The Nonstrategy Benchmark: What the Models Say About Rotating Chemistries

1. Introduction: the thing everybody recommends

Rotate your chemistries. It appears in every label, every certification syllabus and every trade article, and this journal has assumed it repeatedly without ever examining the comparison.

1.1 Read what a sequence is

Use one product until it stops working, then use another. That is not a strategy, and the modelling literature says so by calling it the benchmark.¹

1.2 What this article argues

That the evidence for rotation is weaker than its universality suggests, that mixtures perform better under conditions the modelling specifies, and that structural pest control sits in the part of the parameter space where that advantage does not hold. Sections 10 to 14 are the case.

2. The four concepts

The vocabulary, which is more precise than trade usage.

There are four fundamental strategy concepts: sequences, rotations, mosaics and mixtures.¹

Sequences involve temporally nonoverlapping exposure with the solo use of insecticide A until control failure due to resistance evolution, and then the solo use of insecticide B.²

Rotations, also referred to as cycling, alternation, or periodic application, involve the periodic pre-planned switching between insecticide formulations over time, such that one insecticide is temporarily replaced with another.⁶

Mosaics involve spatially nonoverlapping exposure due to a negatively correlated use of insecticides in different places, and independent use involves randomly overlapping exposure with no correlation in use.²

Mixtures are the use of two insecticides in a single formulation such that the target insect inevitably encounters both insecticides simultaneously.⁶

2.1 Note that independent use is a category

Two products both in circulation with nobody coordinating which goes where, which is what a building serviced by several contractors over several years actually produces.²

It has a name in this literature and it is not one of the four strategies, because it is not a strategy. That observation is ours.

2.2 The distinction that matters

Rotation separates the two products in time. Mosaic separates them in space. A mixture separates them not at all, which is why an individual insect meets both and has to survive both.

2.3 And one term is used loosely

One source notes that in public health the term combination would more generally refer to the use of two different insecticides within the same household, but not in the same control tool, which is a different arrangement from a mixture.⁶

3. The sequence, and what it is called

Because the label does a lot of work.

A sequence is what happens with no resistance management at all: the product is used until it fails, and then it is replaced.²

3.1 It is the control group

Which is why the literature uses it as the benchmark. Any strategy worth adopting has to beat doing nothing in particular.¹

3.2 And it is what most operators actually do

Not by decision but by default. A product that works continues to be used, and a product that stops working is replaced, which is a sequence whatever the operator believes they are doing.

That observation is ours.

4. The finding

The sentence from §1, in its place.

Modelling of mosaics and rotations tended to show minimal gains on sequences.¹

4.1 Minimal gains, across a body of work

The statement is a summary of a literature rather than a single result, carrying seven citations in the original.¹

4.2 And it is stated about a class of models

Not about field outcomes. No source here reports a trial in which rotation was compared to uncontrolled use in a real population over the time resistance takes to develop.¹

4.3 What it does not say

That rotation is useless or harmful. Minimal gains are gains, and a strategy that buys a little time at no cost is worth having.

What it undermines is the confidence with which rotation is presented, which is our reading.

5. Three explanations and none of them the principle

A 2024 paper takes up exactly this question.

Three explanations have been offered for how rotations delay resistance evolution: counterselection from resistance cost, the relaxation of selection and intergenerational redundant kill. We show that all three explanations can make sense of the comparison of rotations with another resistance-management strategy but have failed to elucidate the principle at work.⁴

5.1 That is a remarkable admission

A practice recommended universally for decades, with three competing accounts of its mechanism, none of which the reviewing authors consider to have identified the principle.⁴

5.2 And the framing it sits in

The same paper notes that the principles of resistance management have previously been reduced to moderation, saturation and multiple attack, and that alongside mixtures and mosaics, rotations have been presented as a strategy of multiple attack.³

So all three of the alternatives to a sequence are versions of the same idea, which is why their differences from each other turn out to be small.

5.3 The paper is not hostile to rotation

It presents itself as a perspective on the value of rotations and goes on to offer its own account of how they work.³ It is a clarification rather than a demolition.

6. What the three explanations are

Briefly, because this journal has used all three.

