Development of insecticide resistance

Last time we briefly covered some basic theories about insecticide resistance, which is that as long as genetic variation and the selection pressure - insecticide applications are present, insecticide resistance is inevitable. It's a matter of time before resistance will develop, and the process cannot be controlled, although can be delayed by the proper use of the insecticide. Today we are going to look at the development of insecticide resistance in some depth.

Let's start with some simple calculations. Suppose that there are one million bedbugs in a local area, and we are going to treat the area with an insecticide. The number one million is chosen so that we can get some meaningful number of resistant individuals to start with.

First, the following assumptions are made:

initial population is 1 million

the insecticide provides 90% control of the susceptible individuals

0.01% or 100 (1 million x 0.01%) initial resistant individuals

After the first insecticide application

100 resistant individuals survive

10% of susceptible individuals survive, 10%x(1,000,000-100)=99990

population after first treatment=99990+100=100090

percentage of resistant individuals after first treatment=100/100090 = 0.10%

By repeating the same calculation for each insecticide application, table 8.1 was created. Then by changing the percentage of control to 80%, table 8.2 was created.

# of treatments	starting population	# of resistant bedbugs	# of survived susceptible bedbugs	final population	% of resistant bedbugs
1	1000000	100	99990	100090	0.10%
2	100090	100	9999	10099	0.99%
3	10099	100	1000	1100	9.09%
4	1100	100	100	200	50.00%
5	200	100	10	110	90.91%
6	110	100	1	101	99.01%

Table 8.1 - insecticide provides 90% control of susceptible individuals

# of treatments	starting population	# of resistant bedbugs	# of survived susceptible bedbugs	final population	% of resistant bedbugs
1	1000000	100	199980	200080	0.05%
2	200080	100	39996	40096	0.25%
3	40096	100	7999	8099	1.23%
4	8099	100	1600	1700	5.88%
5	1700	100	320	420	23.81%
6	420	100	64	164	60.98%

Table 8.2 - insecticide provides 80% control of susceptible individuals

A few conclusions can be made from these two tables:

After each round of treatment, the number of susceptible bedbugs decreases, while the percentage of resistant bedbugs in the population goes up. Therefore, the more frequently the same insecticide is applied, the quicker resistance will develop. Similarly, the longer residue an insecticide has, the longer the bedbugs are exposed to the insecticide, and the quicker resistance will develop.

By comparing the two tables, we can also conclude that the more susceptible bedbugs get killed, the higher the selection pressure becomes, and the quicker resistance will develop.

Resistance will start to spread only after reproduction starts. Without reproduction, all the susceptible bedbugs get eliminated, only 100 resistant bedbugs survive, and the population is reduced from 1 million to 100, which is actually a great achievement. But once reproduction starts, the population can get back to 1 million within a few months, based on the calculations I demonstrated in my previous post "How fast can bedbugs multiply - Part 1". And that would be a population that cannot be controlled by the same insecticide anymore.

The rate at which resistance develops also depends on how fast the insects reproduce. A faster reproduction rate means that more generations of the insects will be exposed to the insecticide, and therefore will speed up the selection process.

Consequently, an important conclusion can be reached at this point, which is that insecticides should be applied as infrequently as possible. Firstly, this will slow down the selection process, and hence will slow down the development of resistance. Secondly, this will keep some susceptible individuals alive, so that the susceptible trait will be passed onto the offspring. This strategy is commonly practiced in agriculture. A grower would first establish an economic threshold, and insecticides will not be applied as long as the economic loss remains below the threshold. The concept is also a major component of Integrated Pest Management.

Once reproduction starts, another important factor, the frequency of the resistance alleles will be involved. Let's take a look at some background theories first.

An allele is any of a number of alternative forms of a gene. Most insects have two copies of each allele. For illustration purpose, we label resistant allele with R, and susceptible allele with S. So the combinations, which are termed genotypes, are RR, RS, or SS.

Resistance can be dominant, semi-dominant, or recessive. If dominant, then only one copy of resistant allele is required to have full protection against the insecticide, since the R allele masks the S allele. In this case, the genotypes are either RR or RS, but RR has higher degree of resistance than RS has. If recessive, then two copies of the same allele are required to have full protection, and the genotype is SS. Finally if resistance is semi-dominant, the alleles are still RS, but the dominance is incomplete and only provides partial protection, i.e., protection against lower amount of toxin.

During reproduction, each parent contributes one of the two alleles to their offspring. The combinations as well as the chance of passing the resistance trait to the offspring are shown in table 8.3. By comparing the percentages in the two columns, we can see that when resistance is dominant, the chance of passing the trait to the offspring is much greater than when resistance is recessive.

