20190428_ML

Source: BBC Radio 4

URL: N/A

Date: 28/04/2019

Event: More or Less: Insectageddon

Credit: BBC Radio 4

People:

Professor Casper Albers: Statistician, University of Groningen
Ruth Alexander: Journalist, BBC News
Tim Harford: Presenter, BBC Radio 4's More or Less programme
Professor Jane Hill: Professor of Biology, University of York
Dr. Francisco Sánchez-Bayo: Environmental scientist
Professor Jessica Ware: Entomologist, Rutgers University

Tim Harford: Is "Insectageddon" heading our way? Recently, we heard this...

Woman's voice: There could be no insects left on Earth in 100 years, according to the first global scientific review...

Man's voice: ... that the world is truly under threat, scientists say, from a catastrophic collapse of Nature's ecosystems.

Tim Harford: But some loyal More or Less listeners questioned the findings.

Woman's voice: Widespread reports of insects in decline by 2 and half percent per annum - how can anyone estimate this? Has someone counted them all? If so, how? Is this true?

Woman's voice: All insects to become extinct in 100 years? Seems to me to be too fantastical to be remotely likely.

Tim Harford: Oh, our listeners... So loyal but so sceptical. But are they on to something? Here with us is our Insectageddon reporter, Ruth Alexander - hello, Ruth.

Ruth Alexander: Hello, Tim. Let's start at the beginning, shall we? This new piece of research is a meta-analysis by two authors, Francisco Sánchez-Bayo and Kris Wyckhuys. Now, they haven't counted any insects themselves but they have instead analysed lots of studies by people who have indeed gone out and counted insects. And the way they chose these studies is quite interesting. They did a search on a scientific database, using terms such as "insect", "survey" and "decline". But Casper Albers, a statistician from the University of Groningen in the Netherlands, questions this method.

Casper Albers: So what they did is they only looked for papers that had the keyword "decline" in the title or abstract. It could be that there are also papers that actually found an increase, but if you look for this keyword you won't find those papers. So they kind of introduced a bias in the way they looked for the papers.

Ruth Alexander: We asked Francisco Sánchez-Bayo why he'd only looked for the word "decline" - perhaps there are other studies showing an increase, a recovery or stability. But he seems to have simply assumed it wasn't worth looking for those terms.

Francisco Sánchez-Bayo: We typed those keywords - "decline". But I must say all the surveys report declines, stable populations and increases - all of them. Because that's what happens with every - every group. The majority of the surveys, of the 73, I can say properly 50 were taken from references in the other surveys. We not only looked at declines but also even the papers that talked about declines talked about increases and talked about stable populations, so it's the whole lot.

Ruth Alexander: Now if you want to say something about the global insect population, you want to cover the whole globe, but unfortunately not every area is equally studied - 60 out of the 73 studies examined insects in North America and Europe. The analysis includes more studies on insects in the UK than in the whole of Africa, Asia, South America and Australasia put together - that's two thirds of the globe's landmass. China and Australia, for instance, only included studies of managed honey bees.

Tim Harford: Hmm. I spot a problem here, then. If you're trying to come up with a global figure, you're not going to get a fair result by looking at this very narrow set of countries.

Ruth Alexander: Exactly. And an insect expert, a Professor of Biology at the University of York, Jane Hill, agrees with you.

Jane Hill: Yeah, well, I think that is an inappropriate thing to do, and if you look at where they got their studies from, they can't really extrapolate globally because most of the studies are from Europe. And so the concern, I think, that I have with the paper is about extrapolating widely beyond the studies they examined.

Ruth Alexander: The study acknowledges the issue, but author Francisco Sánchez-Bayo is convinced that they can still extrapolate to the rest of the world.

Francisco Sánchez-Bayo: So these predictions apply particularly to the developed countries, because that's were they come from, but we have no qualms whatsoever in having extrapolated that to the rest of the world, because we see that the rest of the world follows the same trend and it's caused by the same drivers.

Tim Harford: Hmm, maybe. But we do seem to be going well beyond what the data actually show, here.

Ruth Alexander: Yes. Now, let's get to the big claim that the study makes.

Dramatic male voice: Insect Armageddon.

Ruth Alexander: All the headlines were based around the claim that insects were dying out at 2.5% per year and could be extinct in a hundred years' time. Now, to understand where this claim comes from, you need to understand the different ways you can count insects in a study. You can try to count one specific type of insect - say, a honey bee - or you could try to count all of the insects in a particular area, using a trap, and then you can weigh them to come up with what's called a biomass figure. The 2.5% figure and the scary Armageddon headlines all come from these biomass studies.

Tim Harford: How many of these studies are we talking about?

Ruth Alexander: Three.

Dramatic male voice: Insect Armageddon, in three studies.

Ruth Alexander: Lead author Francisco Sánchez-Bayo.

Francisco Sánchez-Bayo: We have three studies that we added to the surveys, and these studies deal with the loss of biomass. So those three studies have been done in Germany, the UK and Puerto Rico, and the three of them came to the same conclusion, that the biomass loss in the last 30 years or so has been 2 and a half percent per year. So that's a shocking number. So it's so consistent and for a tropical country like Puerto Rico or the ones in Europe, right? And it's practically the same.

Tim Harford: I don't know - if you're relying on just three studies for your global Armageddon, you need to be pretty sure the studies are gold-standard - you'd want the decline to be backed up by measurements taken in the same way for a large number of years.

Ruth Alexander: Yes, and unfortunately, that's not the case. The Puerto Rico study looks at just two periods - around 1976 and 2012. This isn't showing a consistent decline but simply giving two snapshots of what was happening in the area. Professor Jane Hill thinks this is problematic.

Jane Hill: Comparing two datasets from several decades apart, which are essentially snapshots, is that we don't really know what's been going on between those time periods, and we know that the populations of insects - the numbers that you find at a site - can go up and down quite a lot, depending on the local weather conditions.

Ruth Alexander: There just aren't enough solid numbers here. Insect expert - they're called entomologists, you know - dragonfly adorer and Professor at Rutgers University in the US, Jessica Ware.

Jessica Ware: The data is patchy, and it's not the fault of any of the authors, that were all individual entomologists or evolutionary biologists. And we have paid up a small pot of money to measure the total number of mosquitoes in Australia for a two-month period.

Tim Harford: Okay, so the data are patchy, but is the overall trend correctly tracked, here? Is there basically a decline?

Ruth Alexander: Well yes, despite the limits, you will find that the general consensus is that many types of insects are in decline. But the so-called "Insect Armageddon" isn't supported by the data, according to Jessica Ware and statistician Casper Albers.

Jessica Ware: As a scientist, I would say: no, we can't make that claim. We don't have enough data.

Casper Albers: They simply don't have that proof.

Ruth Alexander: We put these criticisms to Francisco Sánchez-Bayo, but he says that even if they don't have the data to prove their claim statistically, that doesn't mean they shouldn't make the claim.

Francisco Sánchez-Bayo: Even if we don't have enough data to prove it statistically, whatever, we know that this is happening. So it's better to do it now than ten years later when we may have more serious problems.

Tim Harford: Insectageddon correspondent, Ruth Alexander. Thank you, Ruth.