Monthly Archives: June 2022

Sensory Worlds Beyond Our Imagining

An Immense World; How Animal Senses Reveal the Hidden Realms Around Us, Ed Yong, Random House/Bodley Head, June 2022

This book is an enormous achievement. A thrilling read, taking us into the Umwelt, or perceptual world, of numerous mammal, fish, reptile and insect species. A major work of scholarship, with over a thousand references to a 45-page bibliography, as well as accounts of interviews with numerous researchers and visits to their laboratories. An exploration of many ways of sensing the world, some of which we share, while others are beyond our imagining. The evolving interplay of perception and action, communication and deception, environment and response. And an enhanced insight into what it is like to be a bat, a bird, a blue whale, a beetle, or a human.

From the wealth of detail in the book, a consistent grand narrative emerges. Some physical process interacts with living matter. This is the raw material for sensation. Sensory abilities then shape a creature’s Umwelt, being developed according to the demands of its environment. But every perceiver is itself an object of perception to others, and we have colour displays and camouflage, smells as signals and identifiers, sound as communication to others and, by echolocation, back to the creature who generates it, and the same is true of other senses that we do not share, such as the detection of tiny electrical fields. Senses combine and even, we suspect, merge, and what we ourselves perceive is but part of an immense pattern. But the heedlessness with which we amplify our own signals disrupts this pattern, contributing to our destruction of nature, and we ourselves are the poorer for it.

Let me offer a few samples from the book’s wealth of detail.

Yong starts with taste and smell, two ancient senses that operate by direct molecular contact. It is not long before he surprises us. Snakes use their forked tongues to smell in stereo. Humans are poor compared with other mammals at detecting smells at low levels, but are rather good at telling different smells apart. No one knows how smell relates to chemical structure (contrast this with how seeing relates to the wavelength of light, hearing to frequency, or touch to pressure). As every well-trained dog-owner knows, smell is central to the Umwelt of dogs, but I would never have guessed that the same is true of elephants. And the molecules involved in smell include opsins, which are central to vision. As Yong puts it, in a way we see by smelling light.

File:Ed Yong.JPG
Ed Yong on a tardigrade, in Micropia, via Wikimedia

It is sometimes said that dogs, and most other mammals, are colour blind. This is only half true; they have just two kinds of colour receptors, while we (like other apes, and our cousins the Old World monkeys) have three.[1] The colours that we perceive arise from subtle interactions between these receptors. Thus some neurons are excited by blue cones but inhibited by red or green, while others are stimulated by red but inhibited by green. So the colours are we experience are the result of a kind of neuronal arithmetic, below the level of conscious awareness. I still feel surprise when someone superposes red and green beams of light on a screen, and I see the result as pure yellow. Our colour vision can be represented by a triangle, with red, blue, and green at the corners, and yellow halfway along one edge.

But hummingbirds, like many birds and reptiles, can see four kinds of colour; red, green, blue, and UV. If human colour vision can be represented by a triangle, that of a hummingbird is a pyramid. And while for us the overlap of red, green, and blue are enough to produce white, four kinds of sensor need to be activated to look white to a hummingbird.

Among humans, different individuals have slightly different sensitivities, and some women show a degree of four-colour vision, having inherited different pigments from the sites on their two separate X chromosomes that are responsible for colour vision. Words cannot convey this added perceptual dimension, but the fact that they really do have four-colour vision can be demonstrated by discrimination tests. (This by the way answers a philosophical riddle that intrigues me. How do I know that you are seeing colours the same way that I am? It turns out that there is a real chance that you aren’t.)

Many flowers that appear white to us are coloured in the UV, and appeal to insect pollinators with green, blue, and UV three colour vision. And UV vision evolved in insects long before there were flowers, so the flowers evolved the pigments to attract the pollinators.[2] Thus the ability to see directly influences the evolution of what is there to be seen.

