It has been hard to escape media reports of plummeting insect populations over recent years (so-called ‘insect Armageddon’). Of course, the picture is far more uncertain and nuanced than newspaper headlines suggest (shocker!). Nonetheless, there is solid evidence of worrying long-term declines in insect numbers for some parts of the world. This is extremely concerning, not least because insects are the glue that holds the natural world together.
Given we often lack even the most basic data on insect populations, it will come as no surprise that our understanding of the factors that are causing declines is far from complete. But they are likely to be numerous and complex.
For biodiversity loss globally, conservation groups have traditionally recognised five key threats: habitat change (ranging from its complete removal through to more subtle degradation effects), species overexploitation (fishing, hunting), invasive species and diseases, climate change, and pollution (which usually refers to chemicals, including pesticides and fertiliser).
But as often turns out to be the case, reality can be more complicated (or even quite different) to long-held conventional wisdom. Some potential causes of wildlife declines appear to have been overlooked historically, or at least, their impacts have gone largely unstudied.
One of these neglected areas is undoubtedly the consequences of the surge in artificial lighting.
As night follows day
Insects have evolved on Earth over hundreds of millions of years. Throughout this time, life has experienced dependably predictable rhythms of days into nights. The distinction between bright days and dark nights was one of the few very constants in an ever-changing environment.
But this distinction between night and day has become blurred.
Since the end of 19th century, humans have increasingly shone ever-brighter lights into the darkness. That trusty pattern of night following day is no longer so reliable. What’s more, nocturnal creatures now frequently encounter a relative evolutionary novelty: artificial light at night.
Between 2012 and 2016, satellite measurements revealed that the global area affected by lighting grew by 2% each year. Places that were already illuminated got brighter, also by about 2% per year. Coming hand in hand with new developments and infrastructure projects, lighting is intruding ever deeper into global biodiversity hotspots. Some of Earth’s most threatened ecosystems, such as mangrove forests and the Mediterranean, are the most afflicted.
There is another important change underway: the rise of LEDs.
LEDs were meant to be a good thing. Energy-efficient. Good for the planet. Alas, it hasn’t exactly turned out this way. Paradoxically, more efficient lights haven’t necessarily reduced energy consumption. They may have even increased energy use. Enter: the rebound effect. As lighting technology gets more efficient, we pass up much of the potential savings. And instead, install ever-brighter lights, illuminating even more places.
A moth to a flame
Moths are certainly not the only group of insects to suffer from light pollution, but they are perhaps the best-studied in this context.
The large ecological and evolutionary diversity shown by moths means that they can be used as a proxy for insects more broadly. No single group can ever be a perfect representative, but I reckon that moths, are about as close as we are going to get (the extensive data we have on them also makes them uniquely useful).
Some insect groups are going to be especially sensitive to lighting. For instance, fireflies and glow-worms, that require dark nights if they are to have any hope of wooing potential mates with their own little lanterns. On the other hand, the minority of insects that are day-flying will presumably be more or less unaffected by light pollution.
The plight of insects around lighting is often symbolised by the famous attraction of moths to flames. But the ways through which lighting can affect insects are far more complicated than adults bumbling uncontrollably towards sources of light.
As we demonstrated in our recent review paper, lighting can have disruptive impacts during the entire life cycle of insects. Many fundamental biological processes are controlled by the day-night cycle (e.g., reproduction). The absence of dark nights, to take one rather unfortunate example, can mean male moths become impotent.
Importantly, this means that insects that are only active by day (as adults, e.g., butterflies) could still be negatively affected by light pollution. At least, in theory. Perhaps through a disrupted circadian rhythm, or via other life stages that are nocturnal.
Counting caterpillars
To discover what impacts lighting has on insect numbers, I have spent many days and nights dressed head-to-toe in hi-vis collecting data alongside roadsides over the last three years.
I started this process a few weeks into my PhD back in October 2019, making various Freedom of Information applications to get the streetlight inventories for the surrounding counties. Overlaying these to satellite imagery gave me the opportunity to spot areas that might have habitat with sections that were both lit and unlit. This was to form what’s called a matched pairs experimental design. The idea here is that locations are carefully matched to keep all other factors constant, except for the presence of lighting.
