[Jingle]
Interviewer: Adam Levy
Hello, and welcome back. This week we’re looking at how many lives are cut short by diesel emissions around the world.
Interviewer: Kerri Smith
And how a few error prone bots can help humans work together.
Interviewer: Adam Levy
We’ll also take a look at how caterpillars do digestion differently. This is the Nature Podcastfor May the 18th2017. I’m Adam Levy.
Interviewer: Kerri Smith
And I’m Kerri Smith.
[Jingle]
Interviewer: Kerri Smith
Inside your gut and mine, and every human’s is a bustling community of millions of microbes: our microbiome. This community of bacteria helps us digest food. They chase away bad bugs, and they’re also a high fashion choice of research topic.
Interviewer: Toby Hammer
Microbiomes are really hot right now.
Interviewer: Kerri Smith
Toby Hammer is at the University of Colorado in Boulder.
Interviewer: Toby Hammer
Almost every day there’s a new study showing how the microbiome can be super important to a particular aspect of animal biology. Even my mum, my grandma, have read about the microbiome.
Interviewer: Kerri Smith
Interviewer: Toby Hammer
Going into the project, I was really interested in why different species of caterpillar would maybe have different microbiomes and how that relates to what plants they feed on. I was expecting them to have a normal microbiome like every other animal that I was aware of.
Interviewer: Kerri Smith
But when Toby and his colleagues looked at their samples, a normal microbiome is not what they found.
Interviewee: Toby Hammer
The more samples I ran, the more it seemed like the main result would actually be that they don’t have much of a gut microbiome. So it was a surprise to me.
Interviewer: Kerri Smith
This is not what Toby expected. Nowadays, he says, everyone just assumes that every animal has a microbiome. Well, almost everyone assumes this.
Interviewee: Jon Sanders
My name is Jon Sanders. I’m a postdoc with Rob Knight at UC San Diego.
Interviewer: Kerri Smith
Jon Sanders studies the evolution of microbiomes and their hosts and has a particular soft spot for ants. He wasn’t surprised to see Toby’s results because his own experiments point the same way. Back when he was doing his PhD, he stuck a few ants under the microscope…
Interviewee: Jon Sanders
… And sort of by chance I happened to look at a couple of other ants we had lying around the lab and I thought that I was messing up somehow because I didn’t see any bacteria in those ant guts at all. To me it was just very surprising. I sort of just assumed there would always be bacteria in guts.
Interviewer: Kerri Smith
Being less reliant on a team of bugs could be good news for a caterpillar. After all, Toby says, their main goal is to pile on simple nutrients as fast as they can before their fleeting adult life begins. The human microbiome helps us digest our more complex diet but maybe a mass of bugs would hinder a grub nibbling on plants.
Interviewee: Jon Sanders
The caterpillar stage is when they acquire the vast majority and often all the nutrients that they need for growth and for reproduction, so it’s a really critical stage in their life cycle. And for a lot of groups they don’t even feed at all as adults. Any microbial growth that’s harvesting those nutrients, then you’re essentially competing with microbes in your gut for nutrients. So, if you can exclude microbial growth, then you don’t need to share those nutrients with your microbiome. You can keep them all to yourself.
Interviewer: Kerri Smith
Knowing more about the bugs’ own bugs could help scientists working in pest control. Caterpillars can damage crops such as maize and fruit trees and stored foods such as grains. One important chemical actually acts on the gut.
Interviewee: Jon Sanders
This Bt pesticide that is so widely used in agriculture actually seems to work, not by direct toxicity on the insects themselves, on the caterpillars, but sort of by making their guts leaky and allowing bacteria that happened to be there to kind of get into the host’s bloodstream, essentially, and take over and kill the insect that way.
Interviewer: Kerri Smith
Unfortunately, the new finding might mean pesticides like this wouldn’t work so well on caterpillars without bacteria loaded guts. But, says Toby…
Interviewer: Toby Hammer
I think it’s at least important in understanding these pests and maybe the lack of a gut microbiome could inform new strategies for controlling caterpillars.