Counterselection from resistance cost. While product A is out of use, the allele conferring resistance to it declines because it carries a fitness penalty. Our reversion article examined exactly this and found five mechanisms that prevent the decline in practice.⁴

Relaxation of selection. While A is out of use, the resistant genotype simply stops gaining ground, which requires no fitness cost at all.⁴

Intergenerational redundant kill. An individual surviving A in one generation meets B in the next, so the two chemistries kill overlapping sets of survivors across time.⁴

6.1 They are not the same claim

The first requires a fitness cost and predicts reversion. The second requires nothing and predicts only that things get no worse. The third requires overlap between generations and predicts an effect that scales with rotation frequency.

Which means they make different predictions and could be distinguished, and that distinguishing them is apparently what has not been done. That is our reading of the paper's complaint.

6.2 And the first one is the shakiest

Our reversion article's central finding was that resistance frequently does not decline when selection stops, for reasons including cross-resistance, modifier genes, gene duplication and recessive costs.

If counterselection is the mechanism, rotation's benefit is contingent on a fitness cost that often is not there.

7. The mixture result

What the modelling recommends instead.

A quantitative genetics model of polygenic resistance in a malaria vector concludes that mixtures are likely to be the most robust because no replacement decisions are required, supporting the conclusion that full-dose mixtures are overall the 'best' IRM strategy.⁵

7.1 The hypothesis was stated in advance

Based on the results from monogenic models, we would hypothesize the difference between the rotation and sequence strategies to be small, and for the mixture strategy to perform best.⁵

Which is good practice and which also tells you that the small rotation-sequence difference was already the expected result before this model was run.

7.2 Note the reason given for robustness

Not an evolutionary property but an operational one: no replacement decisions are required.⁵

A strategy that cannot be implemented wrongly has an advantage over one that depends on somebody deciding correctly and on time, and §20 finds the same argument running the other way.

8. And the cross-resistance finding

The result that surprised us most.

The benefit of full-dose mixtures versus rotations or sequences increased with increasing cross resistance.⁷

And the benefit of full-dose mixtures decreased with negative cross resistance, which is expected as negative cross resistance maintains lower resistance levels for all strategies.⁷

8.1 Cross-resistance is defined by what survives both

A mechanism conferring resistance to one compound that also confers it to another, which our articles on metabolic resistance describe as the common case when the mechanism is a broad-spectrum detoxification enzyme rather than a target-site change.

8.2 The first half is the surprise

Cross-resistance is the thing that makes multiple chemistries less independent, which intuitively should damage any multiple-attack strategy. Instead it widens the mixture's margin.⁷

9. Which is counterintuitive

And worth trying to understand.

Our reading is that cross-resistance hurts rotation and sequence more than it hurts a mixture, because those strategies depend on the second product remaining effective after the first has failed.

9.1 Cross-resistance destroys that assumption

A population selected to resist A is already partly resistant to B when B is introduced, so the time bought by switching shrinks.

A mixture never relied on that in the first place: both products act on the same generation, and the cross-resistant individual has to survive both simultaneously rather than in succession.

9.2 We offer that as an interpretation

The source states the result without explaining it,⁷ and the reasoning above is ours.

9.3 Which matters here

Because our articles on layered resistance and on cross-resistance found that structural pest populations carry broad-spectrum detoxification mechanisms, which is the high cross-resistance condition.

10. The condition

The part of the mixture result that qualifies everything above.

At lower exposure and higher effectiveness, mixtures can perform relatively better than sequences at delaying resistance, which previous studies have concluded is when mixtures work.¹

And the reverse: sequences can evolve resistance more slowly than mixtures with higher exposure and lower effectiveness.¹

10.1 So the recommendation has a domain

Two variables, exposure and effectiveness, and the answer flips depending on where a setting sits in that space.¹

10.2 Which is rarely how the advice is given

Resistance management guidance is generally stated as a rule rather than as a rule with a domain of validity, and that observation is ours.

11. The limiting case

The clearest statement of the failure mode.