Temperature detection overlaps other senses. There are sensors that can detect hot or cold temperatures, but can also be stimulated in different ways. The sensor for painful heat can be activated, on the skin as well as in the mouth, by habanero peppers, while menthol feels pleasantly cool in the mouth or the smoker’s throat. Menthol also happens to be addictive.

The ability to detect heat at a distance is useful to species that suck blood, from bedbugs and mosquitoes to vampire bats. Ticks can detect body heat from up to 13 feet away, and common insect repellents work with them, not by interfering with smell, but by blocking the heat sensors. These sensors are in spherical pits on their legs, and the pits are covered by a film with a hole in it. So the sensors give directional information as well as detecting the heat source. Rattlesnakes and other pit vipers also have pits with a narrow opening, falling on a sensing membrane that carries around 7000 nerve endings. They can detect the presence and approximate direction of an increase in temperature of 0.001oC, which means that a viper can locate a rodent 1 m away. Information from the temperature sensing pits is combined with information from the eyes, so maybe for them the sense acts as an adjunct to vision, rather than on its own.

Touch organs can be modified for special purposes. The emerald jewel wasp paralyses cockroaches by stinging their brains, and has a touch sensor at the end of her sting to locate the exact location. A wide range of mammals, including the opossum, a marsupial, as well as rats and mice, use touch sensors at the base of whiskers to explore their surroundings several times a second. Since each whisker has its own connection to the somatosensory cortex, this builds up a map of the surroundings. So whisking, as it is called, is perhaps more like seeing than like touch. It would seem to be a very ancient trait indeed, since the last common ancestor of placentals and marsupials was back in the age of dinosaurs. The whiskers of seals are so shaped and angled as to minimise the forces on them as they move through the water, which would otherwise overpower the pressure waves caused by passing fish.

Sound detection is fast, precise, 24 hour, and useful for detecting predators or prey. Sound is also used in communication, as in the finding and assessment of mates. But mating calls come at the cost of giving away one’s location. There is a species of parasitic fly that has developed ears remarkably similar to those of the crickets it preys on, to eavesdrop on their mating calls. On Hawaii, which was once seriously infested by such flies, the crickets have fallen silent. Once again, the overlapping Umwelten of prey and predator drives evolution.

Surprisingly, the first insects were almost certainly deaf, since hearing has evolved separately among them at least 19 different times, on many different parts of the body, having in general been developed from organs that respond to vibration and pressure.

We can only detect parts of how animals use sound to communicate. Birdsong contains more structure than the human ear can resolve, unless it’s played back slowly. In fact, the structure within each note may be more important to the birds than the order that the notes are played in. Whales and elephants both use what we call infrasound, vibrations too low in pitch for our ears to hear, as a way of keeping in touch over long distances. Mice, however, communicate using ultrasound, frequencies too high for us to detect. The terms infra and ultra are arbitrary, relating to our own capacities, but since we are deaf to such sounds they were not even detected, let alone studied, until a few decades ago, and may be much more important in nature than we realise.

Echolocation in particular may be much more extensive than our knowledge of it, when it uses frequencies that our own ears cannot detect. Even bat echolocation, although suspected much earlier, was not clearly demonstrated until 1938. It enables bats to navigate and catch insects in complete darkness. This is an impressive feat; the bats need to generate short pulses of high-frequency sound, and then detect the direction and timing of the faint echo from a small moving target. Some bats can even tune their ears to respond to a frequency slightly different from the one they are emitting, and detect the movement of their prey using the Doppler effect. [3]

But moths are not merely passive prey. They have ears that can detect bat cries, and dodge and loop to evade capture. Tiger moths produce clicks of their own, which confuse the bats. Some moths even have long flexible tails at the end of their wings, which may also add to the confusion.

Infrasound echolocation by dolphins was detected in the 1950s, and since the 1960s the U.S. Navy has been training them to find sunken equipment and mines, and aiming to reverse engineer their abilities. Humans avoid walking into obstacles using echolocation, and some blind people have developed this to a high skill, building up a model of the world in their visual cortex, in much the same way that most of us do so using sight.