Allele frequency is a measure of how common an allele is in a population. Similarly genotype frequency is a measure of how common a particular genotype is in a population. Assuming that the ratio of the genotypes RR:RS:SS is 1:2:1 (by the way, the phenotype ratio is 3:1 for dominant resistance, and 1:3 for recessive resistance), then the frequency of the R alleles is (2x1+1x2)/8, or 0.5, and the frequencies of the genotypes RR, RS, and SS are 0.25, 0.5, and 0.25, respectively, simply because the ratio is 1:2:1.

The Genotype frequency is generally determined by the allele frequency (subject to some conditions). For instance, if the frequency of the R alleles is 0.5, then the chance of getting a R allele from each parent is 0.5, hence the chance for the offspring to get a RR is 0.5 x 0.5, or 0.25. This is similar to the calculation of probability of getting heads in coin tossing. The chance of getting heads once is 0.5, and of getting twice is 0.5 x 0.5, or 0.25, and so on. By the same token, if the percentage of the individuals in a population that are resistant to each one of two insecticides is 1%, then the percentage of the individuals that are resistant to both insecticides is 1% x 1%, or 0.01%. The genotype frequency can also be calculated using the Hardy-Weinberg model. If the frequencies of the allele R and S are p and q, respectively, then the frequencies of RR, RS, and SS are p², 2pq, and q², respectively.

parents	offspring genotypes	% of resistant offspring
		dominant	recessive
RR x RR	RR, RR, RR, RR	100%	100%
RR x RS	RR, RR, RS, RS	100%	50%
RR x SS	RS, RS, RS, RS	100%	0%
RS x RS	RR, RS, RS, SS	75%	25%
SS x RS	RS, RS, SS, SS	50%	0%
SS x SS	SS, SS, SS, SS	0%	0%

Table 8.3

Basically, since resistance is usually recessive, hence only the RR individuals are resistant to the insecticide. And because the frequency of the genotype RR is determined by the frequency of the R alleles, therefore, the key is to keep the frequency of the R alleles under control. There are a few ways to do this. The first is to minimize the use of insecticides so that we can keep some S alleles in the population, as mentioned above. The second is to intentionally keep a few areas unsprayed to provide refuge to the susceptible individuals. This is called the refuge strategy. The third is to raise susceptible individuals in the lab and then release them to the field periodically. For instance, if we have 100 RR individuals, 200 RS, and 100 SS, then the frequency of the R alleles is (100x2+200x1)/(200+400+200), or 0.5. By introducing another 200 SS individuals into the population, the frequency is lowered to (100x2+200x1)/(200+400+200+400), or 0.33. We can also look at this from a different angle. When a RR individual in the field mates with a SS individual released from a lab, the offspring is RS, and is susceptible to the insecticide. The last two strategies will work only if the resistance is recessive and if resistant individuals and the susceptible ones are nearby so that there is a good chance for them to mate with each other. The strategy does not work well with dominant resistance, since the offspring (RS) will still be resistant, although the degree of resistance is lower than that of RR individuals.

Since a bedbug population is normally confined to one residence, and no new susceptible individuals are introduced into the population, therefore with each round of treatment, the insecticide kills only the susceptible individuals, and effectively increases the frequency of the R alleles and causes resistance to spread (within the population). In addition, bedbugs can have 3 to 4 generations a year, and at the time of writing the average number of treatments required to eliminate them from one household is 3, meaning that multiple generations are often exposed to the same insecticide. The combination of these factors can result in rapid spread and high degree of resistance (several thousand-fold resistance to pyrethroids in some cases, again according to the University of Kentucky report). By the way, I suspect that the average number of treatments required could be higher than 3 and resistance might have already played a role. For instance, if a person had a PCO treat his residence twice, later he found out that the treatments failed and had another PCO treat his place three times. The first PCO wouldn't record this as a failure since the person did not call back, and the second PCO would record it as a success after 3 treatments, instead of 5.

To make things worse, although the strategies mentioned above are very important and have been proved effective in agriculture, it might be difficult to practically use them to fight bedbugs. The reason is simple, we have no tolerance for bedbugs and the only acceptable number is zero, and it's quite unlikely that anyone is willing to keep or accept some susceptible bedbugs in his/her home (just imagine that a PCO comes to your home and releases a few hundred non-resistant bedbugs...). This makes our task even more challenging. However, the good news is that, as I repeated many times in the past, while it's almost impossible to control the reproduction of agricultural pests, it is possible to stop bedbugs' reproduction process by protecting ourselves (for how to protect yourself, see my previous post "How to isolate your bed"). Resistance will start to spread only after reproduction starts.

Next time we'll look at some more practical strategies.