As exposure approaches 100% or effectiveness approaches zero, mixtures look increasingly like independent use, and so the contrast is between resistance to the two identical insecticides evolving simultaneously in a mixture or one after the other in a sequence (which would make a mixture take half as long as a sequence).¹

11.1 Half as long

Not a marginal loss. At the limit, the strategy recommended as best takes half the time to fail as the strategy called the nonstrategy benchmark.¹

11.2 It is a limiting case and not a prediction

Nothing is at 100 per cent exposure with zero effectiveness. The value of the limit is that it shows the direction the comparison moves as a setting approaches those conditions.

12. Why the mixture halves

The mechanism, which is simple once stated.

At full exposure with a weak product, every individual meets both chemistries and neither kills efficiently, so selection acts on resistance to both at the same time.¹

12.1 Two clocks running together instead of one after the other

A sequence runs the first clock, then the second. A mixture runs both at once. If neither product is contributing the redundant killing that makes a mixture work, running them together simply spends both at the same rate.

12.2 And this is where redundant kill enters

The paper's title distinguishes itself from that concept, arguing the effect of exposure has been neglected in a literature that framed mixture performance intuitively in terms of redundant killing.¹

Which is a claim that the field had the right answer for an incomplete reason, and our reading is that the exposure term is what supplies the missing half.

12.3 Which is the whole basis of mixture theory reversing

The reason a mixture should work is that an individual resistant to A is killed by B and vice versa, so resistance to either alone confers no survival. That requires each component to kill well on its own. Remove that and the mechanism is gone.¹

13. Which half of the space is structural pest control in

The argument this article exists to make.

Our reversion article established that structural pest control indoors has every condition that slows reversion, including the absence of untreated refuges. A building is a closed system with little adjoining untreated habitat.

13.1 Exposure means proportion of the population reached

Not dose. The variable in the modelling is what fraction of individuals encounter the insecticide at all, and its opposite is a refuge: individuals that breed without ever meeting the selection.¹

Which is why agricultural resistance management devotes so much attention to refuges, and why an indoor treatment has so little to work with.

13.2 Which is high exposure

If the treatment covers the harbourages and the population lives entirely inside the treated space, the proportion of individuals encountering the chemistry is high.

13.3 And effectiveness is frequently low

Which is why a second chemistry is being considered at all. A contractor reaching for a mixture or a rotation is usually doing so because the first product is performing poorly.

13.4 Both variables point the same way

High exposure and low effectiveness are jointly the condition under which a mixture loses to a sequence.¹

So the vector control conclusion that full-dose mixtures are the best strategy may not transfer to a treated building, and that is our argument.

14. Stating that carefully

Because it is a strong claim from indirect evidence.

We have no structural pest modelling. Every source here concerns mosquitoes or an agricultural moth.⁵⁸

The exposure claim is an inference. Nobody has measured what proportion of a cockroach or bed bug population contacts a residual deposit, and our detection articles suggest the figure would be hard to obtain.

And the limiting case is a limit. Section 11.2.

14.1 What we are claiming

Not that mixtures are wrong for structural work. That the published recommendation carries a condition, that the condition is stated in terms of two variables, and that a treated building plausibly sits on the unfavourable side of both.

Which is a reason to want the measurement rather than a reason to act as though we have it.

15. The ranking is not stable

The honest caveat, from a simulation study that examined it.

Including multiple cycles, insecticide decay and cross resistance resulted in micro-mosaics performing second best (behind full-dose mixtures), while including just insecticide decay and cross resistance, omitting multiple cycles, resulted in micro-mosaics becoming the worst performing.⁷

15.1 Best but one to worst

On the basis of one modelling choice.⁷

The same study reports micro-mosaics moving from performing worse than rotations under a single cycle, to equally well with multiple cycles, to better once natural mortality is included.⁷

15.2 Three features, and each changes the answer

Multiple cycles, insecticide decay and natural mortality.⁷ None is exotic; all three are properties any real deployment has.

Which means a model omitting them is not a simplified version of reality but a different question, and that reading is ours.

15.3 Which is a warning about all of this

These are model outputs, and the ordering of strategies depends on which realistic features the modeller chose to include.⁷

15.4 The authors say so themselves

Noting that while the quantitative effect may not be dramatic, the inclusion of multiple cycles and natural mortality altered the qualitative results.⁷

15.5 And they report a reassurance too

That agreement between two models was high, giving confidence in the model predictions as they are robust to the assumptions made when calculating insecticide selection.⁷

16. And one bad result was a modelling artefact

An instructive disclosure.

