I suppose birds are the animals that most commonly spring to mind when people think about animal navigation.

In researching Incredible Journeys/Supernavigators, I spent many months reading about how birds find their way and interviewed some of the leading researchers in this field.

Homing pigeons have for a century or more been the focus of a great deal of research, and over that time it’s become clear that these very remarkable animals have a complex navigational toolkit at their disposal.

This includes both a sun compass and a magnetic one. These are coupled with the ability both to learn complex routes and to make mental ‘maps’ of the area around their home lofts - based on visual landmarks.

Smell seems to play a part in pigeon navigation too. Indeed it may be the key to their astonishing ability to set a homeward course even when released at a completely unfamiliar place hundreds of kilometers away. (There is also some intriguing circumstantial evidence that they make use of infrasound for the same purpose, but that is another story - see chapter 13 of my book for more on this subject.)

It was the Italian scientist Floriano Papi of the University of Pisa who first proposed that pigeons made use of an ‘olfactory map’ some 40 years ago. This was a very controversial idea at the time, and it remains so today, even though it is now much more widely accepted. In fact it is one of the biggest - and oldest - puzzles facing researchers in the animal navigation field.

The basic idea is simple, even if it may at first seem pretty implausible.

Young pigeons in their home loft are exposed to passing currents of air that carry to them a variety of scents (or ‘odorants’ to use the ugly scientific jargon). Assuming that the odorants reaching them vary in a regular way according to the wind direction, the birds learn to associate each smell with a specific direction.

When the birds are later released in an unfamiliar place, they sniff the air and see whether they recognise any familiar smells. If they do, they assess whether the smells are stronger or weaker than those they detected at the home loft (in other words, whether they are more concentrated or less). If they are more concentrated, then the birds can tell that they must have been moved towards the source of that particular smell. They will therefore know that, if they want to head towards home, they should go in the opposite direction to the one they associated with that same smell when at the loft. So, if they learned to associate the smell with a south-westerly wind, they would head in a north-easterly direction. And they would do the opposite if the smell was less concentrated.

Lots of animals - including many insects and even humans - can follow a scent trail, but what what I’m describing here is quite different. If Papi was right, the pigeons are working out how near or far they are from the source of a smell and using that information to work out how to reach a different place - their home loft. (Though of course they may also be attracted to the smell of the loft itself when they get close enough to it.)

Making use of a single odorant is the simplest form of olfactory navigation. But - so the argument goes - there’s no reason in principle why pigeons shouldn’t simultaneously detect several different smells and then build some kind of a cognitive map based on their spatial distribution in the environment.

Such a map could work in two different ways.

According to the simpler version, pigeons learn to recognise the very particular combinations of smells (or bouquets) associated with different locations as they explore their environment . These characteristic bouquets then provide the basis of a mosaic map - a kind of patchwork quilt of unique smell-signatures dotted across the landscape, all of which the birds can recognise and place.

The mosaic map theory is superficially attractive and seems quite plausible, but plainly a map that relies on prior experience would be of no use when a bird is displaced to a place it has never previously visited.

In search of an explanation for this mysterious long-distance homing behaviour, scientists have come up with a much more complex theory - the so-called gradient map.

The idea is that pigeons can actually plot a course for home by reference to the relative concentrations of different odorants. In order to do so they would have to build a map-like representation of the spatial distribution of the concentration gradients of several different smells across the landscape.

Here’s how this might work.

Suppose there are three different smells - A, B and C - emerging from different locales. One might be a forest of pine trees, another might be an oil refinery and the third might be the open sea. At any given time and place these odorants will be present in varying combinations - depending on multiple factors like the wind direction, temperature and time of day.

Let’s assume that the bird has learned to associate each of these smells with a particular wind direction, and has also learned how to factor all the other variables into its calculations. Then, as it travels around and samples the air, the bird will be able to use the relative concentrations of the three smells to estimate reasonably accurately how far it is from each of their sources and in what direction they lie. By combining all this information (in effect, integrating the three vectors) it will be able plot a course for home.

Critics of the gradient map theory have complained that atmospheric turbulence (as well as other factors) would surely make it very difficult in practice for the birds to extract much reliable information from olfactory cues.

In trying to answer to these criticisms, proponents of the gradient map theory have for some time been trying to show that relatively stable concentration gradients of certain odorants do actually occur in the real world.

Now an interesting new piece of research has strengthened their case.

The researchers sampled the air at a pigeon loft near Pisa in Italy, and at three sites in forested country around the loft. They also sampled the air at an altitude of 180m (the height at which pigeons typically fly) from an aircraft flying in a concentric circles around the loft.

Using spectrometry, they found that certain natural and man-made volatile organic compounds (VOC) displayed ‘repetitive daily patterns’ that varied according to wind speed, direction and chemical oxidation.

One VOC (dimethyl sulphide - or DMS) showed a strong East-West gradient, and others a strong North-South gradient. The roughly 90-degree angle of intersection of these gradients suggests that a bird could in principle compare them to determine its (approximate) position on a gradient ‘map’.

The scientists also tracked the paths followed by 143 pigeons released at three different sites around the loft, using GPS loggers.

Although (frustratingly) they weren’t able to sample the air when the birds were actually released, they used a meteorological computer programme to model where the air masses which the pigeons encountered had come from. Whether the air had recently passed over the sea was of particular interest since it would then probably be laden with DMS, a characteristic smell generated by marine microorganisms. (DMS was in fact one of the most abundant VOCs detected in their sampling exercises at the loft site - which is quite near the sea.)

Assuming that the birds at their loft would associate DMS with westerly winds coming in off the sea, the scientists predicted that the paths they followed on release would be influenced by the concentration of DMS to which they had been exposed. Specifically, they expected that low DMS concentrations (indicating that the sea was further away) would encourage the birds to head in more westerly direction. This turned out to be the case.

So what does all this actually mean?

The new study bears out earlier claims that stable concentration gradients of different odorants do exist in the real world, and that they could provide the basis of a gradient map that a bird might use. The tracking data also suggest that DMS may be one of the odorants on which pigeons actually rely for navigational purposes.

Further experiments - probably lots of them - will be needed before we can be confident that pigeons really do use their sense of smell to generate a gradient map that allows them to navigate home from unfamiliar, distant locations. But this is progress!