Roush 1997). If resistance evolves to the currently used Bt toxin, then in addition to reducing the usefulness of this toxin, the pyramiding strategy is also undermined. Hence, evolution of resistance to the current single- toxin plants would not only have immediate negative consequences but also far reaching implications for IPM. Current efforts to manage resistance to Bt crops center on the refuge/high-dose strategy (Gould 1998). The basic assumption is that resistance to Bt is recessive and conferred by a single locus with two alleles, resulting in three genotypes: susceptible ho- mozygotes (SS), heterozygotes (RS), and resistant homozygotes (RR). Under ideal conditions, refuges with host plants that do not produce Bt enable survival of SS pests, whereas only RR individuals survive the high doses delivered by Bt crops. The relatively large numbers of SS adults emerging from refuges mate with rare RR survivors from the Bt crop. Their hybrid prog- eny (RS) are killed by Bt crops, which purges R alleles from pest populations and delays evolution of resis- tance. Simulation models and limited empirical evidence suggest that the effectiveness of the refuge/high-dose strategy depends strongly on the genetic basis of re- sistance, initial frequency of the resistance alleles, level of mortality caused by transgenics, life-history costs associated with resistance, extent of random mat- ing between the resistant and susceptible adults, and refuge size (Tabashnik 1990, Liu and Tabashnik 1997, Roush 1997, Gould 1998, Peck et al. 1999, Shelton et al. 2000; Y.C. and B.E.T., unpublished data). These fac- tors vary among insect species exploiting a particular transgenic crop; also, the spectrum of pests feeding on 1 Department of Entomology, University of Arizona, P.O. Box 210036, Tucson, AZ 85721. 2 Arizona Cotton Research and Protection Council, 3721 E. Wier Avenue, Phoenix, AZ 85040 Ð2933. 0022-0493/01/0315Ð0325$02.00/0  2001 Entomological Society of America