When insecticides were allowed unique properties, the rotation strategy performed badly. However, this is due to the deployment rules of the rotation model, which prevent immediate redeployment: if one insecticide has failed there is no longer anything available to be rotated so the simulation is terminated.⁵

16.1 The strategy did not fail, the simulation stopped

Which is an entirely different thing, and the reported result would have been misleading without the explanation.⁵

17. Which the authors disclosed

And that is worth saying plainly.

They reported a result unfavourable to one strategy and then explained that it was an artefact of their own implementation rather than a property of the strategy.⁵

17.1 This journal spends a lot of time flagging sources

So it should note when a source flags itself. The disclosure is what makes the rest of the paper credible, and it is the standard we try to meet in our own limitations sections.

18. The one experiment we found

Against a literature that is almost entirely theoretical.

A study on a moth that has demonstrated an ability to develop resistance to many different classes of insecticides, including bacterial proteins and two named newer compounds, tested mosaics against rotations.⁸

18.1 The result

Results of both population density and resistance development indicated that insecticide rotation every generation was better for IRM than if the insecticide was rotated every third generation or if the three insecticides were applied as a mosaic.⁸

18.2 An experiment is worth noting for its rarity

Resistance evolution takes generations, so a comparison of strategies requires running a population for a long time under controlled selection, which is expensive and which is why the literature is overwhelmingly theoretical.

That is the structural reason the models are not tested, and it is not going to change. It makes the model-dependence in §15 a permanent condition rather than a temporary one.

18.3 Rotation beat mosaic

Which is consistent with the modelling summary in §4 only in the weak sense that both are multiple-attack strategies with modest differences between them.¹⁸

The experiment did not include a mixture arm or a sequence arm, so it does not test the comparison this article is about.

19. Rotation frequency

The more interesting half of that result.

Rotating every generation beat rotating every third generation.⁸

19.1 Which discriminates between the explanations in §6

Counterselection from fitness cost should favour longer intervals, since an allele needs time out of selection to decline. Intergenerational redundant kill should favour shorter intervals, since it depends on survivors of one chemistry meeting the next.

The observed direction favours the second, and that inference is ours from the two sources.⁸⁴

19.2 One experiment, one species

We are not going to build much on it. But it is the only empirical comparison we located and it points somewhere specific.

19.3 And it has a practical reading

If shorter rotation intervals are better, then rotating annually, which is what a service schedule naturally produces, is the weak version of the practice.

20. Rotation's actual advantage

Which appears in the literature and is not about evolution.

The fact that rotations have preplanned insecticide replacement already in place may make them more operationally robust than sequences whose unpredictable replacement intervals require rapid response.⁵

20.1 A sequence requires noticing failure

And acting on it promptly, which requires monitoring that our articles on resistance monitoring and on treatment failure both found to be rare in practice.

20.2 Which our own journal has documented

Our article on treatment failure found that a failed treatment is usually reported as a reinfestation, and our article on resistance monitoring found diagnostic testing effectively absent from structural practice.

A strategy whose trigger is detecting failure has, in this trade, no trigger.

20.3 A rotation requires only a calendar

The switch happens on schedule whether or not anybody has detected a problem, which removes the decision that is most likely to be made late or not at all.

20.4 Which is the same argument as §7.2

Where a mixture's robustness was attributed to requiring no replacement decisions at all.⁵

Both arguments are about implementation rather than biology, and they favour different strategies by the same reasoning.

21. Which is not an evolutionary argument

And should not be presented as one.

If rotation's benefit is largely that it forces a switch somebody would otherwise postpone, then it is a management control rather than a resistance mechanism, and the case for it rests on human behaviour rather than on population genetics.

21.1 And it would explain the universality

A recommendation that survives decades without its mechanism being settled is more likely to be surviving on something other than its mechanism. If what rotation reliably delivers is a scheduled decision rather than a genetic effect, that is enough to keep it in every guideline whether or not the population genetics works out.

That is speculation and we mark it as such.

