The Denominator: What a Resistance Ratio Is Divided By

1. Introduction: the number under the line

This journal has reported resistance ratios repeatedly: a strain is so many hundred or thousand fold resistant. We have never asked what the comparison is against.

1.0b And it is named after a person

Which is how it appears in methods sections and which tells you something about how the material travelled: not through a repository but through an individual.²

1.1 One colony, one army base, one year

Supplying the denominator for much of a literature.²

1.1b And the question is not pedantic

A ratio is the field's principal quantitative claim about resistance, appearing in abstracts, extension bulletins and trade articles, and it is the figure that persuades people a product has stopped working.

1.2 What this article argues

That the reference strain does two jobs it cannot both do well, that a 2024 paper says the resulting comparisons confound selection with drift, and that reported ratios for one compound in one species span three orders of magnitude. Sections 11, 14 and 18 are the case.

2. What a resistance ratio is

Stated as a formula in one paper: the ratio equals the lethal dose for the field strain divided by the lethal dose for the reference strain.²

2.1 With a significance test attached

One paper judged whether a ratio was meaningful by checking that the confidence intervals around the two lethal doses did not overlap, which is a stricter treatment than simply reporting the quotient.²

2.1b And the doses themselves come from a dose-response curve

Several concentrations, mortality at each, and a fitted estimate of the dose killing half the sample, which is the standard the whole comparison rests on.²

2.2 Which is worth noting as good practice

Since a ratio computed from two uncertain numbers has uncertainty of its own, and most quoted figures arrive without any.

3. So the number has two halves

Our observation.

Everybody treats a resistance ratio as a measurement of the resistant population. It is a measurement of two populations, and an error in either moves it.

3.0b And only one half is usually described

Papers report where the resistant strain came from, when, and from what kind of building. The reference strain is named and the naming is taken to be sufficient, because everybody in the field knows which colony is meant.¹

3.1 And the denominator is small

A susceptible strain has a low lethal dose by definition, so the ratio is a large number divided by a small one, and small absolute errors in the small number produce large changes in the result.

3.1b And the two halves are measured on different material

One population has been in culture for decades and handles well. The other arrived from a building weeks earlier. They are not equally easy to work with, and the bioassay does not care about that.

3.2 Which is arithmetic rather than a criticism of anybody

But it is worth holding in mind through everything below.

4. Where the denominator comes from

Most comparative studies in this species used the Harlan strain, sampled in Fort Dix, New Jersey in 1973, as the susceptible reference strain.¹

4.1 Most is the word to notice

Not all, but enough that a critique of the reference strain is a critique of the field's comparative literature rather than of one laboratory.¹

5. The provenance

Recorded in the methods sections, where such things live.

The strain was collected in Fort Dix, New Jersey in 1973 and has been maintained at a named university since December 2008, with two named resistance mutations absent from it.⁵

5.0a The dates are recorded carefully

Collection year, transfer year and maintaining institution all appear in methods sections, which is better documentation than most of the trade material this journal reads.⁵

5.0b And its custody has moved

Collected in 1973, held by the collector, supplied to one laboratory in 2005 and to another in 2008, which means the colony in any given paper is a descendant population rather than the original one.²⁵

That framing is ours, though the dates are sourced.

5.1 And it has been in culture continuously

Described as in laboratory culture since 1973 and therefore not having experienced pesticide exposure for several decades.⁶

6. Fed on the man who collected it

The detail in §1.

Maintained in colony, feeding on the named researcher, since 1973.²

6.0b And it solves a real problem

Bed bugs need blood, and a colony maintained for decades needs a reliable source. Other laboratories in our sources use rabbit blood through glass feeders or heated chicken blood through a membrane.⁵³

6.1 Which is a remarkable piece of scientific infrastructure

A reference population for a global literature, sustained for decades by one person allowing it to feed on him.

6.2 And it is a fragile arrangement

Whatever the merits, a comparator maintained that way depends on a single individual continuing to do it.

Both are ours.

7. And supplied through a trade association

Which we note without comment.

One paper records acquiring the susceptible strain from the named researcher at the national pest management association in 2005.²

7.1 Which is not a conflict of interest

Supplying insects is not the same as funding a result, and the strain's value lies precisely in being susceptible, which is the opposite of what a trade body would want to demonstrate.

We note the route because provenance is what this article is about.

8. Fifty years without selection

Which is the property that makes it useful.

