Sounds Wild and Broken, page 11
Each tachinid fly species has its own sonic preferences, some preferring short trills, others rapidly delivered chirps, and each is sensitive to a particular range of frequencies. For prey, this specificity means that there is an advantage to sounding different from other species. Natural selection therefore favors sonic diversity. By sounding different from the crowd, singers avoid being assaulted by hordes of parasitic maggots, a strong incentive. Specialized hearing by parasites can generate regional variants in both the songs of hosts and the hearing preferences of the parasites. The diversity of rain forest trees here is explained, in part, by a similar process. Any tree species that becomes too common is cut back by fungi, herbivorous insect mouths, or viruses. Rarity buys a measure of safety. Over time, this results in more diverse communities.
Tachinid flies seek the sonic signature of a small number of species. Most other predators that hunt by ear, though, have ears and palates with more catholic tastes. Any large night-singing insect here advertises its location to listening crested owls. Calling frogs are picked off by slate-colored hawks that lurk in the vegetation along rivulets. Wolf spiders feel the tremors of singing insects both in the air and through vibrations in their legs. When an ornate hawk-eagle soars over the canopy, its ears, eyes, and talons seek mammals and birds alike, from doves to macaws, squirrel monkeys to spiny wood-rats.
These generalist predators also shape the sounds of their prey. If you’ve ever tried to creep up on a singing tree frog or katydid, you’ve experienced the sudden silence that a passing shadow or rustle of vegetation can elicit. But prey do not only fall silent when danger swoops in, they often give alarm calls, a seemingly paradoxical response. But by calling, prey signal to the predator that it has been seen. With no possibility of a stealthy, unexpected attack, the predator is often better off leaving the area to seek less wary prey. Alarm calls are also part of the cooperative networks that bind animal societies. Calling animals warn others, benefiting their own progeny and kin, storing up social capital with neighbors, and helping their group thrive while others perish.
The function of alarm calls is embedded in their acoustic structures. When a bird-hunting hawk lances through the forest, small songbirds often give high, thin see calls. Other birds dive for cover, responding to the alarm in one-tenth of a second. Hawks swoop at prey at up to fifty meters per second, and so vocal alacrity and split-second responses are essential if prey are to dodge the attack. The structure of the see call conveys alarm to others while minimizing risk to the caller. High, pure tones with tapered starts and stops are cryptic, like camouflage for sound, giving the hunter little information about where the caller is located. The calls are hard to find because they lack abrupt onsets that provide binaural cues about position and are shrill enough to be at the edge of the hawk’s hearing. The high sounds also attenuate quickly in vegetation.
If a predator lingers, the element of surprise gone, songbirds switch to repeated, lower pshht! pshht! calls, vigorous harsh sounds that travel far and plainly advertise the birds’ presence. These calls draw other songbirds within earshot into mixed-species crowds that mob the predator. Birds often dive-bomb the hawk or owl from behind, swooping through branches, then veer away on nimble wings. The recipient of such mobbing behavior often moves on.
Alarm calls are not generic. They do not merely convey the presence of danger. Some birds recognize the voices of mates and kin, responding more vigorously to alarm from these familiars than to strangers’ calls that, to human ears, sound identical. Alarm calls of birds and mammals can also contain information about the species and proximity of the predator. Snakes, small owls, small bird-hunting hawks, and large hawks or eagles all elicit different calls from their prey. The signal for a distant predator is different from one whose strike is potentially imminent. Animals with highly developed social networks—crows, ravens, prairie dogs, monkeys—also use alarm calls to communicate the individual identity of a predator and the threat posed by this individual. Representation in sound of individual identity of predators demonstrates sophisticated cognitive capabilities. These animals recognize individuals, remember salient characteristics of each, then communicate this knowledge to others using sounds that carry information within their forms. In cleaving humans from all other animals, Descartes believed that non loquitur ergo non cogitat, “he does not speak therefore he does not think.” If the philosopher had opened his ears and imagination to the alarm calls of the birds outside his window, his logic might have been reversed, loquitur ergo cogitat.