21.2 That is still a case

A good one. Our article on technician training described a trade where the commercially rational investment is not the technically optimal one, and a strategy that is robust to being implemented by busy people has real value.

21.3 But it changes what the evidence should be

Compliance data rather than selection modelling, and we found none of it.

22. The definitional problem

A methodological point the review raises that applies beyond this.

The authors note that there is an idealised 'mixture concept' that is the subject of theoretical analysis and modelling, which can be different from a practical implementation that is described as a mixture, with the concept set up to represent the most mixture-like-mixture, as distinct from other strategy concepts.¹

22.1 So the model and the product are not the same object

A commercial formulation containing two actives may not behave like the idealised mixture the modelling evaluates, and the conclusion does not transfer automatically.¹

22.2 The same applies to rotation

An idealised rotation switches on schedule with no residue overlap. A real one leaves the previous product's residue in place while the new one is applied, which is neither a rotation nor a mixture.

That extension is ours, and our chlorfenapyr article found a case where the residue overlap would matter chemically as well as evolutionarily.

23. What a contractor should take from this

Four points, stated modestly.

Rotation is worth doing and is not the guarantee it is presented as.¹

Shorter intervals are better than longer ones, on the one experiment we found.⁸

A registered mixture product is a reasonable choice and is not automatically superior. The advantage has a domain and §13 argues our setting may sit outside it.¹

And rotating to a product the population is already cross-resistant to achieves nothing. Which is the practical form of §8 and requires knowing the mechanism, not just the class.

23.1 That last one is the hardest and the most useful

Rotation is normally described in terms of mode of action classification, which is a statement about the target site. Metabolic cross-resistance runs across mode of action groups, so a rotation that looks correct on the classification can be no rotation at all.

Our synergists article and our chlorfenapyr article both describe enzymes with broad substrate specificity, and the second of those is the case where rotating to a different group changes the relationship with those enzymes entirely rather than escaping it.

24. And what we are correcting in this journal

Being specific.

Our reversion article opened by stating the assumption rotation depends on, which was that resistance decays when selection stops. It then found that the decay frequently does not happen.

24.1 What we should have added

That counterselection is only one of three proposed mechanisms for rotation, that the other two do not require a fitness cost, and that the failure of reversion therefore weakens rotation's case without destroying it.⁴

24.2 And the broader correction

We have treated rotation as established practice with a known basis. The 2024 paper's position is that the principle at work has not been elucidated.⁴

25. The Manitoba position

Short.

Nothing in this literature is Canadian and none of it concerns structural pests. The local application is that our bed bug and cockroach work involves populations with the layered resistance our earlier articles describe, and §8 says cross-resistance is precisely the condition that changes the comparison.⁷

25.1 And one thing the climate contributes

Generation time. A population inside a heated Winnipeg building runs its life cycle year-round, which our overwintering and thermal refugia articles describe, so the number of generations under selection per calendar year is set by the building rather than by the season.

Every result in this article is expressed in generations rather than years, so a warm building converts a resistance timeline into a shorter calendar one. That is our observation and it applies whichever strategy is chosen.

25.2 What we could not find

Any resistance management strategy modelling for structural pests, any Canadian guidance beyond the generic instruction to rotate, and any data on what practitioners here actually do.

26. Limitations and open questions

This is a modelling literature. With one experimental comparison, in one agricultural moth, that does not test the comparison this article is about.⁸

Most sources were read as abstracts and extracts. The reviews and the quantitative models reach us through published abstracts and excerpted sections rather than full papers.¹⁵

The settings are vector control and agriculture. Both differ from a treated building in exposure, population structure, generation time and the availability of untreated refuges.

The rankings are model-dependent. Section 15 is the source's own warning.⁷

And one review notes a gap in the modelling itself. That a comprehensive review of IRM modelling highlighted a lack of models including quantitative resistance and cross resistance.⁶

Sections 1.2, 2.1, 3.2, 4.2, 6.1, 6.2, 9, 10.2, 12.1, 13, 14, 19.1, 19.3, 21, 22.2, 23 and 24 are our reasoning. The cross-resistance interpretation, the exposure argument for structural settings, the reading of the rotation frequency result against the three explanations, the implementation-versus-biology distinction and the extension of the definitional problem to rotation are ours rather than sourced positions.