Interviewer: Kerri Smith
When Toby looked back through the old literature on caterpillars, he did find some reports that they were missing a microbial community.
Interviewer: Toby Hammer
Before this current microbiome era, it wasn’t particularly interesting if an animal didn’t have a microbiome and what’s changed is, I think, our expectation that that’s the norm, that everything has a microbiome and lots of aspects of animal biology are made up by microbes.
[Music]
Interviewer: Kerri Smith
That was PhD student Toby Hammer of the University of Colorado. His research report is available on http://biorxiv.org/. That’s spelled, BioRxiv. You also herd from Jon Sanders of the University of California, San Diego. There’s a News story about the research available now for free at nature.com/news.
Interviewer: Adam Levy
Later on: sex-specific asthma and the earliest life on earth. That’s in the Research Highlights but before we get to that, reporter Anand Jagatia is here to explain how it might be good to have a little bit of noise mixed in with your signal.
[Traffic noises]
Interviewer: Anand Jagatia
Noise is usually a bad thing. Ah, that’s true in everyday life, and in science where it’s defined as meaningless data that accompanies the information we actually want. But there are situations where noise can be good. For example, in evolution, genetic mutation is a kind of noise. And it gives natural selection something to play with, and when you’re searching for information a bit of error can be beneficial. Down at the library it could be the book next to the one you pick that turns out to be really useful. Nicholas Christakis directs the Human Nature lab at Yale University. He’s been looking at the effect of noise on human coordination and he’s discovered that it can actually help people perform better on group tasks. I asked him to explain the challenge he gave to his volunteers.
Interviewee: Nicholas Christakis
Well what we did here was we had a group of people, about 20 people, and they were playing a classic coordination game. In this game every person is dropped into a network and they’re introduced to their neighbours and each of them is told they have to pick a colour that is different than they’re neighbour’s. They have to coordinate their colour selection with their neighbours. So everyone every second or so is switching their colours around and at some point they might say, well I picked the colour that’s not in conflict with any of my neighbours and in fact everyone in the network may feel that way, or they picked the best colour they could pick and yet the group as a whole haven’t solved the problem. You also see this in firms. So you may also have a situation in which every subdivision within a firm thinks that they’ve optimised their job – you know, engineering thinks it’s done its job and marketing thinks it’s done its job and sales think they’ve done their job but actually the firm as a whole is not performing.
Interviewer: Anand Jagatia
Right, so I guess this is the idea that at a local level people might be doing the right thing and doing a sensible job but because they can’t see the bigger picture, on a global level, things just aren’t working out optimally.
Interviewee: Nicholas Christakis
Yeah, they only have local knowledge. They can only see what they’re doing and what their neighbours are doing. But what we did in this experiment is we – in kind of a sneaky way – replaced some of the humans with bots and furthermore what we did is we introduced what we call a kind of domai, so what we did is we made these bots a little bit noisy. So, if we made perfect bots and the bots played the colour coordination game to perfection, it actually made the groups worse off, and if we made the bots too noisy so they were constantly making errors, it actually also made the groups worse off. But if we gave the bots a little bit of noise, it unlocked these kind of local optima, these ways in which people saw the problem locally but didn’t solve it globally. Adding a little bit of noise to those bots actually helped the groups to converge on a solution.
Interviewer: Anand Jagatia
So a little bit of noise, so adding mistakes, actually helped people from when they were kind of stuck in these situations where they were gridlocked and everybody thought they were doing the right thing.
Interviewee: Nicholas Christakis
That’s right.
Interviewer: Anand Jagatia
One of the things you also mention in the paper is that the bots, by trying out these different – by making these tiny mistakes kind of encouraged people around them to do the same thing.
