Insecticide Rates of Application and Public Protection

At the Florida Mosquito Control Association 2002 spring meeting, the potential for "sublethal applications" of an insecticide to increase insecticide resistance in mosquito populations was raised. Some mosquito professionals have advised against sublethal applications in the belief that such applications increase the likelihood of the development of resistance. Exposure to insecticide can only increase resistance in a mosquito population if the insecticide causes 1) the death of susceptible mosquitoes, and 2) if some aspect of the differences between susceptible and resistant mosquitoes is genetic in origin.

Unfortunately, we do not have sufficient information for most mosquito control products to predict the application levels to evaluate the population dynamics of resistance and susceptibility. It is critical to define the degree and type of genetic control of insecticide resistance and susceptibility. The simplest and most often used genetic model is that resistance and susceptibility are controlled by a single genetic locus. One can evaluate several different models of dominance and recessiveness with a single genetic locus. These range from a system where susceptibility is dominant, resistance is dominant, or they are equal or codominant so heterozygous SR genotypes have a phenotype between an SS and RR genotypes (an intermediate resistance or part resistant mosquito).

Let's define resistance and susceptibility so that after a specified dose, we will consider it 100% lethal if all dead mosquitoes are susceptibles, any survivors are resistant. Then if susceptibility is dominant, a 100% lethal dose will certainly kill all exposed SS and SR mosquitoes, and the population becomes instantaneously resistant to our dose, i.e., only RR mosquitoes are left. On the other hand if resistance is dominant, then only the SS mosquitoes die, the population becomes SR and RR. If RR resistant individuals are rare (likely before the insecticide) then most of the R genes will be present in heterozygote SR mosquitoes (purists can look this up in any standard population genetics treatment). The many SR mosquitoes mate with one another, and with the rare RR mosquitoes. The point is that even when one kills all the SS mosquitoes, it takes several generations for RR individuals to increase significantly, and SR with SR matings will produce some SS susceptibles. However resistance will increase. If one killed all the susceptibles, the codominant model will lead to a resistant population at a faster pace than the latter model but more slowly than the first model. Of course if there is extensive immigration into the population, bringing in more susceptibles, the development of resistance will be even slower. Any treatment of managing resistance comes to the same conclusion, keeping more susceptibles around will slow the development of resistance.

At the simplest level, keeping more susceptibles in the population with sublethal doses, reduced application rates, untreated refugia etc. will generally delay the development of resistance. (i.e., see Leeper, Roush and Reynolds 1986. Preventing or managing resistance in arthropods. In: Pesticide Resistance: Strategies and Tactics for Management. National Academy Press, Washington D. C. at

The challenge for mosquito control is to have information on the optimal application rates for every product and the mosquito targets. However, the immediate threat from sublethal doses is that the impact on the target population does not reduce the mosquito population to a level to protect the public. What is needed is information on the genetics of resistance for different species and insecticides, and better predictability on the dynamics of the population genetics. We also require a cost benefit analysis. What is the optimal dose to delay resistance in populations while achieving acceptable mosquito control? Clearly doses exceeding label rates are environmentally unacceptable, and as the highest doses allowable these may serve to promote resistance in most resistance models compared to lower doses. However if susceptibles continue to immigrate into the population, resistance will be slowed at the highest allowable doses. Optimally, the best dose would be the lowest dose that provides the required control while delaying resistance. I expect that there may be situations where this dose might be lower then the labeled rate. It would be nice to have more research on this issue.

Walter J. Tabachnick, Ph.D.
Director, FMEL