Resistance is futile!
Pathogens and pests evolve resistance to our control measures, from antibiotics and pesticides to crop rotation and pest-resistant crop varieties. Slowing the evolution of resistance is an important practical application of evolutionary biology.
An iconic agricultural example, discussed in my forthcoming book, is the "high-dose/refuge strategy" to slow the resistance of crop-eating insects to the bacterial toxin, Bt, which has been genetically engineered into corn, cotton and other crops. The "high dose" refers to crop Bt levels high enough that only insects with two resistance genes (genotype rr) can survive. Bt-free refuges serve as a source of so many susceptible (ss) insects that any rs mutants that arise will mate with them (producing susceptible ss and rs progeny) rather than with each other (with 25% of their progeny resistant rr).
But rs mutants could arise in the Bt-free refuge, not just in the Bt crop. If, in the refuge, the fitness of rs mutants is as high as that of ss insects (i.e., if there is no cost to Bt resistance), then rs individuals could become common enough that two of them could mate, producing rr progeny that could then devastate the nearby crop. So it would be good if, in the refuge, rs insects had lower fitness than ss insects.
This week's paper shows one way that this goal might be achieved. "Fitness Cost of Resistance to Bt Cotton Linked with Increased Gossypol Content in Pink Bollworm Larvae" was published recently in PLoS One.
Cotton plants have natural chemical defenses against insects, including gossypol. The researchers found that rs and rr mutants tend to absorb more of the gossypol they eat. As expected, gossypol slowed their growth, which would tend to reduce fitness.
The rs insects were more susceptible to gossypol than ss insects, so they would tend to stay rare in a high-gossypol refuge. They also looked at rr insects, although once these arose, the crop is doomed. Those with two copies of the same r gene were more susceptible to gossypol than ss insects were, but one strain with two different r genes was fairly resistant to gossypol.
Effects of gossypol on growth were fairly small -- about 5% for ss to 30% for rr. Would higher gossypol levels be better?
It's important to remember that the purpose of the refuge is to saturate the local mating market with ss insects. Higher gossypol levels would help keep rs mutants rare, but we don't want gossypol high enough to also reduce ss populations. For example, if we reduced ss populations in the refuge by 50%, we would need a refuge twice as large.