Interviewee: Nicholas Christakis
Yes, that’s right. What we found is that the bots, in addition to making the life easier of the humans to whom they were connected, also actually changed the behaviour of the humans among themselves. So they led to cascading benefits. So not only did the bots facilitate the task for the people to whom they were directly connected, but actually those people then in interacting with further individuals, changed their way of playing and as a result the whole group was able to work better together and in some ways this is evocative of a kind of teaching function of artificial intelligence. And what we’re interested in here is the use of artificial intelligence, not in replacing individuals, but rather in affecting the network interactions among groups of individuals. What’s the role of artificial intelligence not in our cognition but in our social interaction?
Interviewer: Anand Jagatia
So obviously in real life we don’t go around trying to coordinate coloured dots with our neighbours. Could you give an example of a network where maybe you could put some kind of bot or some kind of artificial intelligence that you could add some noise into the system and it could be beneficial?
Interviewee: Nicholas Christakis
We’re in the point of transition right now between a world in which humans interact solely with each other and a world in which machines interact solely with each other. And we’re going to have a transition period in which humans and machines are interacting on the same plane. One of my favourite examples of this might be to consider what’s going to happen when we have autonomous vehicles on the road and it might not be the case that we want the machines to work perfectly for example. We might want to understand a world in which humans and machines are interacting on a level playing field and explore ways in which the AI with which those machines are equipped could be optimised, given the fact that they are interacting not just with other machines but also with humans.
Interviewer: Anand Jagatia
This is a very unscientific way of putting it but it kind of seems like you’re saying if we want to put machines into situations where they want to interact with humans we have to make them a bit more human like and less machine like, we have to make them, perhaps, make more mistakes.
Interviewee: Nicholas Christakis
I’m hesitant to leap too far beyond our data but I do think one can imagine that we don’t want machines to be perfect. In fact, in our own experiments we found that if we made the machines perfect in several ways it was harmful. For instance, computers can instantaneously change their colours and we found if they did that it annoyed the hell out of the humans. So we had to slow the machines down so that they played at the tempo that humans play. And furthermore we found that if we made the machines always pick the optimal colour, that is to say the colour that had the fewest conflicts with its neighbours, we once again found that that degraded the performance of the group, so that actually we needed to add a little noise to these bots and when we added a little noise, we found that actually they helped the humans to unlock their own potential and not get stuck in these local minima.
Interviewer: Adam Levy
That was Anand taking to Nicholas Christakis who’s at Yale University in Connecticut. There’s also a News & Views article about the study. Both can be found at nature.com/nature.
Interviewer: Kerri Smith
Stay tuned for the News, the Bronze Age equivalent of going viral and century old cancer samples get sequenced. But now it’s the Research Highlights read by Corie Lok.
[Jingle]
Interviewer: Corie Lok
The earliest life on earth may have sprung from hot springs. Researchers studied rocks from Western Australia that were first formed 3.5 billion years ago. These rocks contain some of the earliest evidence for life. The researchers found deposits called geyserite, which are produced by hot springs on land. Near these deposits were layered rock structures called stromatolites which are signatures of microbial life. The researchers conclude that life on land may have emerged near hot springs about 600 million years earlier than previously thought. Find out more from the journal Nature Communications.
[Jingle]
Interviewer: Corie Lok
Asthma is more common in boys than girls but after puberty this trend reverses with women being affected more than men. Researchers think that they have come up with a possible biological basis for this, at least in mice. They found that male sex hormones like testosterone inhibit the development of a specific type of immune cell in mice. In humans these cells help to trigger asthma and other allergic inflammatory responses. The research team found that male mice had fewer of these immune cells and were less likely to develop inflammation in their airwaves than female animals. You can find the study in the Journal of Experimental Medicine.
[Jingle]
Interviewer: Kerri Smith
Whatever your sex, asthma can be made worse if you live somewhere with polluted air. In cities, vehicles belch out many harmful pollutants with diesel vehicles a particular culprit. Some cities are even considering banning diesel vehicles altogether. Adam takes a look at how big a problem diesel is and how difficult it is to measure diesel emissions.
Interviewer: Adam Levy
There’s a study out this week that’s painting a global picture of diesel emissions. It estimates that about 110,000 premature deaths a year are down to particular gases emitted by diesel cars, buses and trucks. That’s in the context of 3 million deaths annually from all sources of air pollution. Countries around the world already have regulations in place to limit diesel pollution, but as lead author Susan Anenberg explains, vehicles driving on our streets, often exceed these limits.
