20141010_ML

Source: BBC World Service

URL: http://www.bbc.co.uk/programmes/p027tqp4

Date: 10/10/2014

Event: More or Less: Species in Decline?

Credit: BBC World Service

People:

• Professor Steve Buckland: Professor of Biostatistics, St. Andrews University
• Louise McRae: Project manager, Living Planet Index
• Wesley Stephenson: Broadcast journalist, BBC News

Wesley Stephenson: Hello, this is More or Less on the BBC World Service, I'm Wesley Stephenson. Let's take a trip into the jungle... And if the WWF and the Zoological Society of London are right, there are now fewer animals here than there were 40 years ago. And they've fallen by an average rate of 50%. In fact, it's not just in the jungles but all over the world. This news prompted startling headlines, and it also got the More or Less inbox buzzing. We've had a lot of emails like this one from Steve Welch:

Steve Welch: The headline figure of wildlife falling by half seemed immediately suspicious, and I see from the text that only some types of animals are included. For example, there are no insects. But what does falling by an average of 52% mean? Please enlighten us.

Wesley Stephenson: More or Less listeners are a sceptical lot, so let's have a look at what the figure means and how it was arrived at. It came from the recently released Living Planet Index, and the project manager is Louise McRae from the Zoological Society of London.

Louise McRae: The exact figure is 52% and this is a decline in the size of vertebrate populations over 40 years. The way we came to this figure was that we gathered over 10,000 different population trends from over 3,000 species, and what we've done is we've calculated an average of these trends to get this figure. Now what it means is the rate of decline, so it means that populations have roughly halved, over this period.

Wesley Stephenson: So does it mean that if you counted all the vertebrates in 1970 and counted them all in 2010, the number is 50% less?

Louise McRae: It's not quite the same - so it's not that we've lost half of all individual animals, because some of the populations are very large and hundreds of thousands of individuals, and some are very small. So it's all relative, that it just means that the rate at which populations are declining is that they've halved, on average, over that time.

Wesley Stephenson: Let's imagine a scenario. You have two populations, one of frogs and one of birds. And let's say, in 1970 there were 10 frogs and 100 birds, giving a total of 110 animals. By 2010, the frogs had declined by 80% and the birds by 20%, leaving two frogs and 80 birds, a total of 82 animals. The average decline of these two populations is 50%, but the total number of animals hasn't halved - which is probably not what most people would have understood from the headlines.

Another thing about the index, as loyal listener Steve points out, is this is just an index of vertebrates and doesn't include insects or plants, and it's not as if either the WWF or the Zoological Society of London have actually been out to count the animals themselves. This is a figure modelled using available data, some of which is patchy.

Louise McRae: Our data set is a sample of all the vertebrates that are out there in the world. Now, unfortunately this isn't a representative sample of what species that we would expect to see - this is because people, historically, have monitored species such as birds and mammals, which are very easy to count, and in places like Europe and North America these monitoring schemes have been going on for a very long time, so we have a lot of data from these regions. Now, most of biodiversity is actually found in tropical areas, so we were trying to balance the data set, so that we can actually get an accurate picture of what is actually going on. The way we did this was to weight the data, so we're placing more weight on where we see more species.

Wesley Stephenson: But also you're placing more weight on those areas where the data isn't as - as rich as it is, for example, in Europe and other parts of the world.

Louise McRae: Yeah, that was a concern when we started thinking about this method, but what we did, when we saw that we didn't feel we had enough data to make this robust change, we combined the data together. So, for example, for amphibians and reptiles we didn't have enough data to weight them individually so we grouped them into a group on their own. So that was a concern but we addressed it by combining data into sub-groups.

Wesley Stephenson: Now there's nothing wrong with modelling the figures, and Louise and her colleagues feel confident that the way they've extrapolated the data has given them a pretty solid figure, with a margin of error of 10 percentage points either way. But there are some who aren't so confident that the figure really tells us what they think it's telling us. The weighting in favour of tropical areas, where there's less data, has caused concern for people like Professor Steve Buckland, Professor of Biostatistics at St. Andrews University in Scotland, and whose models are used in calculating the figure.

Steve Buckland: Many of the surveys carried out in temperate areas - that's non-tropical areas, like Europe - are done to a very high standard, very good quality. But in temperate areas the biodiversity trends are reasonable encouraging. The problem is the tropical areas account for most biodiversity in the world, and both the quality of time series and the trends tend to be less good in tropical areas.

Wesley Stephenson: And so what's your concern about those?

Steve Buckland: Firstly, the index relies on those populations that are monitored, which are not necessarily representative of populations in total. There are all sorts of reasons for monitoring populations - one of them is that a species, or a population of a species, is thought to be under threat and so a monitoring programme is instigated. Species or populations that are more impacted by man's activities tend to be easier to monitor, also, for example, rainforest populations that are under threat tend to be rainforest populations where there's activity by man. Areas of rainforests that are very inaccessible and remote, we don't have the time series of abundance estimates for those locations, and so when you increase the weight of those monitored populations in tropical areas, the rate of decline increases quite substantially from the old index.

Wesley Stephenson: And does it overplay that, do you think?

Steve Buckland: I think it probably does, simply on the basis that the populations being monitored are not representative. If they were representative of tropical and temperate populations as a whole, then that would be a correct thing to do. But the difficulty is that populations that are not in trouble are less likely to be monitored. Now they've done some assessment of whether there's bias towards surveying threatened species, and they concluded there wasn't evidence of bias from that source. But nevertheless, there probably is bias towards species that show decline, rather than species that are largely unaffected.

Wesley Stephenson: What do you think the 52% figure tells us, then?

Steve Buckland: It tells us that the populations being monitored are in real trouble, and that is a very large number of populations. What it doesn't tell us is that 52% of all vertebrates - animals on the planet - have declined by 52% in the last 40 years.

Wesley Stephenson: So the figure probably isn't as clear-cut as the headlines or the press releases from the WWF and the Zoological Society of London suggest. But Steve Buckland pointed out to me that to caveat the figure in lots of detail would have obscured the message that the WWF and the Zoological Society were trying to get across, namely that man is having a serious impact on wildlife. But as climate change scientists have found in the past, trying to exaggerate your message - even if you believe in the cause - can backfire, and can make people more sceptical about future claims.

That's all for this week - thank you very much for listening. If you have any numbers that you want us to investigate, please email moreorless@bbc.co.uk. You can also listen to other editions of More or Less, exploring subjects from the chances of someone running a sub 2-hour marathon to whether the UK is really poorer than most of the states in America. And you can do this by going to our website: bbcworldservice.com/moreorless.