This is the second of two interviews, to accompany my latest New Scientist feature on how birds sense magnetic fields. Thorsten Ritz was one of two scientists who blew this field of research open in 2000, with a landmark paper that suggested how migrating birds could detect the faint traces of the Earths’ magnetic field. My interview with his partner, Klaus Schulten, is elsewhere on the blog, along with more background on the topic. These interviews are meant to provide bonus extras to the other freelance writing I do, acting as a home for great material that would otherwise be cut and lost. And Ritz’s interview is great material. He talks eloquently about the science of the magnetic sense but more importantly, he talks about what it’s like to work in this field – the reasons why progress has been slow, the thrill of crossing disciplines, and the feeling of doing “19^th-century science”. I’ve edited the transcript for length, but it’s still long – if you read any of it, read that second half (starting from the fourth question). Why did you start looking into the magnetic sense of birds, and can you describe how the ‘radical pair idea’ works? I’m a physicist. I came into this field because I wanted to work out how you could, in principle, detect a very weak magnetic field in a biological system. You don’t have a superconductor or things like that. One idea that has been around since the 70s is the radical pair idea. At the molecular level, electrons have a property called spin. Usually, an electron is paired in an atom with another electron with spin in the opposite direction, so they cancel each other out. But you can take an electron and move it from one molecule to another, for example, by shining light on certain molecules. Then you create a transient state – the radical pair state – which becomes sensitive to the action of external magnetic fields [A radical is a molecule with an unpaired electron – Ed]. That part is fairly well understood. Depending on the alignment of these electrons, you have different spin states – singlet or triplet. It doesn’t really matter how you define it but what matters is that they’re chemically different so that if you’re in the singlet state, you can form different products than if you’re in the triplet state. You can think of it like a chemical switch. Depending on the effect of the external magnetic field, you get a bit more or a bit less of one type of product over the other, or you might only be able to form a product from one state. These are the ways in which, in principle, that magnetic fields can affect chemical reactions. All of this has been studied for some time. The big question is whether this is what’s going on in a bird. If this idea has been around since the 70s, why had no one linked it to magnetoreception in birds? When I looked at this 10 years ago, as a grad student, I could tell that this idea had got a bit of a short shrift from the biologists. From a physicist’s perspective, there really wasn’t any reason why people shouldn’t believe in this idea but I suppose that it’s hard to visualise. My advisors and I thought that it might be a good idea to write a paper to point out the things that one should look for if one believes in this idea. One of them is that the magnetic field would manifest itself as an indirect effect on some light-sensing pathway; maybe it’s an indirect effect on vision. So then we said one should look for areas linked to vision. We also thought about what the molecule would be. Most molecules involved in vision use light to change the shape of the molecule rather than to transfer electrons (for example, the receptors in our eyes). In 1998, there was a new light receptor discovered called cryptochrome and that was the first one that was known to exist in animals and in birds that uses light to transfer electrons. It had the right chemistry. We also suggested a set of indirect tests, such as using oscillating magnetic fields to disrupt this mechanism. And over the past decade, people looked at all these ideas and they checked out. We saw an effect of the oscillating fields. We saw cryptochromes in the eyes of birds. I’d like to know more about where exactly they are – I hope to see them in photoreceptor [light-detecting – Ed] cells but that study hasn’t been done yet. Henrik Mouritsen and the Wiltschkos have shown that there is a brain centre called Cluster N that is closely linked to the eye; if you lesion it, the magnetic compass is disrupted although the birds can still find their way according to the stars and sunset. This may really be a brain centre linked to magnetic processing. None of that proves the radical pair idea but they strengthened this idea. Cryptochromes are common to all sorts of animals, so why is it that only some groups can sense magnetic fields? People have looked at magnetic field effects quite across a range of organisms. In plants, there are growth responses that are controlled by cryptochrome and magnetic effects have been seen. The same is true for circadian rhythms in fruit flies. Reppert’s group have done experiments where fruit flies could be trained to move to one side based on a magnetic field, but not if you took their cryptochrome away. So there’s a possibility that there may be something special about cryptochrome in that it may be a molecule that is capable of detecting magnetic field strength. On the other hand, your question is really an evolutionary one. To answer it you have to look at the evolutionary context and the advantage that is conferred by the magnetic field. That is very hard to answer. Cryptochromes are often involved in day/night rhythms but they’re used in very different ways in different organisms. One of the big things we lack is a genetic model for magnetoreception. Most of the experiments have been done with migratory birds, which are fascinating animals but not ‘genetically tractable’. You can’t do a lot of the things that biologists do to answer these types of question. We just haven’t been able to ask the right questions in organisms like fruit flies, where you could breed some that detect magnetic fields better than others. These studies still need to be done. For humans, there’s a chance we still have the right molecules but we simply don’t use them and they’ve become degenerated over time. At some point in my life, I’d love to do a study with some indigenous people who orient in areas with few landmarks, such as the Pacific islands. This seems like an incredibly difficult area of research, where progress has been slow. What makes it so tough? I’ve asked myself that question a lot of times too. In some ways, I’m grateful that it’s slow because a young scientist like me has the chance to make a big discovery! We shouldn’t overlook the fact that we as humans don’t consciously use that ability and it’s a bit tricky to ask, for example, what do you expect this sense to do? For almost all the other senses, you could answer that, maybe with the exception of echolocation (but even there we have radar, so we have a good idea how you would use that). For a compass sense, we don’t really know what to use it for so it’s hard to see what the best stimulus would be. For all the other senses, the first thing you would look for is a dose-response curve; if you give more of a stimulus, what happens? But the magnetic field is pretty much all-encompassing. You don’t switch it off. It might change its direction but it won’t change its intensity. Basic things that you do in other senses don’t make sense when it comes to magnetoreception. Almost every other sense is linked to an opening in bone structure – eyes, ears and so on – where we get our information. The magnetic sense could sit anywhere in the body because the magnetic field penetrates the body. So there isn’t an obvious location to look for, where we can say, “This is where the magnetic organ is,” or, “Here’s another bone opening and we don’t know what it’s good for”. And finally, I think our field is heavily dependent on behavioural biology results – looking at an animal and testing it in a cage and so on. These experiments are a lot of fun and fascinating because you’re dealing with a whole animal and not just a cell, but they’re hard to interpret because when an animal does something, you don’t know why it does it. And it’s slow – you can only work with migratory birds during migratory season. They’re wild animals, you have to catch them, get permits, and so on. So that slows down research a lot. So why don’t we move to another organism? I speculate that the magnetic sense is just one of many senses that feeds into directional information, and it may not be the dominant one. That’s why migratory birds are so prominently used. During the migratory season, the migratory sense becomes very dominant. You can put them in a cage and they still want to go north and you can measure that. But if you just have an animal roaming around looking for its hiding place, I’m sure it won’t just use the magnetic sense. It will also use vision and smell. So trying to take away these other senses and leaving the magnetic sense may not be natural – you want a mix of things. Even if you believe in the radical pair mechanism, maybe the bird compares its magnetic information to its visual flow. Maybe the bird needs to turn its head to find north so if you keep its head still, it won’t work. That makes it so much harder to design a good reproducible experiment because the more factors you have, the harder it will be to reproduce results. We have very little to guide us as to what’s the right design. All of that means that we have mostly behavioural experiments to date and if a science develops by behavioural experiment s alone, it will be slow. So we need neurophysiology. We need to find that magnetic neuron and then you can get data in a very different way. We simply don’t have that yet, or the genetics, or the molecular biology. These are the steps that drive the field. When you look at vision, you can take retinas and opsins [proteins involved in vision – Ed]and put them in a test-tube and study them. With the magnetic sense, maybe cryptochrome will get us there but we’re not there yet. It’s amazing how many areas of science you need to understand to really get this topic... Yeah, it’s quite amazing how much we know despite of that! There are many animals that are studied quite a lot in biology but it seems that the animals that lend themselves to magnetic studies just aren’t them! We want to understand the magnetic sense, so we have to work with organisms that have it. And they aren’t the organisms that are easy to handle. Why don’t we have transgenic birds at this point? It would be far easier to take a robin and knock out the cryptochrome. We’re not the experts at this but the people who are the experts are typically not interested in these questions. It’s like conversion, person by person. You often have scientists who are behavioural biologists who move into neurobiology or genetics and that pick up these techniques. But the big geneticists haven’t taken an interest in this yet. That sounds fun, though... I kind of like it. I like that this is a field that’s a bit 19^th-century science. You have to work across disciplines and be holistic. When I did these calculations with oscillating fields, we could do calculations at the level of a radical pair and I can do the quantum mechanics. But then you have to make predictions of what an animal’s going to do in a cage, so you have to make sure you’re not completely stupid about signal transduction and neural processing and all these things going on in between, of which you know nothing. So you have to work, which is very different to areas where you don’t cross so many layers, where you work with a neuron or a cell without having to understand animal behaviour. What are the big questions that are still to be answered? For the radical pair idea, there are a couple of obvious things that have to happen. If cryptochrome is the right molecule, you should be able to design an experiment where you see magnetic field effects on cryptochromes. You take them out, put them in a test tube and look at them. There need to be studies to show where the cryptochrome is in birds, and which ones (there are multiple kinds). Then you need to take the cryptochrome out to show that the magnetic sense doesn’t work anymore – that would be an important link. Ideally, you still need to find that neuron where magnetic information is transmitted in the eye – that’s the big one. You want to see that you give a stimulus and you get a response somewhere in the nervous system. These are the steps for probably the next decade. The last one has been quite a good one. Compared to the 80s and 90s, much has happened – a lot of new ideas and concepts and proofs-of-principle. Wolfgang Wiltschko has said to me that he wishes this all happened 20 or 30 years ago so that he could have played a much more active part in it. I’ve always been wrong when I try to give a time window, but I’m quite optimistic. But I think that if we look back, we’ll view the next decade as the one where some of the big discoveries were made. The next question for me is how is the magnetic sense actually used? What piece of the magnetic information is used by the bird? When you’re hiking, you don’t use a compass all the time. You take at a map and you use a compass to turn the map the right way. Then you think “I need to walk towards this mountain or this tree”, and then you put the compass and map away and walk towards the tree with your eyes. It’s possible that other organisms do something similar. If you have two sensory capabilities as it looks like in birds, what do they do? Which one does what? Why are there two? Why isn’t one enough? I think we’ll be busy for a long time in studying that!