The information encoded within calls is a language that crosses species boundaries. By listening to what other species are signaling, birds and mammals assess the presence and identity of predators. Species that are preyed upon are knit together in a communicative network rich with nuanced representation of danger and identity. By giving the calls of other species our attention, we humans can join this network. If a bird stabs the air with a see, look up and see the hawk slicing low through the trees, hoping to surprise its prey. A cluster of scolding songbirds likely has a small owl in its center. A loud, repeated alarm indicates more immediate danger than the same calls given just once. A squirrel or bird giving a repeated harsh complaint and moving slowly through low branches is likely following the progress of a fox or other mammal. As I’ve worked over the years to open my ears, I’ve found that tuning in to the sonic network reveals previously unseen animals: a coyote in the brush at the edge of the park, a pygmy owl deep in the branches of a fir tree, or a hawk sweeping through gaps in the forest understory, in view for only a second.
Alarm-calling behaviors create opportunities for deception. An alarm given when no danger is present can distract and deflect a competitor or predator. If a male swallow suspects that his mate is in a tryst with a neighbor, his chittering call breaks up the assignation. Male lyrebirds in Australia sometimes mimic the sounds of a flock of alarm-calling birds. This causes females to pause as they peruse the territory, giving the males a little more time with potential mates. Some caterpillars give birdlike see alarms when pecked, surprising their attackers and allowing time to escape. Primates and dozens of species of birds give deceptive alarms during moments of intense competition for food. They shriek, then grab the food of fleeing competitors. The champion of this ruse is the African fork-tailed drongo, a bird species that mimics the alarm calls of forty-five other species. Drongos match the sound of their calls to the alarm call typically used by their victims. But this only works on the first heist. To avoid habituation in the victim, the drongos switch to different alarm calls on their second attempt with the same species.
Alarm calls give us a window into the complexities of nonhuman animal sounds. Unlike the many sounds made when animals are feeding, breeding, and interacting with their offspring, alarms are given in relatively simple contexts and thus are easy to study. Carry to the woods a taxidermied owl hidden under a sheet, then yank off the veil. The squirrels call out in fright. String a wire across a field and dangle a stuffed hawk from a pulley. The fake predator swoops, and songbirds shout their see alarms as they scuttle for cover. Prop a loudspeaker in a tree and watch the birds and monkeys as you blast out a prerecorded alarm. Compared with the many social and spatial subtleties of feeding and pair-bonding, these are straightforward encounters, easily manipulated in experiments. It is only in the last two decades that the inner complexities of alarm calls have been documented in the scientific literature, following a small number of pioneering studies in the twentieth century. If alarm calls contain within them such expansive meanings, what might the next decades reveal about the far richer sounds contained within other social signals? There is ample evidence from birds and mammals that songs carry information about the body size, health, and identity of the singer. Whether these sounds contain information that transcends the body of the singer, referring to objects outside the body as alarm calls do, is presently unknown. Might we also expand the breadth of our studies of subtle meanings in sound to include the insects, fish, and frogs? We know that some of these other species have individually recognizable sounds, but we have little idea whether these variations encode more information.
The diversity of sounds that I hear in the rain forest is partly the result of the grisly attentions of predators and parasites. Without them, trilling insects would be more monotone, and the vocalizations of birds and mammals would lack range and subtlety. Another threat to singing animals is acoustic competition. In such a loud and crowded place, the sounds of other species singing over one another are a potentially serious problem. These acoustic competitors may not dump hordes of writhing parasitic larvae or tear off your head with their hooked beaks, but if your sound cannot be heard over the din, your genes are likely to face oblivion nonetheless.
With hundreds, sometimes thousands, of species making sound in the rain forest, the masking effects of noise are severe. Animals here face challenges unlike those of other climes. In the Rocky Mountains, insects are silent for most of the year and give only weak chirps and clicks in the height of summer. They and the other mountain animals almost never overwhelm one another with their voices. The wind is the primary acoustic foe in the mountains. Even in the lush forests of the southeastern United States, in the most biodiverse temperate forests in the world, most of the year passes without intense sonic competition. Springtime birds can be voluble, but they don’t make a clamor that blocks the sounds of others. Only in the hot days of midsummer do the cicadas fill the air at amplitudes that make my ears ring, surpassing, if they were in a factory, the legal threshold for required hearing protection. In the same forests on late summer nights, katydids unite in a low pulsation loud enough to make humans raise their voices as they converse. These choruses come after the birds’ and frogs’ breeding seasons, but any other insect species trying to be heard faces a barrier of sound. A challenge that exists, at best, for only a few weeks in the temperate zone is omnipresent in the rain forest. Evolution’s response to this difficulty takes many forms, most of which promote diversification of sound.