Interviewee: Susan Anenberg
Most vehicle markets around the world are regulating diesel emissions but we do still have a problem where some of these regulations are allowing higher real world emissions – so, higher emissions under real world driving conditions as compared to what the vehicles are showing during emissions testing for certification.
Interviewer: Adam Levy
Of course, tests don’t always reflect the real world. That’s not a surprise. There are many reasons diesel engines might behave differently in tests than under real use. If an engine isn’t properly maintained, for example, it may end up above the lab limit. Or, engines may be affected by different weather conditions.
Interviewee: Susan Anenberg
It could be simply deficient certification test procedures that are non-representative of real world driving conditions.
Interviewer: Adam Levy
A typical test for emissions monitors a cocktail of pollutants coming out from a vehicle’s exhaust while the vehicle travels at set speeds in set conditions – a far cry then from the variation that would be experienced in the real world. Susan’s study focused on nitrogen oxides, known as NOxfor short and extra NOxemissions will cause extra human health problems.
Interviewee: Susan Anenberg
We wanted to know: what are the public health implications of all diesel trucks, buses and cars emitting nitrogen oxide emissions under real world conditions? And to do this we had to get a sense of the real world emissions coming from individual vehicles. So we first developed new emissions factors for all types of diesel vehicles – cars, trucks and buses – in all different regions.
Interviewer: Adam Levy
These emissions factors are used to work out how much vehicles emit in various circumstances. The team then took this information and used it to work out how much excess NOxdiesel vehicles are responsible for around the world. Nobody can measure every single vehicle so to get this estimate...
Interviewee: Dan Carder
You really have to do a considerable amount of extrapolation.
Interviewer: Adam Levy
Interviewee: Dan Carder
You need to be very careful that the activity that you’re measuring and the emissions from those activities are truly able to represent a global fleet.
Interviewer: Adam Levy
Dan Carder knows how demanding it can be to get the right data on diesel vehicles. It was his team’s tests that revealed certain Volkswagen diesel cars were way over the NOxlimits. Now, the Volkswagen emissions were so high because Volkswagen was actively cheating emissions tests. There’s no reason to suspect most other car companies of this kind of foul play. But Dan’s experience shows how important it is to check vehicles aren’t exceeding limits in real world conditions, even if they’ve passed tests. Looking across 11 different regions, Susan and her colleagues were able to work out how much NOxpollution is out there and how much of it is in excess of the limits.
Interviewee: Susan Anenberg
About one third of all diesel NOxemissions within these 11 major vehicle markets are in excess of certification standards so it follows that about one third of the health ramifications of diesel NOxemissions are due to those excess emissions. One third of the health impacts of diesel vehicle NOxemissions could have been avoided if these vehicles emitted the same levels in the real world as they showed during certification testing.
Interviewer: Adam Levy
That’s 38,000 premature deaths a year. So if excess emissions are such a problem across the industry does that mean it’s not even possible to get emissions under the limits?
Interviewee: Susan Anenberg
It is possible based on the in-use emission testing studies, there are some vehicles that are performing der those emission limits. This is not a question of technical feasibility. So this is a problem that is fixable.
Interviewer: Adam Levy
Susan points out that there are already more stringent tests in some countries that do pretty well at keeping real world emissions to the limits. But Dan Carder isn’t convinced tests will ever fully capture the complexities of driving conditions out on the roads.
Interviewee: Dan Carder
I don’t think you will ever find a perfect scenario either in the laboratory or in the real world. Even when you introduce these new real world approaches, you’re cycles are still somewhat contrived.
Interviewer: Adam Levy
Dan imagines that future tests could one day use miniature sensors on whole fleets of cars to crowd source real world emissions data. But he finds this current study eye opening. He usually just thinks about what comes out of engines, not what happens when these pollutants enter people’s lungs.