An initial pool of over 500 potential field sites was whittled down to 27 sites that met the stringent pairing criteria. These sites contained either hedgerows or grass margins – home to two different groups of caterpillars that can each be sampled using easily standardised methods.
The results were striking (published in Science Advances, and summarised in a Twitter thread here). Of the 2,478 caterpillars that I counted, the majority were from unlit areas. Lighting reduced the numbers by between a half and one third. The lit areas, which had been illuminated for at least five years, had almost universally had lower numbers of caterpillars.
This complements a previous study from the Netherlands that found that experimental lighting installed in nature reserves drove down the numbers of adult moths after several years.
I’d previously estimated that 100 metres of (unlit) hawthorn hedge at one of my study sites was home to 21,000 caterpillars. This particular community of caterpillars is vital food for songbirds (e.g., blue tits, which alone have been estimated to feed 35 billion caterpillars to feed their chicks in Britain each spring), as well as hedgehogs, predatory insects, and parasitoids. If I run the same back-of-the-envelope calculation for the section at this same site lit by white LEDs (only 120 metres away along the road), this had ‘only’ about 8,600 caterpillars. That may still seem like a lot. But it’s a huge reduction in the prey available in these areas illuminated by streetlights.
A devastating driver or an indistinct influence?
It is clear from our results that lighting can be a very important determinant of insect populations, at least at local scales. The next logical question is a little trickier: how far can light pollution go to explain insect declines?
This is a question that we began to tackle in our study. I used those spatial streetlight datasets to produce estimates for the total area within the study region that is directly lit by streetlights. This turned out to be extremely small, especially for more rural habitats (e.g., 0.23% of arable and 0.68% of broadleaved woodlands were directly lit). With so much of the wider countryside being unaffected by direct lighting, can this factor really explain national-scale insect declines?
Probably not. But that does not exclude the possibility that light pollution has a contributory role, at least.
As I mentioned earlier, the area affected by lighting is growing year-on-year. This is something I experienced first-hand at several of my field sites. As more of the countryside becomes urbanised, those percentages are only going to creep up. It is also worth noting that streetlights are not the only cause of light pollution (though they are the principal source of direct lighting in rural areas).
No-one is suggesting that light pollution is the only cause of insect declines.
In a special issue in the journal PNAS, David Wagner and colleagues described contemporary insect declines as ‘a death by a thousand cuts’. While a tad melodramatic, it serves as a useful reminder that insect declines will be the manifestation of many, often intertwined, factors. The relative importance of these different causes will vary for different insect groups, across geographical areas, and also act unevenly through time.
But there is something that sets light pollution apart from most of other biodiversity threats: it should be really quite simple to solve. The impacts of lighting can be minimised relatively easily and cheaply. It is a tractable problem.
So, whether light pollution is up there with the most frequently recognised causes of insects declines, or if it has a minor effect in most situations, this does not necessarily matter. Insects are in trouble. And we should be doing all we can.
In the cold light of night
As mentioned earlier, LEDs are being embraced globally thanks to their superior energy efficiency. In practice, this typically means that white LEDs are taking the place of traditional sodium lighting that has lit our roads for many decades (with its characteristic warm, yellow or orange hues). In the UK, 55% of streetlights are now LED. In a couple of decades, I suspect the switch will be largely complete, with almost all our streetlights being LEDs.
The evidence from our new study (where we compared the impacts of older, yellow-ish sodium lighting with newer white LEDs) is that this ongoing switch is likely to be harmful to insects. This is in line with physiological predictions: insects are well-known to be more sensitive to bluer wavelengths of light (which are largely absent from sodium lighting).
And it’s not just insects.
Biologically, the colour of light (i.e., the wavelengths that are emitted) is extremely important for determining exactly how light affects a whole bunch of biological processes. The trend towards outdoor lighting emitting light over a wider range of wavelengths (often producing cooler white light) is predicted to have more negative impacts for a wide range of different organisms (including humans).
So, despite being initially hailed as a beacon of sustainability, LEDs arguably have been something of an environmental disaster. Thanks to the rebound effect, as discussed earlier, and also this shift towards whiter lighting.
But I shouldn’t get too carried away with hating on LEDs. They have a trump card.