Two laboratory colonies that had never been exposed to pyrethroids, one collected more than thirty years earlier and one more than twenty, both showed complete mortality at a discriminating dose.³

8.0a Which is the only property it is ever tested for

That it dies. Nothing in our sources reports checking whether it is otherwise typical of the species as it existed in 1973 or as it exists now.

8.0b And a second old colony agrees with it

A Florida colony collected more than twenty years before that study also gave complete mortality, which is the only independent check on the reference we found anywhere.³

8.1 So the strain does the job it was chosen for

It dies reliably, which is what a susceptible reference has to do.³

9. Which is not the same as unchanged

Our argument.

A colony in culture for fifty years is not a population held still. It is a small population under a new selection regime: laboratory temperature, artificial feeding, a fixed harbourage, and whatever genetic drift a confined colony accumulates.

9.0a And the comparison being made assumes it has not moved

Every ratio treats the reference as a fixed point. If it has shifted in either direction over fifty years, every ratio computed against it shifts with it.

9.0b A closed colony is a small population

Which is the condition under which drift acts fastest, and fifty years is a great many bed bug generations.

That is our reasoning and §29 records that we found nothing testing it.

9.1 The selection pressure was removed, not the evolution

Insecticide exposure stopped in 1973. Everything else about being a laboratory animal started.

9.2 And we found no source addressing this directly

Which §29 records as the largest gap in the article.

9.3 And the strain is characterised for the thing that matters

Two named knockdown-resistance mutations are recorded as absent from it, which is a direct check on the property it is relied upon for, even if it says nothing about anything else.⁵

10. It also defines ancestral

A second role, in the genomic work.

One study notes that it considered derived alleles as opposed to ancestral ones, carried by the Harlan strain from which the reference genome was sequenced.⁴

10.1 Which is a practical choice rather than a mistake

A genome project needs an inbred, available, well-characterised colony, and the reference strain was all three by the time sequencing became possible.

That reading is ours and no source criticises the decision.

11. Which is a second job for one colony

And the two are not independent.

The same 1973 material defines what susceptible means in a bioassay and what ancestral means in a genome. A trait present in that colony is therefore both normal and original by construction.

11.0b And the two roles have opposite requirements

A good susceptible reference should be as sensitive as possible, which argues for an unusual population. A good genomic reference should be as typical as possible, which argues for the opposite.

11.1 The authors note the consequence for their method

That if adaptation involved standing genetic variation rather than new mutation, alleles already segregating in the population before selection began, then their pipeline may have filtered out valid candidates.⁴

11.1b Hard sweeps and soft sweeps behave differently here

A new mutation rising rapidly leaves a clear signal against an older reference. An allele already present and rising slowly does not, and the second is described as a recognised route to insecticide resistance.⁴

11.2 Which is the circularity stated by the people doing the work

A resistance allele present at low frequency in 1973 would look ancestral, and would not be detected as derived.

Section 11.2 is our restatement of §11.1.⁴

12. The 2024 objection

All the comparative studies mentioned so far were conducted on strains having very different genetic backgrounds.¹

12.1 Made about a body of work including well-regarded studies

The papers named are the ones that established much of what is believed about metabolic resistance mechanisms in this species.¹

13. Distance as a confounder

The reference strain was compared to strains sampled from distant locations, 800 kilometres away for one and 440 for another, which probably differ in many traits unrelated to resistance phenotype.¹

13.0b And the reference is fixed while the comparators vary

Every study reaches for the same denominator and supplies its own numerator, so the distance between them differs in each case while the comparison is reported the same way.¹

13.1 Distance is a proxy rather than the mechanism

Nothing about kilometres causes a difference. What distance stands for is separate population history, which accumulates differences of every kind.

14. Selection or drift

The conclusion.

Consequently, it is unclear whether the genes or transcripts identified in these studies are differentially expressed as a result of selection by insecticides or other selective factors, or as a result of nonadaptive genetic differentiation due to genetic drift.¹

14.1 Nonadaptive is the important word

Differences arising from population history rather than from anything selecting for them, which is the default expectation for two populations separated by hundreds of kilometres.¹

15. Which is a serious charge

Our assessment.

The claim is not that earlier studies were sloppy but that their design cannot separate the effect of interest from an alternative explanation, which means their positive findings are consistent with resistance and also consistent with two populations simply being different.

15.0b The objection is about inference rather than data

The measurements in those studies are presumably sound. What is contested is what may be concluded from a difference between two populations that differ in more than one way.