27. Conclusion

The instruction to rotate chemistries is given everywhere and its evidential position is not what that suggests. Modelling of mosaics and rotations has tended to show minimal gains over sequences, and a sequence is using one product until it fails and then switching, which the literature calls the nonstrategy benchmark.¹ Three explanations have been offered for how rotations work, namely counterselection from a fitness cost, relaxation of selection and intergenerational redundant kill, and a 2024 assessment concludes that all three can rationalise particular comparisons and none has elucidated the principle at work.⁴

Mixtures do better in the models, with one vector study concluding that full-dose mixtures are overall the best strategy and that their advantage grows as cross-resistance increases. But the advantage has a domain. As exposure approaches complete and effectiveness approaches zero, a mixture stops delivering redundant killing and simply selects for both resistances at once, which at the limit makes it take half as long to fail as the benchmark it was supposed to beat.¹⁵

A treated building is a closed system in which most of the population meets the residue, and a second chemistry is usually reached for because the first is performing poorly. High exposure and low effectiveness are jointly the condition under which mixtures lose, so the vector control conclusion may not transfer here. We cannot demonstrate that, because nobody has modelled resistance management strategy for structural pests and nobody has measured what proportion of an indoor population a treatment actually reaches. What we can say is that the advice arrives without its conditions attached, that the conditions are stated in terms of two variables, and that a contractor rotating annually because the label says to rotate is doing the weakest version of a practice whose mechanism is disputed.

