Anyone who drove long distances in the 1980s may remember stopping to scrape insects from the windscreen — and noticing, years later, that it was no longer necessary. Standardised trapping across German nature reserves found flying insect biomass had fallen 76 percent in just 27 years.

Every family car once carried a rag with exactly one job, and it was not for polishing.

Anyone who drove long distances in the 1980s may remember the ritual. Pull into a petrol station, soak the squeegee, and attack the windscreen with the grim energy of someone scrubbing a barbecue plate. Moths, aphids, hoverflies, the occasional beetle that arrived like a thrown pebble. Over the following two decades, that chore quietly stopped being necessary.

Researchers call it the windscreen phenomenon.

What the German traps weighed

Volunteers from the Krefeld Entomological Society spent 27 years running Malaise traps, which are tent-shaped mesh structures that intercept whatever flies into them, across 63 protected areas in Germany. Catches were weighed rather than sorted species by species, sidestepping arguments about who is good at telling one small brown fly from another.

Caspar Hallmann and colleagues at Radboud University ran the analysis, and their paper in PLOS ONE reported flying insect biomass down 76 per cent over the period, and down 82 per cent at the height of summer. Weather, land use and habitat change were all tested as explanations. None of them accounted for the fall.

This is one study, in one country, inside reserves that were supposed to be the safe bit. It cannot speak for the rest of the planet.

The aerodynamics objection

“The aerodynamics of a refrigerator, and these days it stays clean.” That was German entomologist Martin Sorg on his Land Rover, talking to Gretchen Vogel of Science, whose 2017 report took the obvious rebuttal seriously. Modern cars are slipperier, the argument runs, so bugs ride the airflow over the roof instead of ending up smeared across the glass.

It is a fair thing to wonder about, and it has a tidy workaround.

Nobody has ever tried to make a number plate aerodynamic. It is a flat rectangle bolted to the front of a car, doing the same blunt job in 2026 that it did in 1986.

Kent Wildlife Trust went at the question head on, recruiting classic car owners so that the vehicles in its 2019 survey ranged in age from 1957 to 2018. Paul Tinsley-Marshall and his co-authors found the relationship running the opposite way to the one everybody assumed. Newer cars squashed slightly more invertebrates than older ones, a small effect but a statistically significant one, though the team wants more vintage vehicles in the sample before anyone leans on it.

Counting splats on number plates

Drive somewhere, then count what is stuck to the plate when you stop. That is the entire method. The Royal Society for the Protection of Birds ran it nationally in 2004, and the Kent repeat 15 years later, badged as Bugs Matter, found splat density roughly halved, from about 0.2 insects per mile to 0.1.

Since 2021 the survey has run every summer through a phone app, with drivers photographing their plates at the end of a journey. Results for 2025 were released by the invertebrate charity Buglife. Drawn from more than 25,000 journeys, they put splat rates 59 per cent lower across the UK than in 2021, an average annual fall of 19 per cent. Andrew Whitehouse of Buglife pointed out that a warm spring and a hot summer should have given insect numbers a lift, and that the data showed a fifth straight year of decline instead.

Kent is faring slightly worse than the national average. Kent Wildlife Trust puts the county’s annual average drop at 23 per cent since 2021, drawn from nearly 2,000 journeys.

Take those figures seriously, but take them carefully. The trust says plainly on its own project page that the survey has not run for long enough to decide whether the numbers add up to a trend. Citizen science buys volume, and volume comes bundled with noise: different drivers, different roads, different speeds, and five summers of weather that refuses to repeat itself.

What is driving the decline

So what is doing the killing? No single cause explains it, which is part of what makes the problem so awkward to legislate against.

The most systematic attempt at unpicking it came from Sebastian Seibold and a large German team, who sampled 150 grassland and 140 forest sites between 2008 and 2017. Writing in Nature, they reported grassland arthropod biomass down by 67 per cent in a single decade, with individual numbers falling 78 per cent and species counts by a third.

The telling detail was where the steepest losses landed. Sites ringed by more farmland declined faster, regardless of how carefully that particular meadow or woodland was managed. That shifts the blame outward, from the individual field to the whole landscape. An organic paddock surrounded by intensive agriculture is an island, and islands lose species.

Add pesticides, fertiliser runoff, hedgerow removal, street lighting and a warming climate, and you have a list of suspects rather than a culprit.

What a plate can and cannot tell you

Splats are a crude instrument. They register mass and motion, nothing else. A plate will tell you that fewer things are hitting the car, and it will not tell you which ones stopped turning up, because a hoverfly and a honeybee leave much the same mark.

Sorting out which insects have gone, and from where, still needs traps, nets, ethanol jars and people who can identify a wasp on sight. Frustratingly slow work, and considerably less fun than driving around counting smears.

The plate survey is growing regardless. It crosses the Channel in April 2026, running in France alongside the UK and the Republic of Ireland, on the reasoning that a hoverfly heading north out of Normandy pays no attention to the border. The season also opens a month earlier, which should catch spring insects that earlier rounds missed.

Drivers noticed this for decades and wrote none of it down. That part has been fixed.