15.1 And it applies to a substantial body of work

Several named studies from 2011 and 2012 are cited as sharing the problem.¹

15.1b And the authors are describing a field they work in

Which is the kind of criticism that carries more weight than an outsider's, and which this journal has rarely been able to quote.

15.2 It does not mean the resistance is not real

Whether these populations survive pyrethroids is established by bioassay, not by transcriptomics. What is in question is the mechanism attribution.

16. And the fix is available

Demonstrated by the same paper.

It used two strains obtained from a single commercial supplier, described as closely related and varying in resistance, which controls the genetic background the earlier work could not.¹

16.0b And the source of the strains matters too

Both came from one commercial supplier, which removes the geographic separation that §13 identifies as the confounder.¹

16.1 Closely related is the operative phrase

Two populations differing mainly in the thing being studied, rather than in that and everything else.

16.2 Which took thirty years to arrive

The reference strain has been in use since the 1970s and the design objection is published in 2024, which is a long time for an obvious confound to sit unaddressed.

That is our observation, and §29 notes we cannot say whether it was raised earlier in venues we did not search.

17. The range of reported ratios

Which is the second half of this article.

17.1 And they concern one insecticide class

Pyrethroids throughout, in one species, which removes the easiest explanation for a spread this wide.

18. Seventeen to ten thousand

For the same compound class in the same species.

One paper reports its two relatively recent strains as not particularly high, 17 to 199 fold, and contrasts that with another study where field strains were 6,945 and 10,178 times higher than their susceptible reference.⁴

18.0b And the lower figure is described dismissively

Not particularly high is how the authors characterise ratios of 17 to 199, which for most purposes would describe a population a product cannot control.⁴

18.1 Both figures come from the same sentence

One paper describing its own results as modest by explicit comparison with another's, which is what makes the spread visible rather than something we assembled from separate sources.⁴

19. The warning attached

Given by the authors themselves.

Resistance ratio should be compared with caution, since the bioassays or analytical methods can be widely different and have an impact.⁴

19.0a It appears in a discussion rather than a limitations section

Placed where the authors compare their results with somebody else's, which is where the problem becomes unavoidable.⁴

19.0b And it names the reason precisely

Not sampling error or biology, but that the bioassays and the analytical methods differ between laboratories.⁴

19.1 Which is an unusually direct statement about a field's own numbers

And one that almost never travels with the figures when they are quoted.

20. A ratio with no upper value

From an earlier paper.

One colony showed no control mortality and only 21.6 per cent mortality at the highest concentration tested, so the resistance ratio was at least 6,123.³

20.0b Which is a different kind of statement from a measurement

Every other ratio in this article is a quotient of two numbers. This one is a bound, and it appears in the literature alongside the others as though it were the same kind of thing.

20.1 At least, because the dose could not be raised further

A lethal dose cannot be calculated for a population that does not die, so the figure is a floor and the true value is unknown.³

20.1b And the comparison they reach for is telling

That their figures appear to sit at the same order of magnitude as the highest resistance recorded in any insect species, which is a statement about the ceiling of what has ever been measured.³

20.2 The authors say so

Describing their ratios for two compounds as underestimates, while noting they appear to be of the same order of magnitude as the highest levels seen in other species.³

21. And ratios that fall in captivity

The finding that unsettles the whole measurement.

Very high ratios have frequently been observed for recently collected field strains, but often decline after rearing in the laboratory without insecticide exposure.⁴

21.0b And it is described as frequent rather than occasional

Very high ratios in recently collected strains are noted as a common observation, with the decline on rearing given as the usual sequel.⁴

21.1 Which our reversion article would predict

Selection stops when the animals enter the laboratory, and if resistance carries a fitness cost the susceptible types recover ground from the first generation.

22. Which makes timing a variable

Our reading.

If resistance declines during laboratory rearing, then a ratio depends on how long the field strain sat in culture before being tested, which is a property of the experiment rather than of the population it was collected from.

22.0b Which is the opposite problem from the reference strain

There the concern is a population held too long in culture. Here it is that the field strain is held in culture at all before being measured.

22.1 And it is rarely reported alongside the ratio

We did not find a generation count attached to any of the figures in our sources.

22.1b And it complicates any comparison across studies

Two laboratories testing the same field population at different points after collection would get different ratios, without either doing anything wrong.⁴

22.2 It also connects to our reversion article

Which examined resistance decay after selection stops and found fitness costs the proposed mechanism. This is that phenomenon operating inside the measurement rather than in the field.