References

1. Beyond redundant kill: a fundamental explanation of how insecticide mixtures work for resistance management. Open-access journal article in a pest management science title. Used for the statement that there are four fundamental strategy concepts, namely sequences, rotations, mosaics and mixtures; for the finding that modelling of mosaics and rotations tended to show minimal gains on sequences, which was often viewed as the nonstrategy benchmark where there is a switch in the solo use of insecticides when they fail due to resistance evolution; for the methodological point that there is an idealised mixture concept which is the subject of theoretical analysis and modelling and which can differ from a practical implementation described as a mixture, the concept being set up to represent the most mixture-like mixture as distinct from other strategy concepts; for the result that sequences can evolve resistance more slowly than mixtures with higher exposure and lower effectiveness; for the explanation that as exposure approaches 100 per cent or effectiveness approaches zero, mixtures look increasingly like independent use, so the contrast becomes resistance to two identical insecticides evolving simultaneously in a mixture or one after the other in a sequence, which would make a mixture take half as long as a sequence; and for the finding that at lower exposure and higher effectiveness mixtures can perform relatively better than sequences at delaying resistance, which previous studies have concluded is when mixtures work, a result often framed in the logic of redundant kill while the effect of exposure has been neglected. https://pmc.ncbi.nlm.nih.gov/articles/PMC10092901/
2. Bibliographic record for the above paper, national library of medicine database. Used for the listed keywords covering additional kill, epistasis, mixtures, modelling, population genetics, redundant kill and resistance evolution; and for the figure legend classifying the fundamental strategy concepts, describing sequences as involving temporally nonoverlapping exposure with the solo use of one insecticide until control failure due to resistance evolution followed by the solo use of another, and describing mosaics, independent uses and mixtures as involving simultaneous use with different spatial patterns of exposure, with mosaics involving spatially nonoverlapping exposure due to negatively correlated use in different places and independent use involving randomly overlapping exposure with no correlation. https://pubmed.ncbi.nlm.nih.gov/36098048/
3. What is the value of rotations to insecticide resistance management? Journal article in a pest management science title, publisher page, read as abstract. Used for the framing that in recent years there has been a resurgence of interest in the use of insecticide mixtures, particularly based on new theoretical models, and that the paper presents a perspective on the value of rotations focusing on the interpretation of influential theoretical models; for the statement that the principles of resistance management have previously been reduced to moderation, saturation and multiple attack, and that alongside mixtures and mosaics, rotations have been presented as a strategy of multiple attack in using more than one insecticide; and for the authors' stated approach of comparing rotations with sequences, mosaics and mixtures under the principle of multiple attack and arguing that rotations delay resistance through alternation. https://scijournals.onlinelibrary.wiley.com/doi/10.1002/ps.7939
4. Publication database record reproducing the abstract of the above rotations paper. Used for the statement that three explanations have been offered for how rotations delay resistance evolution, namely counterselection from resistance cost, the relaxation of selection and intergenerational redundant kill; and for the authors' conclusion that all three explanations can make sense of the comparison of rotations with another resistance-management strategy but have failed to elucidate the principle at work. https://ngdc.cncb.ac.cn/openlb/publication/OLB-PM-38173134
5. Insecticide resistance management strategies for public health control of mosquitoes exhibiting polygenic resistance: a comparison of sequences, rotations and mixtures. Journal article in an evolutionary applications title. Used for the presentation of a flexible quantitative genetics model calibrated primarily for a named malaria vector but adaptable to other species, investigating how insecticides could be deployed temporally as sequences, rotations or mixtures; for the stated hypothesis, based on results from monogenic models, that the difference between rotation and sequence strategies would be small and the mixture strategy would perform best; for the model assumptions that resistance is a classically quantitative genetic trait encoded by many genes of very small effect and that generations are discrete and nonoverlapping; for the conclusion that mixtures are likely to be the most robust because no replacement decisions are required, supporting the conclusion that full-dose mixtures are overall the best strategy; for the disclosure that when insecticides were allowed unique starting resistance, heritability and fitness cost the rotation strategy performed badly, but that this was due to the deployment rules of the rotation model, which prevent immediate redeployment so that if one insecticide has failed there is nothing left to rotate and the simulation terminates; and for the observation that the preplanned insecticide replacement built into rotations may make them more operationally robust than sequences, whose unpredictable replacement intervals require rapid response. https://onlinelibrary.wiley.com/doi/10.1111/eva.13546
6. Preprint version of the above comparison paper, preprint server. Preprint rather than the final published version, flagged accordingly. Used for the definition of rotations, also referred to as cycling, alternation or periodic application, as the periodic preplanned switching between insecticide formulations over time such that one insecticide is temporarily replaced with another; for the note that mixtures are also occasionally referred to as combinations or, for transgenic crops, as pyramiding, and that in public health the term combination would more generally refer to the use of two different insecticides within the same household but not in the same control tool; for the authors' definition of mixtures as the use of two insecticides in a single formulation such that the target insect inevitably encounters both simultaneously; and for the note that a comprehensive review of resistance management modelling highlighted a lack of models including quantitative resistance and cross resistance. https://www.biorxiv.org/content/10.1101/2022.10.28.514211.full.pdf
7. Simulating dynamic insecticide selection pressures for resistance management in mosquitoes assuming polygenic resistance. Open-access journal article. Used for the demonstration of how the inclusion of insecticide decay, cross resistance and multiple cycles affects the comparative performance of named resistance management strategies, run over a stated period with two deployment intervals; for the finding that including multiple cycles, insecticide decay and cross resistance resulted in micro-mosaics performing second best behind full-dose mixtures, while including only insecticide decay and cross resistance resulted in micro-mosaics becoming the worst performing; for the findings that the benefit of full-dose mixtures versus rotations or sequences increased with increasing cross resistance and decreased with negative cross resistance, the latter being expected since negative cross resistance maintains lower resistance levels for all strategies; for the observation that agreement between two models was high, giving confidence that predictions are robust to assumptions made in calculating insecticide selection; and for the disclosure that while the quantitative effect may not be dramatic, the inclusion of multiple cycles and natural mortality altered the qualitative results, with micro-mosaics moving from performing worse than rotations under a single cycle, to equally well with multiple cycles, to better once natural mortality was included. https://pmc.ncbi.nlm.nih.gov/articles/PMC12058183/
8. Testing insecticide resistance management strategies: mosaic versus rotations. Bibliographic record and abstract for a journal article, national library of medicine database. Used for the note that the named agricultural moth has demonstrated an ability to develop resistance to many different classes of insecticides including proteins produced by a named bacterium and, more recently, to two named newer compounds; and for the conclusion that results of both population density and resistance development indicated that insecticide rotation every generation was better for resistance management than rotating every third generation or applying the three insecticides as a mosaic. https://pubmed.ncbi.nlm.nih.gov/20552665/