Box 6: Enemy-risk effects, between-plant movement, and
insecticide resistance management
Predator-induced between-plant movement by herbivores can disrupt schemes that are intended to delay the evolution of resistance to insecticides. A significant recent change in agricultural pest management has been the introduction of crop plants genetically engineered to produce their own insecticidal proteins, derived from the bacterium Bacillus thuringiensis (“Bt”; Tabashnik et al. 2013). Although Bt-crops can reduce the need for widespread applications of insecticides, planting a crop that constitutively produces an insecticidal toxin is a recipe for rapid evolution of resistance. To reduce this risk, evolutionary biologists working with regulators and seed companies designed and implemented the “high dose, refuge” strategy of resistance management. Assuming a monogenic basis for resistance with susceptible allele S and resistance-conferring allele R, a ‘high dose’ means that both susceptible homozygotes (genotype SS) and heterozygotes (RS) are killed on Bt plants. Only the rare resistant homozygotes (RR) can survive. The ‘refuge’ refers to a planted block of non-Bt plants, which are expected to produce relatively large numbers of SS individuals. The rare RR homozygotes surviving on Bt plants are then expected to mate with one of the abundant SS individuals developing in the refuge, and the offspring (genotype RS) are subsequently killed on the Bt crops, removing R alleles from the population. In this way, the models suggest, resistance can be dramatically delayed (Tabashnik et al. 2013).
A key problem, however, has been farmer compliance with planting the block of non-Bt refuge plants (Carroll et al. 2012; Garcia et al. 2016). In response to this, seed companies have introduced the notion of a “refuge in a bag”: planting seed is sold as a mixture of Bt and non-Bt seed, which generates a field with spatially interspersed Bt and non-Bt plants. This approach is now being adopted on a global scale (Tabashnik et al. 2013; Carrière et al. 2016). But if pests move frequently between plants in response to unsuccessful predator attacks, two problems are introduced (Mallet & Porter 1992; Carroll et al. 2012; Carrière et al. 2016). First, the efficacy of the refuge may be eroded. The refuge in a bag idea relies on the expectation that individual non-Bt plants, surrounded by Bt plants, can still support the development of SS individuals. If, however, SS individuals move between plants, individuals beginning their development on a non-Bt refuge plant may move to a Bt plant and be killed (Head et al. 2014). Second, the efficacy of the high dose may be eroded. RS heterozygotes, which must be killed under the high dose strategy, can survive, favoring a rapid increase in R allele frequency, in either of two ways. First, herbivores may begin their lives on a non-Bt plant, where the highly vulnerable early developmental instars can be passed safely, and then move to Bt plants as later instar larvae, which are often more tolerant of Bt toxins, allowing RS individuals to survive (e.g., Head et al. 2014). Second, young RS individuals who start their feeding on a Bt plant may be exposed to toxins, but if they move to non-Bt plants before they ingest a lethal dose they may survive. Thus, enemy-risk effects of predators that cause increases in herbivore movement, even on the very small spatial scale required to move between adjacent plants, can have major effects on the evolutionary trajectory of pest populations.