To save insects, we need to understand their decline

Recently many studies have revealed how badly insect populations are faring.

This year one from the United States found that butterfly numbers had dropped by 22% over the past two decades. A second from the United Kingdom reported that bumblebees had their worst year since records began.

Research like this is important for showing that insect numbers are declining, but they aren’t designed to pinpoint why it’s happening. This makes it challenging to know how to stop the declines. Professor Andy Purvis, a researcher at the Natural History Museum, is part of a team that’s hoping to change this.

“Ideally, we would have been doing regular stock takes of insects everywhere over the last few decades, and put simply we haven’t been,” says Andy. “Nearly all of the places where we do long-term monitoring of insects are places that were already heavily modified ecosystems when we started looking.”

“And most of the other places we’ve been monitoring are those that are protected from modification, so they’re also giving us a biased picture of what’s going on.”

This doesn’t mean that these studies are wrong, but they simply don’t give us true before and after snapshots. This means we can’t be certain what effect different impacts such as land-use change, climate change or pesticides are having on insect numbers. It also makes it almost impossible to understand how these impacts might be combining to exacerbate each other.

Andy and his colleagues have been working on a framework that can take in all of the data we currently have about why insects decline and overlay this with our understandings of potential drivers of these. From there we can then hopefully predict how insect populations might react in the future, as threats worsen or conservation steps up.

Their first few steps in laying out this framework is published in the journal Science.

Why are insect populations declining?

Figuring out what’s going on with insect populations is not easy. Insects are typically small, neglected animals which despite their importance have historically flown under the radar.

But studies that have looked at their numbers paint a mixed and worrying picture. For example, the UK charity Bumblebee Conservation Trust found that 2024 was the worst year for bumblebees since records began. Their research showed that on average the 24 species of bumblebee in Great Britain declined by 22.5%, with some species dropping by a massive 39%.

On the other side of the Atlantic, a similar scene is playing out with butterflies. A new study looking at these charismatic insects found that over the past 20 years their numbers had declined by 22% on average. Some species, such as the Florida white, are seemingly on the brink of local extinction.

“This is troubling because insects account for about 90% of the world’s animal species as far as we know”, explains Andy. “On land and in freshwater they dominate wild animal biomass.”

“This means they’re hugely important in how ecosystems work, and given that we depend critically on ecosystems and what they provide us with for our quality of life we should know how they’re doing.”

But knowing that insects are declining is only one part of the story. We also need to know why they are declining and how they will respond in a changing future. We’re currently lacking this data and time is running out.

According to Andy, there are two ways to deal with this problem. “One is to start doing the regular kind of stock take that we should have been doing all this time with lots of experiments to understand how human impacts affect the insects,” says Andy. “Great, we could and probably should do that. But we won’t know the answer for decades when we need to know the answer now.”

“Instead, we’ve got to be a bit more creative.”

Predicting threats to insects

The general idea of the team’s work is to produce an evidence-based model that will accurately say how different insect populations in different parts of the world will react to various threats. It’ll work for an individual threat but also, more importantly, show what will happen when threats are combined.

This is no simple task. The team are taking different strands of the story from thousands of published papers. Some provide simple number counts of insects, others controlled experiments, there’s expert testimony from scientists who are studying different groups, and raw data. They are then trying to bring these disparate sources together to give a much more holistic idea of insect populations.

This data can then be combined with the already highly accurate data that exists on, for example, climate and land-use change.

“The project’s been going for about three years now, and we’re starting to fit together the bits of the jigsaw,” explains Andy. “We’d love to have more pieces, but we’re more convinced than ever that this is a valid and potentially powerful approach to understanding and tackling these declines, and that it is very much preferable to not trying.”

The next step will be to use the model to predict past trends. This will allow the team to use previously published papers on insect numbers to check if the predictions are right.

Then, it’ll simply be a matter of time. They’ll use the new model to start predicting how insect populations will react in the future to, for example, rising temperatures or habitat change. This will ultimately then be proved right or wrong.

“We’re putting together the framework, and we’re starting to use it to make predictions,” says Andy. “Then we’ll use the bits that pass that test to predict what’s going to happen in the future. And then the future will come along and kick us when we’re wrong. That’s how models improve.”