Interviewee: Dan Carder
Coming from my perspective, we tend to find ourselves always searching for the next technology in hopes of reducing emissions and I think one of the things we don’t get to see enough of in my little world is what that health impacts is.
Interviewer: Adam Levy
And the health impacts of adopting better tests around the world could be huge, Susan’s study estimates that better testing could save around 170,000 lives a year by 2040 and she says that we’re finally beginning to see some countries moving towards more stringent testing, that better captures what vehicles do in the real world.
Interviewee: Susan Anenberg
So I am optimistic that in the coming years we’ll see some improvements towards reducing real world diesel NOx emissions.
Interviewer: Adam Levy
That was Susan Anenberg, partner at Environmental Health Analytics which is a consultancy form in Washington DC. You also heard from Dan Carder who’s at West Virginia University in the US. Susan’s paper is in the usual place.
Interviewer: Kerri Smith
News now, and I’m joined in the studio by Richard van Noorden, European Bureau Chief. Hi Richard.
Interviewee: Richard van Noorden
Hi Kerri.
Interviewer: Kerri Smith
Long time, no see.
Interviewee: Richard van Noorden
Yeah, it’s good to be back here.
Interviewer: Kerri Smith
Now, first, the Bronze Age equivalent of something kind of going viral – a pottery craze called the ‘Bell Beaker’ culture – that’s made it into the News section this week.
Interviewee: Richard van Noorden
Yeah we’re back in the Bronze Age, 4,500 years ago when, yes this bizarre fad for bell shaped pottery swept across prehistoric Europe – all the way across to Spain and Portugal and as far east as Hungary – and archaeologists have been debating what this craze has told them. For more than a century they’ve wondered about this. Does it just mean that this was the Bronze Age’s hottest fashion: that different people shared these ideas and started making these bell shaped pots? Or is it evidence for an immense migration of ‘Beaker folk’ across the continent?
Interviewer: Kerri Smith
And for listeners who can’t see the lovely picture that you have in print of these pots, could you just describe what they are for us?
Interviewee: Richard van Noorden
Yes, it’s a piece of pottery, decorated, and it looks a bit like a bell with a curved bottom and you basically use it probably for drinking. And these were found in burials all over Europe along with other things that defined this nebulous ‘Bell Beaker’ culture. There’s copper daggers, there’s flints. It’s not exactly clear that it was one culture. Some people call it the ‘Bell Beaker’ phenomenon because they don’t want to say that there was one distinct culture involved.
Interviewer: Kerri Smith
So, as you say, archaeologists have been looking at this question for decades. Did the people who made the pots spread and therefore the pots spread or did their ideas spread without them? And now a new genetic analysis…
Interviewee: Richard van Noorden
It turns out that ancient genomes are giving us the answer and there’s a study posted to BioRxiv, the pre-print server, that analyses genomes from an enormous 170 ancient Europeans and then compares that to hundreds of other genomes from ancient and modern Europeans. Now, this genomes analysis has shown that in Euorpe, skeletons found buried with these artefacts are actually quite genetically distinct. There’s no evidence that, as has been proposed, that there’s a well spring of ‘Beaker folk’ in Iberia, Spain and Portugal that migrated across the rest of Europe, not at all, so it looks like it was people sharing ideas and they all just loved making these new-fangled pots. But in Britain there seems to be evidence for an invasion or a migration of ‘Beaker folk’ that came in and seemed to have, within a few hundred years, completely displaced the Neolithic farmers who were there. These are the farmers who built Stonehenge and these famous relics, but it looks like that by about 2000 BC, they’ve been replaced by other groups who, according to this study, look like they were these ‘Beaker folk’.
Interviewer: Kerri Smith
And the Beaker – I love how, by the way, how many times you’ve said ‘Beaker folk’ with a completely straight face. So, well done for that. These are being used as like a proxy, basically, for a culture, and so archaeologists who have been studying this have had a hard time choosing between the two hypotheses. Would that be right?