Darwin was puzzled by so-called electric eels, which use electric shocks to stun their prey. After all, evolution regards present organs as the result of a series of incremental improvements, but what use would the electric organ have been in its feeble first stages of development? It took a century to find the answer. Many fish possess a lateral line, sensitive to pressure, and in some cases this has been modified to detect electricity. This confers an obvious advantage, since any living thing moving through water generates tiny electric currents. And electrodetection gains in sensitivity and acuity when combined with the ability to generate one’s own more powerful electric field. So we have passive electroreception and active electroreception, just as we have hearing and the use of echoes. The cells that detect the electrical fluctuations are hair cells, basically similar to the same cells that detect pressure waves on the lateral line, or pressure oscillations in our own ears. Active electroreception operates in every direction, will work as well in cloudy as in clear water, and is so sensitive that some fish can be trained to detect the difference between a clay pot full of river water, and one also containing an insulating glass rod.

Passive electroreception is extremely widespread among vertebrates, being used by sharks, catfish, and salamanders, while the platypus has over 50,000 electroreceptors in its bill. Bumblebees can detect the electric fields that surround flowers, and it may well be that electroreception is much more common than we as yet realise among insects, equipped as many are with touch- and current-sensitive hairs.

Yong concludes his list of the senses with the ability to detect magnetic fields. This is a difficult area, if only because magnetic effects are extremely weak, and show subtle variations on a global scale in direction, intensity, and angle of dip relative to the surface. To complicate things further, no one even knows what the magnetoreceptor would look like or how it operates. Some bacteria grow small crystals of magnetite, and can distinguish North from South, but no one has managed to find similar structures in the birds and animals that are known to use magnetic fields as an aid to migration. One current theory invokes what are called radical pairs, molecules raised to high-energy states by the influence of light, but such states are short-lived, and I as a chemist would require a lot of convincing.

Senses interact. Mosquitoes are attracted by body warmth, but only if they can smell carbon dioxide. Electric fish that have learnt to distinguish shapes using their electric sense are then able to do so by sight, and vice versa. I have already mentioned how the mental maps constructed by blind people, who have learnt to navigate using echolocation, reside in the visual cortex. And finally, we need to remember that sensation requires discriminating between the signals that come to an organism from outside, and those that it generates by itself.

So all complex animals, including ourselves, perceive only a small part of the immense world of possible sensations, and construct their own Umwelt from the part accessible to their own senses. But we show brutal insensitivity in how we influence this world. We brighten the night sky and blur the distinction between the seasons, confuse forest insects with the sound from our machines, scatter huge amounts of material that must distract the sense of smell, and make the very oceans reverberate, so that whales navigating by infrasound end up stranding themselves in response to naval sonar.

And when did you last see the Milky Way?


Ed Yong’s earlier book, I contain multitudes: the microbes within us and a grander view of life, was a New York Times bestseller, and in June 2021 he received a Pulitzer Prize for Explanatory Reporting for his writing on the COVID-19 pandemic


1] Red colour perception evolved by accidental duplication of the gene that codes for the green-sensitive receptor, followed by Darwinian selection for a new use – the ability to detect ripe fruit, or tender young leaves, against a green background. A nice example of how evolutionary change generates new information.

2] Yong does not tell us how we know this, and to do so would have required a chapter in itself. But, in brief, the methods involve cross-species comparisons, and, these days, the use of molecular biology to reconstruct family trees for the relevant genes. The simplest assumption is then that a trait prevalent in one particular clade was present in its last common ancestor.

3] This is the familiar increase in the perceived frequency of a sound wave when the distance between source and observer is decreasing. Here the decrease is in the length of the round trip, from bat to target and back again, but with further fine tuning from the motion of the prey, and even the flapping of a moth’s wing has a detectable effect.

This post first appeared in 3 Quarks Daily

%d bloggers like this: