Not Attraction: What Two Thousand Years of Light Trapping Got Wrong

1. Introduction: an old method with no explanation

Every insect light trap this journal has written about works on a principle nobody had established.

1.1 Perpendicular to the source

Which is not the direction anything attracted to something travels.¹

1.1b And the claim is about mechanism, not about whether traps work

They catch insects, which two thousand years of use establishes better than any study could. What was never established is why.

1.2 What this article argues

That the mechanism inverts the usual description, that the practical consequence is about fixture geometry rather than lamp colour, and that the finding covers a narrower question than it first appears to. Sections 9, 19 and 23 are the case.

2. How old

Older than most things this journal discusses.

Artificial light is an ancient method to trap insects, with the earliest written records dating back to the Roman Empire around 1 AD, and the reference list includes an item titled for the very first light trap, dated 1565.²⁴

2.0b Which makes the gap remarkable

Two thousand years of practice, several centuries of deliberate trap design, and a large body of observations on how catch varies with wavelength, moon phase and weather, all accumulated around a mechanism nobody had established.²

That is our framing, and it is worth holding alongside the articles in this journal that have found similar gaps in much younger practices.

2.1 With a literature to match

Efforts to improve light trap efficiency have generated many observations about nocturnal phototaxis, including data on the effects of wavelength, the moon, sky brightness and weather.²

2.1b Which is data about outcomes rather than about causes

Knowing that catch rises with certain wavelengths and falls with a full moon constrains what the mechanism could be without identifying it, which is why the observations accumulated for so long without settling the question.²

2.2 And several qualitative models

Which the same paper says were proposed in consequence.²

3. What was believed

Explanations of why nocturnal insects fly erratically around fires and lamps have included theories of lunar navigation and escape to the light, with researchers also suggesting attraction to the heat of the light, or blinding causing erratic movement and crashes.¹⁵

3.1 Two of them assume the insect wants to be there

Lunar navigation and escape to the light both describe an animal pursuing something. The other two describe an animal overwhelmed. The finding in §6 belongs to neither group, since the insect is neither pursuing nor overwhelmed but executing a reflex correctly under conditions that make it wrong.

That classification is ours.

4. Why none of it could be tested

Stated as the reason the question stayed open.

Without three-dimensional flight data to test them rigorously, the cause for this odd behaviour has remained unsolved.¹

4.0b And the observation is genuinely hard to interpret by eye

An animal circling a lamp looks like an animal trying to reach it and failing. Distinguishing that from an animal holding an orientation requires knowing which way the body is pointing at each instant, which is precisely what a two-dimensional view of a moving insect does not give.¹

4.1 Which is a methods problem rather than an interest problem

A great many people have watched insects at lamps. Watching does not distinguish an animal heading somewhere from an animal failing to hold a heading.

That framing is ours.

5. What was done

Two settings.

High-resolution motion capture in the laboratory and stereo-videography in the field, used to reconstruct the three-dimensional kinematics of insect flights around artificial lights.¹ The laboratory lights were a UV LED bulb, a UV-blue actinic tube and a cool white LED bulb.⁵

5.0b Two settings answering two objections

A laboratory gives the resolution needed to recover body orientation. A field site gives animals that arrived on their own rather than being released. Running both addresses the complaint each would attract alone.¹

5.1 So the finding rests on trajectories rather than counts

Which is what distinguishes it from two thousand years of catch data.

6. What was found

Quoted in §1.

Insects within a controlled environment did not head towards the light source, but predominantly orbited it.⁴ Simulated agents predominantly travelled orthogonally to the direction of the light, as seen in both field and laboratory observations of real insects.³

6.1 With a simulation as corroboration

The guidance model built from the tilting rule reproduced the flight paths recorded, which is what allows the authors to call dorsal tilting sufficient to create them rather than merely consistent with them.¹

6.2 And the simulated agents behaved like the real ones

Travelling predominantly orthogonally to the direction of the light, as seen in both field and laboratory observations.³

7. The reflex itself

Most flying insects display the dorsal light response, a behaviour that keeps their top side to the brightest visual region which was, until humans lit up the night, usually the sky.⁵

8. Why it works under a sky

Because the sky is a reliable indicator.

Under natural sky light, tilting the dorsum towards the brightest visual hemisphere helps maintain proper flight attitude and control.¹

8.0b And the reference is free

An insect needs to know which way is up continuously, and the sky supplies that without any organ dedicated to the purpose, which is presumably why the arrangement is conserved so widely.⁷

That reading is ours.

8.1 Which is an attitude reference rather than a navigation cue

It tells the animal which way is up, not which way to go, and the distinction is what the whole finding turns on.

8.1b Which is why the failure is so complete

A navigation error puts an animal in the wrong place. An attitude error stops it flying properly at all, which is what §10 records.

8.2 And it substitutes for machinery insects mostly lack

The paper notes that the largest flying insects can leverage passive stability to stay upright, which implies the smaller ones rely more heavily on the visual reference.⁴

9. And fails under a lamp

The authors set out three possible consequences.

That an artificial point source could reduce or remove the accuracy of the dorsal light response and mislead insects to tilt their dorsum away from the sky, misdirect lift generation and disrupt flight stability, or inhibit coherent heading control.⁴

9.0b The three possibilities are not alternatives

Misleading the tilt, misdirecting lift and disrupting heading control describe one failure at three levels: what the animal perceives, what its wings then do, and where it ends up. The recorded flights show all three at once.⁴

9.1 And the consequence recorded is continuous steering

Near artificial sources this highly conserved response can produce continuous steering around the light and trap an insect.¹

9.1b Trapped is the word the authors use

Which is closer to a physical description than to a behavioural one: the animal is held by a control loop it cannot switch off rather than detained by anything.¹

9.2 The animal is not going anywhere

It is holding an orientation, and the orientation it is holding produces a circle.

That restatement is ours.

10. The three patterns

Orbiting, stalling and inversion, with the insects consistently directing their dorsal axis towards the light source, even if this prevented sustained flight and led to a crash.⁴

10.0b All three are failures of the same kind

Orbiting is the animal holding the tilt and turning. Stalling is the animal holding the tilt and climbing until it cannot fly. Inversion is the animal holding the tilt when the only way to hold it is upside down. One rule, three outcomes, depending on where the light sits.⁴

10.1 Stalling described

A steep climb as the insect faced away from the light source, losing speed until it ceased to make progress.⁴

11. Inversion is the vivid one

And worth quoting at length.

Inversion of attitude, through roll or pitch, occurred when the insect flew directly over a light source, resulting in a steep dive to the ground. Once below the light, insects frequently righted themselves, only to climb above the light and invert once more.⁴

11.0b And it is the clearest evidence against attraction

An animal trying to reach a light does not dive away from it on passing overhead. An animal holding its back toward the light has no choice, because the only way to point the dorsum downward is to invert.⁴

11.1 Which explains something everybody has seen

The moth that repeatedly drops out of the air beneath a lamp and climbs back into it.

11.1b And the cycle can repeat indefinitely

Righting below the source restores level flight, climbing returns the animal to the position that caused the inversion, and nothing in the loop supplies an exit.⁴

11.2 And it is not a crash into the light

It is a loss of attitude control caused by the light being underneath the animal at the moment it passes over.

12. And the wind result is the practical one

Which we had not expected.

Orbiting was prevalent at low wind condition, below one metre per second, with insects dispersing if a gust of wind arose.⁴

12.0b Which is consistent with the mechanism

A reflex holding an orientation is overwhelmed by a force that moves the whole animal, so a gust does not persuade the insect to leave; it removes it before the orientation can be reasserted.⁴

12.1 So the trapping is fragile

A light that accumulates insects on a still night accumulates fewer on a breezy one.⁴

12.1b And it may explain a familiar complaint

A monitoring programme that records high counts one week and low counts the next, with no change in the building, has a candidate explanation in the weather rather than in the population.⁴

12.2 Which is a confound for every light trap catch series

Our article on interpreting trap catch argued that catch is a function of the trap as much as of the population. This adds wind to the list of things that move the number without moving the insects.

That connection is ours.

13. How the competing theories were ruled out

Two arguments, both neat.

13.1 Neither is a new experiment

Both arguments are made from things already known: what LEDs emit, and what the recorded flight paths look like. The paper disposes of two centuries of speculation without running a study aimed at either.⁷

14. The heat argument

Which disposes of an old explanation.

The effect of LED lighting, which supplies negligible infrared radiation yet still entraps vast numbers of insects.⁷

14.0b And the older sources did emit heat

Which is why the hypothesis survived: incandescent and discharge lamps produce substantial infrared, so heat and light arrived together and could not be separated by observing a lamp of that era.⁷

14.1 A technology change acting as an experiment

Nobody designed the transition to solid state lighting as a test of the heat hypothesis, and it functioned as one.

That observation is ours.

15. The blinding argument

Disposed of by the trajectories themselves.

The predictable light-centric flight trajectory motifs elicited argue against insects being blinded by light.⁷

15.0b And the two rejected theories fail differently

Heat is ruled out by a change in the stimulus, since LEDs removed the infrared and kept the effect. Blinding is ruled out by the character of the response, since the flights are too structured to be a loss of vision. One is an experiment the world ran, the other an inference from the data.⁷

15.1 Because a blinded animal would fly randomly

And what was recorded was orbiting, stalling and inversion, which are structured and repeatable.

16. And why the explanation generalises

The argument for parsimony.

The dorsal light response is described as the most parsimonious explanation of insect light entrapment, being a basal sensory mechanism, which the authors say explains the high prevalence of light attraction across a wide range of insects both diurnal and otherwise.⁷

16.0b And it covers diurnal insects too

Which is a point in the explanation's favour, since the phenomenon is usually framed as a nocturnal one and the reflex is not.⁷

16.0c And parsimony is doing real work in the argument

The competing accounts would each need to apply across every group that shows the behaviour, which for lunar navigation in particular is a strong assumption about animals that do not obviously navigate by the moon.⁷

16.1 A basal mechanism explains a universal behaviour

Where a specialised one would not, since the phenomenon appears across groups that share little else.

17. What this changes about our light trap article

A correction to our own work.

That article treated insect light traps as attracting insects, examined ultraviolet output decay and placement, and recommended positioning on the assumption that the device draws animals toward it.

17.0b And the error was in the explanation rather than the practice

Which is the useful kind of mistake to find, since nothing about how the devices are used depended on the account being right.

17.1 The draw is not a draw

At close range it is a steering failure, and a trap catches because an animal that has entered its vicinity cannot leave.¹

17.1b And the same word appears throughout this trade

Attractant, attraction, draw: the vocabulary of light-based devices is built on a premise that a trajectory reconstruction now contradicts, and replacing it would mean rewriting a great deal of routine description for a distinction most readers will not care about.

17.2 Which we should have flagged as unresolved rather than assumed

Although in fairness the mechanism was unresolved until recently.¹

18. Which is a correction rather than a reversal

Because most of that article survives.

Ultraviolet output still decays, placement still matters, and a trap in the wrong place still catches nothing. What changes is the account of why a well-placed one works.

18.0b Which is how a mechanism finding usually lands on practice

Most of what people do was arrived at empirically and works for reasons that were never the stated ones. A correct mechanism reorganises the explanation and changes a minority of the actions.

18.1 And one recommendation is strengthened

That traps should not be placed where they are visible from outside a building, which under the old account risked drawing insects in and under the new one risks capturing insects that would otherwise have passed by.

18.1b And the underlying reason is stronger now

Under the old account a visible trap was a beacon competing with the outdoors. Under the new one it is a hazard placed in a flight path, which is a more specific objection and easier to explain to a client.

18.2 With the same conclusion by a different route

Which is a good sign for the recommendation.

19. The design consequence

The usual advice about exterior lighting concerns colour: warmer lamps, less blue, less ultraviolet. The mechanism points somewhere else.

19.1 Colour advice is not wrong, only incomplete

Section 25 records a 2025 paper finding intensity and colour do affect attraction, so the point is that a second and largely unaddressed variable sits alongside them.⁶

20. Which is unusual advice

Our reading, before we give the authors' version.

If the problem is that a bright region appears where the sky is not, then what matters is where the brightness is relative to the flying animal, which is a question about fixtures, shielding and mounting rather than about lamps.

20.0b And it reframes what a bright light costs

On an attraction account, brighter means a larger catchment and dimmer is straightforwardly better. On this account brightness matters through whether the source outshines the sky in the insect's visual field, which is a comparison rather than an absolute and depends on how dark the surroundings are.

That reasoning is ours and points toward the same conclusion by a different route.

20.1 A shielded downlight puts brightness below the insect only near the ground

An unshielded globe puts it beside and above insects across a wide volume.

20.2 And an uplighter is the worst case on this account

Because it creates a bright region directly beneath animals flying over.

Sections 20.1 and 20.2 are our inference and neither is tested by anything we read.

21. And it is the authors' own prediction

Stated in the paper, which is why we are willing to build on it.

In contrast, diffused artificial light in the same plane as the night sky should restore appropriate dorsal-light-response, allowing insects to fly normally.⁴

21.0b Which follows directly from §8

If the reflex works because the sky is a broad bright region overhead, then light with those properties should be read as sky and produce normal flight, while a point source anywhere else is read as sky in the wrong place.⁴

21.1 Diffused, and in the plane of the sky

Two properties, neither of which is colour.⁴

21.2 And it is a prediction rather than a result

Should restore, in the authors' wording, with no experiment reported testing it.⁴

22. What it means for a building

Because insect loading on a facade is a pest control variable.

Insects accumulating at exterior lights are the food supply for the spiders our misdiagnosis article described, the reason facades acquire webs and droppings, and the population from which interior entry occurs when a door opens.

22.0b And the effect is at the scale a building owner controls

Unlike the weather, the neighbourhood and the season, a fixture is a decision made by the person paying for the pest problem, which puts it in the small category of drivers our other articles identify as actually movable.

22.1 So exterior lighting is an inspection item

Which our built environment articles have not treated as one.

22.1b With one caution about expectations

Nothing we read measures how much insect loading a fixture change removes, so this is a reasoned intervention rather than a quantified one, and we would present it to a client on that basis.

22.2 And the change is capital rather than chemical

Refitting or shielding a fixture is a one-off cost with no recurring service attached, which is the same shape as the exclusion recommendations our other articles keep reaching.

23. The limits the authors state

We did not test the interaction between range and attraction, and other mechanisms might contribute to the arrival of insects at nocturnal light sources over longer ranges, including interference with the celestial compasses insects use for nocturnal navigation.⁴

23.0b The authors raise the alternative themselves

Which is the mark of a paper reporting a result rather than defending a position, and it is why we treat §23 as part of the finding rather than as an objection to it.⁴

23.1 With a caveat on the caveat

Even at long distances artificial light sources often remain brighter than the night sky and may cause dorsal tilting that would also steer an insect towards a light source.⁴

23.2 And the relevant experiment barely exists

The paper notes that only one experiment has tracked moth trajectories over the distances concerned.⁴

24. Which we want prominent

Because the headline is easy to overstate.

What is established is the mechanism of entrapment near a source. Whether a lamp draws insects from a field is a separate question the study did not address.⁴

24.0b Which matters for how the result is reported elsewhere

Insects are not attracted to light is a headline the paper does not quite support. Insects near a light are not steering toward it is what was shown.¹⁴

24.1 And the two are not alternatives

Section 23.1 allows that tilting could operate at distance as well, in which case the same mechanism covers both and the word attraction is wrong at every range.

24.2 Our summary in §1 should be read with that attached

Insects do not steer toward the light is a statement about the flight paths recorded.¹

25. Spectrum has not gone away

And a later paper says so.

A 2025 study is titled for the finding that intensity and colour of artificial light at night affect insect attraction in a taxon-dependent manner, and notes that current modelling of insect attraction is based on phototactic responses towards sources of relatively broader spectrum lights at a single intensity.⁶

25.0b The two findings are not in conflict

One describes what happens once an insect is near a source, the other how strongly different groups respond to sources of different colour and brightness. A mechanism can be universal while the threshold for triggering it varies.

That reconciliation is ours and neither paper offers it.

25.1 Taxon-dependent is the operative phrase

Meaning a lamp choice that reduces one group's response may not reduce another's.⁶

25.1b Which cuts against simple lamp-swap advice

A recommendation that reduces the response of one group while leaving another unchanged is not a general solution, and the buildings this journal is concerned with host more than one group.⁶

25.2 And the critique of existing modelling matters

If the modelling rests on broad spectrum sources at one intensity, then advice derived from it may not transfer to the narrow band sources now installed.⁶

25.3 We have the abstract and the title

Not the findings, so §25 records that the work exists and what it claims rather than what it showed.⁶

26. What we would change

Three things.

Treat exterior lighting as an inspection item. Section 22.1.

Look at the fixture before the lamp. Shielding and direction, on the reasoning in §§20 and 21.⁴

And stop saying attracted. Which is a description of a mechanism that appears not to be the mechanism.¹

26.1 And one thing we would not change

Any recommendation that already rests on placement rather than on mechanism, since §18 found those survive the correction intact.

27. Our own position

The disclosure.

We sell insect light traps and service them, and §17 says our account of how they work was wrong. The devices still catch insects, which was never in doubt.

27.0b And we are reporting against our own published account

Which is the second time this journal has corrected an earlier article, and we would rather find these ourselves than have a reader find them.

27.1 And the recommendation in §26 sells nothing

A shielded fixture is an electrical job, and the advice reduces the insect pressure a recurring service would otherwise be addressing.

28. The Manitoba position

28.1 The season is short and intense

Insect activity at exterior lights here is concentrated into a few months, which means a facade problem is a summer problem and a fixture change is assessed against a short window.

28.1b And the winter half of the year is irrelevant here

Which is unusual among the drivers this journal examines, most of which operate indoors and year-round.

28.2 What we could not find

Any Canadian work on exterior lighting and insect loading on buildings, any local guidance on fixture selection for this purpose, and any municipal lighting standard that mentions insects.

28.2b One local implication we can state

Exterior lighting on a building here is doing its insect work during the same months as every other outdoor pest pressure, so a fixture assessment costs nothing extra if done during a summer inspection that is happening anyway.

28.3 And the wind finding may matter locally

Section 12 reports orbiting prevalent below one metre per second, and this is not a still place.⁴

29. Limitations and open questions

This is one paper. A single 2024 study, read across its published version, its preprint and coverage of it, which is four references reporting one piece of work.¹²⁴⁵

That is the most important limitation because §§6 to 21 rest on it and a single result overturning a long-standing assumption is exactly the kind that warrants waiting for replication.

The species tested are not named in what we read. Insects and moths appear generically in the abstracts and coverage available to us, and the generality of the finding across groups is asserted from the basal nature of the reflex rather than demonstrated.⁷

The laboratory work used three light types. Two ultraviolet and one cool white, all point sources, with no warm or diffuse comparison reported.⁵

The design recommendation is untested. Section 21.2 records that the diffuse light prediction is a prediction, and §§20.1 and 20.2 are our extrapolation from it.