Interviewee: Richard van Noorden
Exactly, because it’s very confusing. The beakers aren’t found everywhere. It’s sort of discontinuous clumps, where you find them and there’s this enormous range. It’s tricky to know what’s going on. It doesn’t mean archaeologists are going to say ‘great, problem solved’. And some of them are a bit sniffy about some of this study. They agree there’s evidence for a migration into Britain but the idea that these inward migrators completely displaced Neolithic farmers – they say well, in the archaeological records they’re not sure about that. The DNA may say that when you look at later people, there’s very little of the Neolithic left in the DNA but archaeologically there doesn’t seem to be such a huge split and they suggest that as cremation arose around the time of the Bronze Age and the middle of the Bronze Age, that may have cleared away a lot of the bones and evidence that, if we could only sample them, would show actually there was a lot more Neolithic DNA involved in later people in Britain.
Interviewer: Kerri Smith
We’re going to confine ourselves to history for the next story but only in the last hundred or so years because cancer researchers in London have been looking to history to find more samples of very rare cancers.
Interviewee: Richard van Noorden
Yeah, this is in London at Great Ormond Street Hospital, founded in 1852, on the back of donations raised by Charles Dickens in fact and there’s a lot of samples – a huge, 165 year old archive of samples and patient records in this hospital and many of them are these childhood tumours stored in the paraffin wax cubes with all the patient numbers hand written on the sides, going back, as I say, more than a century.
Interviewer: Kerri Smith
Why would that be useful to a modern cancer biologist?
Interviewee: Richard van Noorden
So the problem is that there are very few tumour samples from rare cancers that are available for modern researchers to sequence. We talked to Sam Behjati – he’s at the Wellcome Trust Sanger Institute in Hinxton – and he says the treatment regiments for children with rare cancers are essentially made up. If you’ve got 3 or 4 patients in the whole country, how are you ever going to conduct a reasonable clinical trial? Figure out what treatments may or may not help treat certain mutations… So he dug into the archive of the Great Ormond Street Hospital and picked out 3 samples that are a century old and showed that they could, first of all, agree with the diagnosis of those samples. He picked out a muscle cancer, a blood vessel tumour and a lymphoma and then extracted DNA and sequenced hundreds of genes in each slice that they pulled of this tumour, and indeed they could find the cancer-associated mutations in the samples, suggesting that there’s a rich vein there, not just at Great Ormond Street Hospital, but at the archives of other hospitals for these relics of childhood cancers. The genomes of adult cancers often have hundreds of mutations whereas the genomes of childhood cancers tend to contain far fewer alterations. So you can much more easily hone in on the most important mutations and sift through the background noise of degraded DNA.
Interviewer: Kerri Smith
Richard, anything else that listeners should be aware of in this week’s News section?
Interviewee: Richard van Noorden
Yeah, for our listeners who haven’t heard enough about Donald Trump, we lead on his proposed budget for 2018, but we looked specifically at what he’s proposing to cut in science and interestingly, how that could hurt his own supporters. We look at the science projects that are in the so-called red states, the Republican states, for example we talk about the Gulf of Mexico where a fish called the red snapper has made a comeback. US Government regulations have rebuilt the fish stocks, but Trump wants to eliminate a Mississippi based sea-grant program that’s overseeing a massive study of red snapper stocks and is supposed to guide future management decisions and protect this fishery that hauls in billions of dollars for these Republican dominated Gulf states.
Interviewer: Kerri Smith
Great. Richard van Noorden, thank you very much for coming in. More on each of those stories which are by Alexandra Witze, that’s the red states analysis; Heidi Ledford wrote about the old tumour samples and Ewen Callaway has been looking at the migration of the ideas of the ‘Beaker people’, all on nature.com/news.
Interviewer: Adam Levy
That’s all we have time for this week but for more from us, make sure to follow the podcast on Twitter @naturepodcast, or you can find my personal tweets @climateadam…
Interviewer: Kerri Smith
… And mine @minikerri. See you next week. I’m Kerri Smith.
Interviewer: Adam Levy
And I’m Adam Levy.
[Jingle]