LEDs are far more flexible than traditional lighting types. They can be easily dimmed, integrated with motion sensors, and made into ‘intelligent street lighting‘. The first white LEDs produced a very harsh, cold light. But technological advances mean that different ‘colour temperatures’ are now available. For instance, you can opt for warm white lighting in your bedroom, while having cool white in the kitchen. We would expect that white LEDs with reduced blue wavelengths (so a warmer colour temperature) would be better for insects. More studies are needed to test whether this is actually the case. It seems pretty unlikely that they are going to be worse, though.
And LEDs don’t have to be white, of course.
In some places, entire streets have been bathed in red LED light. These novel ‘red-light districts’ have actually been created for certain bats which can only tolerate red light. Similar red lighting is used in some beachfront areas to try to help prevent confused newly hatched turtles from setting off in completely the wrong direction to their inevitable doom, instead of heading for the ocean (as featured in a harrowing Planet Earth II footage).
For insects, there are many studies that have shown that red LEDs do successfully ease the disruptive impacts of night lighting. But there are also some other cases where red LEDs can be equally harmful as white light. ‘Red light districts’ replacing white LEDs would likely benefit insects. But they would still not be perfect.
The best option for wildlife is to reduce both the extent and intensity of lighting as much as possible.
Bringing back the night
Britain’s policy of austerity showed that significant changes in lighting are possible (sometimes even tolerable with the public, with some studies finding no evidence of negative consequences). Unnecessary lights can be switched off completely, while others can be dimmed for the early hours when few people are awake.
When the lights go out, our wildlife should recover almost immediately.
Another approach is shielding lights, which can easily be added to minimise spillover into adjacent habitats (and indeed into people’s bedroom windows).
But we mustn’t forget that there are good reasons why street lighting exists in the first place. Lighting can benefit society, by reducing vehicle collisions, cutting crime, and increasing feelings of safety.
The safety of women at night has been propelled to the fore this year, following the tragic murder of Sarah Everard. In response to the case, the government has said it will increase funding for street lighting, while several petitions urging for street lighting to be switched back on have gained significant traction. Clearly, decisions about public lighting need to actively involve a cross-section of society, with greater female representation (of course lighting isn’t the core issue here, and far more needs to be done to stem male violence against women).
In other respects, lighting can be bad for people, especially when it’s poorly designed.
Astronomers have long noted that lighting is obscuring our view of the stars. As we touched on earlier, there is some evidence that night lighting can negatively impact human health. Bright streetlights can also create dark shadows, which might actually increase crime and reduce personal safety. In the UK alone, streetlights cost the taxpayer between 160 and 430 million pounds each year (estimated from figures here).
We need more research to understand the harmful impacts of artificial light at night (both on people and ecosystems). But we also need more robust evidence on the societal benefits that lighting can offer us.
A better understanding of the pros and cons of streetlights will mean we accurately trade-off them off. This will allow us to find lasting solutions that maximise the benefits of lighting for people, while simultaneously minimising the harms to insect populations and ecosystems.
Note: rejigged parts of this article are published as a piece in The Conversation. You can hear me discussing the Science Advances paper in a 12-minute interview for Science Friday.
Great article, thanks for your important work!!
I have a question that I often wonder about. Since lights are being mostly strobes, just fast enough so that our eye would not notice, wouldn’t that be something affecting other species as well? I guess so since each species perceives its environment differently. I guess some see strobes instead of light. Do you know if this is the case?
It is my understanding (for UK LED street lighting anyhow) street lighting is not dimmed via pulse width modulation (strobe) but by simply reducing DC current to the LEDs ( they are powered via a DC supply called a driver)
Thanks, Paul! Not sure about the flicking I’m afraid, but thanks Alex for pitching in here.
The point that strobing (invisible to us) could affect other species is an important one, and something that I’m not aware has been studied.
We, the humans should by all means not attempt to interfere with insect’s habitats and also not with their day-night cycle. And we should not trap them by using white lighting. Because bright white LED lights attracts them in large numbers. There are also predators sitting around on these lamps, like spiders with their large nets. Pale orange lights instead would attract much fewer insects as these are not that sensitive to pale orange colours. So, the solution is not broadband-white (because of high portion of green and blue) but rather a narrowband-lighting such as sodium lamps or even LED lighting imitating the narrowband spectrum. This is feasible by using respective dotting, filtering / colouring of LED encasing/dense orange glass in front of white LED arrays.