There’s no thermometer for insects

You can’t count every insect on Earth. There are too many, they’re too small, and most of them live and die unobserved. So scientists do the next best thing: they pool hundreds of long-running surveys — moth traps, beetle pitfalls, pollinator transects — into a single index. And that index tells a consistent story. Across the world’s land, terrestrial insects are vanishing at roughly 9% per decade¹.

Nine percent doesn’t sound like much. Compounded, it’s brutal: at that rate, half the insects alive in 1990 are gone by the 2050s. The most-studied datasets are worse still — German nature reserves lost 76% of their flying-insect biomass in just 27 years² — and one global review estimates 40% of insect species face extinction in the coming decades³.

The number that broke through: Krefeld, 76%

In October 2017, a paper landed in PLOS ONE that turned a niche entomological dataset into a global news story. A team led by Caspar Hallmann at Radboud University, working with the volunteer-run Entomological Society Krefeld, reported that the total biomass of flying insects in 63 German nature reserves had fallen by more than 75% over 27 years of continuous monitoring².

The decline was 76% on average, 82% in midsummer — peak insect season — when the absolute losses are largest. The paper was authored by professional ecologists but the dataset was collected by volunteer entomologists working through the Krefeld Society. Citizen science had quietly produced one of the most consequential biodiversity datasets of the century.

Three things made it land so hard. First, the sheer length of it — 27 years of continuous data, collected the same way in the same places, where most insect studies last a season or two. Second, where it was measured: these were nature reserves, legally protected from the obvious culprits, and the insects were vanishing inside them anyway. And the scale was simply hard to absorb — a 76% collapse in less than three decades, the kind of curve that ends in an empty net within a single human generation.

The number that put it in perspective: Van Klink, 9% per decade

The Krefeld finding was alarming and locally extreme. The natural follow-up question was: is this happening everywhere?

In April 2020, Roel van Klink and colleagues published the most comprehensive answer to date in Science¹. They pooled 166 long-term insect monitoring studies from 1,676 sites worldwide and found that terrestrial insects are declining at about 0.92% a year — roughly 9% per decade. Slower than Krefeld’s headline, but global, and grim. Curiously, freshwater insects told the opposite story: their numbers are rising in many regions, a quiet dividend of the clean-water laws that scrubbed rivers across Europe and North America since the 1970s. The terrestrial declines themselves were steepest in those same well-studied regions, while the tropics — where most of the world’s insects actually live — remain badly under-sampled.

Put the two findings side by side and they don’t contradict so much as nest. The German reserves lost 76% of their flying-insect biomass in 27 years; the global terrestrial average is nearer 9% a decade; freshwater insects are recovering in places. All three are true at once — a savage local extreme, a serious global trend, and a reminder that the damage can be reversed where we choose to reverse it. This page tracks the global terrestrial figure, because that’s the number that belongs to the whole world, and keeps Krefeld in view as the study that first rang the alarm.

Pavement, pesticides, and the dark

The Krefeld team documented the decline but could not pin it on a single cause. They ruled out the obvious explanations within their dataset: land use inside the reserves, weather, habitat type. What they couldn’t rule out — because the data didn’t reach beyond the reserve boundaries — was the landscape context.

The 2019 Sánchez-Bayo and Wyckhuys global review ranked the drivers and reached a stark conclusion: on the current trajectory, 40% of insect species worldwide face extinction in the coming decades³. Agricultural intensification leads the list — neonicotinoid pesticides, fertiliser runoff, the spread of monocultures across what used to be marginal habitat. Behind it come habitat loss and fragmentation, which can render even a protected reserve unviable once it shrinks below a critical size, and a warming climate that scrambles the timing of emergence and breeding. Subtler pressures pile on top: light pollution that disorients night-flying insects, and the steady creep of invasive species and disease.

Pull this thread, the web unravels

It is hard to overstate how much rests on insects. They pollinate roughly 80% of the world’s wild plant species², and the crop pollination they perform on top of that — bees and their relatives — is worth somewhere between $235 and $577 billion to global agriculture every year. They are also the base of the food chain: about 60% of bird species eat them², which is no small part of why North America has lost three billion birds since 1970 — the species vanishing fastest are the insectivores. And out of sight, beetles, ants, and flies do the slow work of turning dead matter back into soil.

A flying-insect collapse, then, is never just an insect problem. Pull the thread and the birds, the wildflowers, and the soil itself come loose with it.

Reading the line forward

The line on the counter heads toward an index of about 50 by 2050 — half the insect abundance of 1990. That figure isn’t published by Van Klink; it’s their measured 0.92%/year rate continued forward from a 1990 baseline, nudged slightly steeper to reflect Sánchez-Bayo’s trajectory of accelerating species-level extinction risk.

What this projection assumes:

• The rate measured 1990–2020 continues at roughly the same pace
• No major policy intervention bends the curve
• The decline doesn’t accelerate beyond linear (climate stress, pesticide intensification could push it steeper; freshwater-style policy recovery could slow it)

We say so plainly in the counter spec. The 2050 number is where current trends point; it’s not a forecast.