23. The discriminating dose alternative

A different method, used in the same literature.

One study assessed presence or absence of resistance in ten populations using a discriminating dose of 0.13 milligrams per square centimetre of technical grade deltamethrin over 24 hours, finding a clear cut and significant difference among populations.³

23.0b And the design is a survey rather than a comparison

Ten populations tested against one fixed exposure, which produces a distribution across a region rather than a single number about one strain.³

23.1 With a useful conclusion attached

Resistance to pyrethroid insecticides is not a local phenomenon, nor is it universal.³

23.2 And the result was a distribution rather than a verdict

Clear and significant differences among the ten populations tested, meaning some were resistant and some were not, in the same region and the same year.³

24. What that method gives up

And what it avoids.

A discriminating dose answers whether a population survives a fixed exposure. It does not produce a ratio, so it cannot be inflated by a small denominator, cannot be reported as at least, and cannot be compared incorrectly against another laboratory's number.

24.0b And it produces a comparable result across laboratories

Because the dose is fixed in advance rather than derived from whatever reference colony a laboratory happens to keep.³

24.1 What it loses is magnitude

Two populations both surviving the dose are both resistant, with no statement about how much.

24.1b And a survey of populations is more useful than a number for one

Knowing that some buildings in a city hold resistant populations and others do not is a planning fact. Knowing that one strain is six thousand fold resistant is not, on its own.³

24.2 Which is often the answer a practitioner needs

Since the operational question is whether the product will work here, not by what factor it fails.

Sections 24 to 24.2 are ours.

25. What this means for our own articles

And this is why we wrote it.

This journal has quoted resistance ratios in articles on layered resistance, on cuticular penetration, on synergists and on fleas, in each case as evidence for how far a population has moved. Every one of those figures is a division by somebody's reference strain.

25.0b And a reader would have no way to check

Because we named the finding and not the comparator, which is the practice this article is criticising in the literature it drew from.

25.1 And one of them reported a thousandfold shift

Our article on cuticular penetration reported a resistance ratio rising by three orders of magnitude between exposure routes, which is exactly the kind of figure this article says should carry its reference strain with it.

26. Which we are not retracting

For three reasons.

The ratios were reported accurately from their sources. The resistance they describe is real and established by other means. And the comparative point most of those articles made, that one strain is far more resistant than another, survives considerable imprecision in the number.

26.0b And the change costs nothing

A clause naming the reference strain adds a dozen words and makes the figure interpretable by somebody who wants to compare it with another.

26.1 What we would change is the presentation

A figure like a thousandfold resistance should be given as a reported ratio against a named reference strain rather than as a property of the population.

26.1a And the same standard should apply to anybody quoting one

A figure without its comparator is not a measurement anybody can check, which is the test this journal applies to every other number it encounters.

26.1b Which changes very few of our conclusions

An argument that a population is resistant enough for a product to fail does not depend on whether the ratio is two hundred or two thousand.

26.2 And an order of magnitude is the most that should be claimed

Which is what §19's warning implies and what §20's floor requires.

27. Our own position

The disclosure.

We have quoted large resistance ratios in articles that also explain why a treatment failed, which is a combination that serves a company explaining a failed treatment. The corrective is §26.1 and we will apply it.

27.0b And we are not in a position to do better ourselves

Testing a local population would require a colony, a bioassay protocol and a susceptible comparator, none of which this company has or could maintain.

27.1 And the incentive runs the other way too

A large ratio is a better story for an article than a discriminating dose result, and this journal has published the former more often than the latter.

28. The Manitoba position

28.1 Nothing here is local

Every strain named in this article was collected in the United States or Argentina, and no Canadian population appears in any of our sources.⁴

28.2 What we could not find

Any resistance bioassay on a Manitoba population of any structural pest, any Canadian susceptible reference colony, and any local surveillance programme.

28.2b And a local reference colony would be a real asset

Any laboratory here maintaining a susceptible population of a structural pest would be supplying something the country appears not to have.

28.3 Which means the ratios we quote are foreign

And our molecular diagnostics article already identified the same gap from the genotyping side.

28.4 And the seasonal collection problem is real here

A field strain collected in a northern winter arrives from a heated interior population, which is one more difference between compared groups that nobody reports.

That is our speculation.

29. Limitations and open questions

We found no study of laboratory adaptation in this strain. Section 9 argues that fifty years of culture is itself a selection regime, and we located no work testing whether the reference colony has drifted, which is the article's central speculation and is unsupported.

