GENET archive


2-Plants: First field system to test Bt resistance management plans developed (1)

----------------------------- GENET-news -----------------------------

TITLE:  Testing Bt refuge strategies in the field
SOURCE: Nature Biotechnology, Volume 18 Number 3
        by Fred Gould
DATE:   March 2000

-------------------- archive: --------------------

Testing Bt refuge strategies in the field

Fred Gould (e-mail: is professorof entomology
North Carolina State University, Raleigh, NC, 27695.

In 1999, an estimated 24% of corn and 5% of cotton grown worldwide 
contained a transgene for an insectical protein derived from the 
bacterium Bacillus thuringiensis (Bt). The prevalence of these so-
called "Bt crops" has raised concerns about the possible widespread 
emergence of Bt resistant pests, which could ruin the utility of Bt 
crops. In 1996, the Environmental Protection Agency (EPA; Washington, 
DC) and industry moved to circumvent this danger by requiring farmers 
to plant a certain percentage of their cotton acreage in non Bt-
producing cultivars. This would provide pests a refuge where they 
could feed on plants lacking toxins, thereby maintaining Bt 
susceptible resistance alleles within the insect population. Since 
that ruling, there have been calls for increases in Bt refuge sizes 
by scientific panels by the EPA2, industry3, and the Union of 
Concerned Scientists4 (Washington, DC), as well as two independent 
groups led by extension entomologists working directly with corn and 
cotton farmers5. However, critics have argued that before going 
forward with such increases, field tests are needed to prove that 
refuge strategies are effective. Now in this issue, Shelton et al.6 
have moved us one step forward in addressing this concern, by 
reporting the results of controlled field tests to evaluate the 
impact of pest refuges on the evolution of insects' resistance to 
transgenic insecticidal crops.

 The premise for pest refuges is based on population genetic theory, 
which predicts toxin-susceptible insects produced in pest refuges 
will mate with toxin-resistant insects that survive on the transgenic 
insecticidal crops, thereby diluting the alleles for resistance and 
prolonging the pest population's susceptibility. The 1996 EPA-
approved plan required that cotton farmers plant 4% of their acreage 
in non-Bt-producing cultivars, and not treat this acreage with any 
insecticides that kill Bt-targeted pests, to ensure their survival in 
the refuge. Farmers who did not want to risk the potential crop 
damage in the 4% untreated refuge were offered the option of planting 
at least 20% non-Bt cotton that could be sprayed. The logic here was 
that conventional pesticides are expected to kill about 80% of the 
pests, therefore the 20% and 4% refuges should preserve an equivalent 
number of surviving Bt-susceptible insects.

As illustrated in Fig. 1, refuges are expected to be very effective 
if the transgenic insecticidal crops produce a "highdose" at least 25-
fold the amount of toxin needed to kill susceptible insects. In fact, 
the EPA science advisory committee concluded that "a refuge/high dose 
approach must be employed within the current understanding of the 

Unfortunately, the only way to directly test the refuge/high dose 
theoryis to use it on a wide scale in some areas and not in others; 
economic and ethical issues, however, make this impossible. Even 
small scale field tests have faced major obstacles because they 
generally involve the release of pest insects with resistance 
alleles. Shelton et al. have circumvented this problem by developing 
Bt-producing broccoli, and using it for field tests of the 
diamondback moth in upstate NY where the moth cannot survive the 

Shelton et al. planted broccoli in plots with various types of 
refuges.They released insects with a known frequency of resistance 
alleles and examined how the refuges impacted both resistance allele 
frequency and pest population density. Unfortunately, Shelton et al. 
needed to release insects with a high frequency of resistance alleles 
(0.12 and 0.80) for practical reasons, even though they realized that 
the refuge/high dose approach relies on initially low resistance 
allele frequencies. This problem made it impossible for Shelton et 
al. to test the refuge/high dose strategy as a whole, but they were 
able to examine some aspects of the refuge component of the strategy.

 One question addressed by Shelton et al. was how far away from the 
Bt plants must the non Bt refuge plants be grown? One extreme is to 
plant a mixture of seeds, with Bt and non Bt plants side by side. A 
criticism of this approach is that larvae which begin feeding on non 
Bt plants might move to, and feed on Bt plants, thereby diminishing 
the number of larvae completely escaping the selective impacts of the 
toxin. Shelton et al.'s field data indicate that this is a legitimate 
concern for the diamondback moth. While this would not be a problem 
for cotton pests, such as the pink bollworm, that don't generally 
move between plants, it could affect other target insects such as the 
tobacco budworm and European corn borer. Fortunately, the current and 
proposed refuge recommendations for these pests do not endorse seed 

Shelton et al. also tested whether insects would evolve resistance 
more rapidly when a refuge was sprayed than when it was unsprayed. In 
the 27 day period between the time that insect releases ended and the 
field measurement of resistance allele frequency was begun, insects 
from 20% sprayed refuge treatment and 20% unsprayed refuge treatment 
showed no difference in changes in allele frequency. (The initial 
frequency was 85%, and first field measure was 76%, based on the 
assumption that only larvae homozygous for resistance survive). The 
second field measurement indicated that the frequency of alleles in 
the sprayed refuge had increased by about 2%, but in the unsprayed 
refuge the frequency had decreased by about 12%. Between the second 
and third measurement the frequencies in both 20% refuge treatments 
had risen 0.5%. The heterogeneity in response to the two refuges over 
time is unexplained, but the overall difference of 10% is certainly 
in line with theoretical expectations.

Shelton et al. also found higher numbers of insects on Bt plants in 
the sprayed refuge treatments which they attribute to lower Bt 
susceptibility of these populations. Unfortunately, differences in 
susceptibility do not really explain the differences in densities 
because in period 1, when the mortality of larvae from the sprayed 
and unsprayed treatments are equivalent, there are already 
differences in the densities. The results from this experiment are 
also hard to interpret because Shelton et al. do not report on the 
efficacy of the spraying.

Although there are obvious limitations to these studies, they do 
confirm, in the field, that the mixed seed approach could be 
problematic and that an unsprayed refuge results in more rapid 
adaptation than an equal size sprayed refuge. Although scientists may 
not be surprised by these results, field studies like this one are 
essential for developing public confidence in resistance management 

James, C. Global Review of Commercialized Transgenic Crops. (ISAAA, 
Ithaca, NY, 1999).
E.P.A. 1998. Scientific Advisory Panel, Subpanel on Bacillus 
thuringiensis (Bt) plant-pesticides and resistance management, 
February 9-10, 1998. Docket No. OPPTS-00231.
ILSI. An evaluation of insect resistance management in field corn: a 
science-based framework for risk assessment and risk management. 
(ILSI International Life Sciences Institute, November 23, 1999).
Rissler, J. and M. Mellon. Now or never: Serious new plans to save a 
natural pest control. (Union of Concerned Scientists, CambridgeMA., 
E.P.A. and U.S.D.A. 1999. E.P.A. and USDA position paper on insect 
resistance management in Bt crops.
Shelton et al., Nat. Biotechnol. 18, 339-342(2000).


|                   GENET                     |
| European NGO Network on Genetic Engineering |
|                                             |
|             Hartmut MEYER (Mr)              |
|          Reinhaeuser Landstr. 51            |
|            D - 37083 Goettingen             |
|                 Germany                     |
|                                             |
| phone: +49-551-7700027                      |
| fax:   +49-551-7701672                      |
| email:                    |