Experimental evidence that refuges delay insect adaptation to Bacillus thuringiensis

Theoretical projections suggest that refuges from exposure can delay insect adaptation to environmentally benign insecticides derived from Bacillus thuringiensis, but experimental tests of this approach have been limited. We tested the refuge tactic by selecting two sets of two colonies of diamondback moth (Plutella xylostella) for resistance to B. thuringiensis subsp. aizawai in the laboratory. In each set, one colony was selected with no refuge and the other with a 10% refuge from exposure to B. thuringiensis subsp. aizawai. Bioassays conducted after nine selections were completed show that mortality caused by B. thuringiensis subsp. aizawai was significantly greater in the refuge colonies than in the no-refuge colonies. These results demonstrate that the refuges delayed the evolution of resistance. Relative to a susceptible colony, final resistance ratios were 19 and eight for the two no-refuge colonies compared to six and five for the refuge colonies. The mean realized heritability of resistance to B. thuringiensis subsp. aizawai was 0.046 for colonies without refuges, and -0.002 for colonies with refuges. Selection with B. thuringiensis subsp. aizawai decreased susceptibility to B. thuringiensis toxin Cry1Ab, but not to Cry1C or B, thuringiensis subsp. kurstaki. Although the ultimate test of refuges will occur in the field, the experimental evidence reported here confirms modelling results indicating that refuges can slow the evolution of insect resistance to B. thuringiensis.