That is the most important limitation because §§9 and 11 would be much stronger with it and are reasoning from general principles without it.

We read abstracts and methods sections. The quotations here come from introductions, methods and discussions accessed as extracts rather than from full papers read end to end.¹⁴

The ratio comparison in §18 may be unfair. Different compounds, different life stages, different exposure routes and different countries sit behind those numbers, which is what the authors' own caution in §19 is about, and we are using the spread to make a point about comparability rather than about any particular figure.⁴

The 2024 critique concerns transcriptomic studies specifically. Its objection is about attributing gene expression differences, and we have extended it to the reference strain's role generally, which goes slightly beyond what the paper says.¹

And one reference is a life table study. Used only for its description of strain maintenance and same-day comparative testing, not for anything about resistance.⁷

Sections 3, 6.1, 6.2, 9, 11.2, 15, 22, 24, 25 and 26 are our reasoning. The argument about the denominator, the account of laboratory culture as a selection regime, the circularity in the dual role of the reference strain and the case for discriminating doses are ours rather than sourced positions.

30. Conclusion

A resistance ratio is a division: the lethal dose for a field strain over the lethal dose for a susceptible reference.² For the common bed bug, most comparative work has used one colony collected at Fort Dix, New Jersey in 1973, maintained in laboratory culture ever since without insecticide exposure, fed for much of that period on the researcher who collected it, and supplied to other laboratories through a national trade association.¹²⁶ It works as a reference in the sense that matters, dying reliably at a discriminating dose alongside a second old colony.³ It also supplied the material for the species reference genome, so the same 1973 population defines both what susceptible means and what ancestral means, and the authors of one genomic study note that if resistance arose from variation already present rather than from new mutation, their method may have filtered out the very alleles they were looking for.⁴

A 2024 paper puts the problem directly. Comparative studies used that reference against strains collected 440 and 800 kilometres away, which probably differ in many traits unrelated to resistance, so it is unclear whether the differences found reflect selection by insecticides or nonadaptive genetic differentiation through drift.¹ That is not a complaint about care; it is a statement that the design cannot separate the effect from its alternative. The same paper shows the fix by using two closely related strains from one supplier. And the numbers themselves do not compare: one study reports ratios of 17 to 199 fold as not particularly high while citing another where field strains ran 6,945 and 10,178 times their reference, with the explicit caution that ratios should be compared carefully because bioassays and analytical methods differ; an earlier paper could only report a ratio as at least 6,123, because its insects would not die at the highest concentration tested; and high ratios are noted to decline after laboratory rearing without exposure, which makes the answer depend on how long the sample waited before being tested.⁴³

None of this makes bed bug resistance less real. Whether a population survives a product is settled by exposing it, and the discriminating dose method does exactly that, answering the question a practitioner actually has: does this work here. What the ratio adds is magnitude, and magnitude is the part that travels badly. This journal has quoted such figures in at least four articles as evidence of how far populations have moved, and we are not withdrawing them, because they were reported accurately and the resistance is established by other means. What we will change is how they are written: a reported ratio against a named reference strain, good to about an order of magnitude, rather than a property of the insect. We should also record that we found no study testing whether the 1973 colony has itself drifted over half a century in culture, which is the question this whole article turns on and the one nobody appears to have asked.