And the 2025 spectrum paper reaches us as a title and an abstract fragment.⁶

Sections 4.1, 8.1, 9.2, 11.2, 12.2, 14.1, 17, 20, 22 and 24 are our reasoning. The attitude-versus-navigation distinction, the reading of inversion, the wind confound for catch series, the correction to our own light trap article and the fixture geometry argument are ours rather than sourced positions.

30. Conclusion

Light has been used to catch insects since the earliest written records of the practice around the first century, and a 1565 item is cited as describing the first light trap.²⁴ The explanations offered over that period included lunar navigation, escape to the light, attraction to heat and blinding, and none could be tested without three-dimensional flight data.¹⁵ When that data was collected, by motion capture in the laboratory and stereo-videography in the field, insects did not steer toward the light. They turned their backs to it and flew perpendicular to it, orbiting, stalling, and inverting into a dive when they passed directly over a source.¹⁴

The mechanism is the dorsal light response, which keeps an insect's top side toward the brightest visual region, reliably the sky until people lit the night.⁵ A lamp is a brighter hemisphere in the wrong place, the reflex tilts the animal toward it anyway, and continuous steering around the source follows.¹ The heat hypothesis is answered by LEDs, which emit almost no infrared and trap insects in vast numbers, and the blinding hypothesis by the fact that the trajectories are structured and repeatable rather than random.⁷ Orbiting was prevalent below one metre per second of wind and insects dispersed on a gust, which is a confound for every light trap catch series anybody has recorded.⁴

The practical consequence points away from the usual advice. If the problem is brightness appearing where sky is not, then geometry matters more than colour, and the authors predict that diffused light in the same plane as the night sky should restore normal flight. That is a prediction rather than a result, and so is our extension of it to shielding and fixture choice.⁴ Two further cautions belong with all of this. The study addresses entrapment near a source and did not test arrival from a distance, where celestial compass disruption may also operate, though the authors note a distant lamp still outshines the sky and could tilt an insect toward it.⁴ And it is one paper. Our own article on insect light traps described those devices as attracting insects, which now looks like the wrong word for what they do.