References

1. Transcriptomic response to pyrethroid treatment in closely related bed bug strains varying in resistance. Journal article in a genome biology title, read as an extract of its introduction and methods. Source for the statement that most comparative studies used the Harlan strain, sampled in Fort Dix, New Jersey in 1973, as the susceptible reference strain; for the criticism that all the comparative studies mentioned were conducted on strains having very different genetic backgrounds, compared against strains sampled from locations 800 kilometres away in one case and 440 in another, which probably differ in many traits unrelated to resistance phenotype; for the consequent conclusion that it is unclear whether the genes or transcripts identified are differentially expressed as a result of selection by insecticides or other selective factors, or as a result of nonadaptive genetic differentiation due to genetic drift; and for the description of the two strains used in that study as provided by a single named commercial supplier. https://academic.oup.com/gbe/article/16/8/evae158/7717632
2. Deep sequencing of pyrethroid-resistant bed bugs revealing multiple mechanisms of resistance within a single population. Open-access journal article, read as an extract of its methods and figure legends. Source for the formula that resistance ratios were calculated by dividing the field strain's median lethal dose by that of the reference strain, with significance judged by non-overlapping confidence intervals; for the account that pyrethroid-resistant field strain bed bugs were collected from a group home in Richmond, Virginia in autumn 2008 and transported to a named university laboratory; for the statement that a pyrethroid-susceptible strain was acquired from a named researcher at the national pest management association in February 2005, originally collected from Fort Dix, New Jersey and maintained in colony feeding on that researcher since 1973; and for the rearing conditions of both strains in environmental chambers. https://pmc.ncbi.nlm.nih.gov/articles/PMC3198472/
3. Insecticide resistance in the bed bug as a factor in the pest's sudden resurgence. Journal article in a medical entomology title, read as an extract. Source for the use of a discriminating dose of 0.13 milligrams per square centimetre of technical grade deltamethrin with 24 hour exposure on third to fifth instar nymphs; for the establishment of two laboratory colonies never exposed to pyrethroids, one collected more than thirty years earlier at Fort Dix and one more than twenty years earlier in Florida, both showing complete mortality at the discriminating dose; for the finding that one field colony showed no control mortality and only 21.6 per cent mortality at the highest concentration tested, so that the resistance ratio was at least 6,123; for the authors' statement that their ratios for two compounds are underestimates while appearing to be of the same order of magnitude as the highest levels of resistance seen in other insect species; and for the assessment of ten populations by discriminating dose showing a clear cut and significant difference among them, with the conclusion that resistance to pyrethroid insecticides is neither a local phenomenon nor universal. https://academic.oup.com/jme/article/44/2/175/862389
4. A major superlocus involved in pyrethroid resistance in the common bed bug. Open-access journal article, read as an extract of its discussion. Source for the observation that two older laboratory strains showed high sensitivity as expected since they were collected before the re-emergence of bed bugs, as was the reference strain; for the report that two relatively recent strains collected in 2008 and 2015 had resistance ratios of 17 to 199 fold based on the dose killing 95 per cent, described as not particularly high compared with a 2019 study where field strains were 6,945 and 10,178 times higher than their susceptible reference; for the caution that resistance ratios should be compared carefully since bioassays and analytical methods can be widely different and have an impact; for the note that very high ratios have frequently been observed for recently collected field strains but often decline after rearing in the laboratory without insecticide exposure; for the statement that the study considered derived alleles as opposed to ancestral ones carried by the reference strain from which the species reference genome was sequenced, that strain having been collected in Fort Dix in 1973; and for the caveat that if adaptation to pyrethroids involves standing genetic variation rather than hard selective sweeps, the analysis pipeline may have filtered out valid candidate alleles. https://pmc.ncbi.nlm.nih.gov/articles/PMC10197226/
5. Susceptibility of insecticide-resistant bed bugs to infection by a fungal biopesticide. Journal article in a pest management science title, read as an extract of its methods. Source for the identification of the Harold Harlan strain as the Fort Dix strain, collected in Fort Dix, New Jersey in 1973 and maintained at a named university since December 2008, with two named knockdown-resistance mutations absent from it; for the contrasting description of two named field strains collected in 2008 which are resistant to pyrethroids and carry both mutations; for the rearing conditions of temperature, photoperiod and relative humidity and the use of defibrinated rabbit blood through custom glass feeders; and for the confirmation that the reference strain was susceptible to deltamethrin while the field-collected strains tolerated relatively high concentrations. https://scijournals.onlinelibrary.wiley.com/doi/full/10.1002/ps.4576
6. Transcriptomics of the common bed bug. Open-access journal article, read as an extract of its methods. Source for the description of the susceptible strain as having been in laboratory culture since 1973 and hence not having experienced pesticide exposure for several decades; and for the contrasting description of pesticide-exposed bed bugs collected during 2009 and 2010 from an apartment that had undergone repeated insecticide treatments without successful control. https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0016336&type=printable
7. Population growth potential of the common bed bug, a life table analysis. Open-access journal article, read as an extract of a table note and acknowledgements and cited only for its description of strain maintenance rather than for any resistance finding. Source for the note that the reference strain was collected from a population maintained in the laboratory since 1973 while two others were field collected; for the statement that tests on both field strains and the laboratory strain were run on the same day for comparison using an equal sex ratio; for the named formulated products and their active ingredient concentrations used in that comparison; and for the acknowledgement recording that the laboratory-susceptible strain was provided by the named researcher at the national pest management association. https://pmc.ncbi.nlm.nih.gov/articles/PMC4553456/