References

1. Why flying insects gather at artificial light. Journal article in an open-access general science title, read as its published abstract and summary. Source for the statement that explanations of why nocturnal insects fly erratically around fires and lamps have included theories of lunar navigation and escape to the light, and that without three-dimensional flight data to test them rigorously the cause remained unsolved; for the method of high-resolution motion capture in the laboratory and stereo-videography in the field to reconstruct three-dimensional kinematics of insect flights around artificial lights; for the central finding that contrary to the expectation of attraction insects do not steer directly toward the light but turn their dorsum toward it, generating flight bouts perpendicular to the source; for the explanation that under natural sky light tilting the dorsum towards the brightest visual hemisphere helps maintain proper flight attitude and control while near artificial sources this highly conserved dorsal light response can produce continuous steering around the light and trap an insect; and for the statement that the guidance model demonstrates dorsal tilting is sufficient to create the seemingly erratic flight paths and is the most plausible model for why flying insects gather at artificial lights. https://www.nature.com/articles/s41467-024-44785-3
2. Open-access repository version of the same article, read as its introduction. Source for the statement that artificial light is an ancient method to trap insects with the earliest written records dating back to the Roman Empire around 1 AD; for the note that efforts to improve light trap efficiency have generated many observations about nocturnal phototaxis including phenomenological data on the effects of wavelength, the moon, sky brightness and weather, and that several qualitative models of how insects gather at light have been proposed in consequence; and for the description of the motion capture data allowing quantitative probing of aerial manoeuvres in free flight. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10827719/
3. Bibliographic record and abstract for the same study in a biomedical literature database, including figure legend text. Source for the statement that simulated agents predominantly travelled orthogonally to the direction of the light as seen in both field and laboratory observations of real insects; and for the presence in the reference list of an item on the effect of spectral composition of artificial light on the attraction of moths and an item titled for the very first light trap, dated 1565. https://pubmed.ncbi.nlm.nih.gov/38291028/
4. Preprint version of the same study, posted to a biology preprint server and read as its full text, which carries detail absent from the published abstracts. Source for the three possible consequences the authors considered, namely that an artificial point source could reduce or remove the accuracy of the dorsal light response and mislead insects to tilt their dorsum away from the sky, misdirect lift generation and disrupt flight stability, or inhibit coherent heading control; for the prediction that diffused artificial light in the same plane as the night sky should restore appropriate dorsal light response and allow insects to fly normally; for the note that the largest flying insects can leverage passive stability to help stay upright; for the descriptions of the three flight patterns, namely orbiting prevalent at low wind conditions below one metre per second with insects dispersing if a gust arose, stalling characterised by a steep climb facing away from the source and losing speed until progress ceased, and inversion of attitude through roll or pitch when flying directly over a source resulting in a steep dive to the ground with insects frequently righting themselves below the light only to climb above it and invert once more; for the observation that insects consistently directed their dorsal axis towards the light source even where this prevented sustained flight and led to a crash; for the statement that insects within a controlled environment did not head towards the light source but predominantly orbited it; and for the stated limitations that the authors did not test the interaction between range and attraction, that other mechanisms might contribute to arrival over longer ranges including interference with celestial compasses used for nocturnal navigation, that even at long distances artificial sources often remain brighter than the night sky and may cause dorsal tilting steering an insect towards them, and that only one experiment has tracked moth trajectories over such distances. https://www.biorxiv.org/content/10.1101/2023.04.11.536486v1.full
5. University news release reporting the study, issued by the institution of one of the authors. Institutional press material, cited for plain-language description and for one methodological detail. Source for the statement that researchers have also suggested insects are attracted to the heat of the light or that they become blinded by it causing erratic movements and crashes; for the identification of the three light sources used in the laboratory as an ultraviolet LED bulb, an ultraviolet-blue actinic tube and a cool white LED bulb; for the statement that the light sources significantly altered the insects' ability to control their bearings as they attempted to tilt their backs toward the lights; and for the description of the dorsal light response as a behaviour that keeps an insect's top side to the brightest visual region, which was, until humans lit up the night, usually the sky. https://www.imperial.ac.uk/news/251217/flying-insects-become-disorientated-trapped-artificial/
6. Intensity and colour of artificial light at night affect insect attraction in a taxon-dependent manner. Journal article in an insect conservation title, read as a title and abstract fragment together with part of its reference list, which we flag because we report what it claims rather than what it found. Source for the statement in its title that intensity and colour affect insect attraction in a taxon-dependent manner; for the observation that current modelling of insect attraction is based on phototactic responses towards sources of relatively broader spectrum lights of a single intensity; and for the existence in its bibliography of work on the neural substrate of spectral preference in a model fly and on the biological impacts of artificial light at night. https://resjournals.onlinelibrary.wiley.com/doi/10.1111/icad.12855?af=R
7. Full text of the same 2024 study hosted on a research-sharing platform, read for its discussion section. Source for the argument against the heat hypothesis, namely the effect of LED lighting which supplies negligible infrared radiation yet still entraps vast numbers of insects; for the argument against the blinding hypothesis, that the predictable light-centric flight trajectory motifs elicited argue against insects being blinded by light; and for the authors' conclusion that they consider the dorsal light response the most parsimonious explanation of insect light entrapment, being a basal sensory mechanism which thus explains the high prevalence of light attraction across a wide range of insects. https://www.researchgate.net/publication/377811697_Why_flying_insects_gather_at_artificial_light