Yelling louder is one solution to communication in noisy environments. This adjustment can happen in the moment or over evolutionary time. Birds, mammals, and frogs all sing louder in clamorous places and match their vociferousness to the amplitude of the background. Whether any insects do the same is presently unknown. Animals that live in persistently noisy places have evolved to make louder sounds at all times. Contrast the soft clicking sounds of the Putnam’s cicada that sings from solitary perches in the uncrowded quiet of Colorado mountain pine forests, with the blast of the Magicicada periodic cicada that calls from dense aggregations of thousands in Tennessee woodlands. The former is gentle, sounding like a fingernail tapping on a dry twig; the latter is almost intolerable at close range, leaving my ears with the ringing, plugged-up feeling reminiscent of the aftereffects of a rock concert. The rain forest is so loud partly because the animals are shouting over one another, often pushing the physiological limits of amplitude.
As human aficionados of quiet dining can attest, you can sometimes avoid the fracas by changing your timing. A five or ten o’clock dinner reservation yields quieter surrounds than one at seven p.m. With only twenty-four hours in the day and hundreds of species jostling for space, this strategy has its limits in ecological communities, but several species are known to rework their schedules to avoid noise. A conehead katydid from Panama usually sings at night, but shifts to daytime chirping in places occupied by another species with a similar song. Experimentally removing the competitor causes the conehead to switch back to nocturnal song. But this is an unusual example. Most insect communities have extensive overlap in the timing of the daily cycle of their songs. A finer-grained division of time is possible, though, even when animals are singing at the same time of day. Some birds and frogs space their songs to avoid overlap. By timing their song phrases to fall in the silent gaps between the phrases of other species, singers avoid masking. This strategy, though, depends on all parties singing at about the same tempo. Birds with similar sounds, therefore, sometimes listen to one another’s timing to squeeze in their songs, but other animals in the boisterous rain forest, especially insects at dusk, sing not in discrete phrases but in overlapping choruses or near-continuous trills.
Time is one way to slice the acoustic pie. Frequency is another. The lowness of the grunt of the crested owl makes it distinct from the higher squeak of tree frogs and the shrill whines of insects. By singing at different frequencies, animals might escape acoustic competition.
As I listen to the night sounds of the Amazon, it seems at first that animals have indeed divided up the frequency spectrum, making room for each species’ voice. From owls to frogs to katydids to crickets, I’m hearing sounds arrayed across a wide spectrum, seeming evidence that evolution has produced a coherent whole, minimizing competition. This idea, proposed by pioneering sound recordist Bernie Krause, is hard to test. The song frequencies of animals may differ for many reasons, not just because competition has led to divergence and, in many cases, species do in fact overlap with one another. The frequencies of animal sounds depend largely on body size, for example. The wide spectrum of calls in the forest may reflect not acoustic competition, but the evolution of different body sizes for different ecological roles. Owls call lower than hummingbirds because they have larger sound-making membranes. The sonic differences between these two groups reflect the ecologies of each—owls hunt large insects, hummingbirds sup on flower nectar—and not a competition-induced division of the sound spectrum. A red howler monkey giving a deep roar at dawn from its treetop perch weighs about six kilograms. Its body is adapted to a diet of leaves and fruit plucked from the rain forest canopy. In the wet forest at the river’s edge, pygmy marmosets trill to one another in high-pitched voices. These are the world’s smallest monkeys, weighing only a tenth of a kilogram. They feed by gouging small holes in tree bark, then lapping at the oozing sap. The marmosets chitter and purr back and forth as they work their trees. Just as a violin cannot make as low a sound as a bass, a pygmy marmoset is physically unable to make as deep a sound as the howler monkey.
The observation that a species-rich place like a rain forest is filled with diverse sounds does not, therefore, amount to evidence for the idea that acoustic competition caused this diversity. A more rigorous test is to ask whether species that sing in the same places have more divergent song frequencies than we’d expect by chance.
A study of the dawn chorus of birds in the Amazon made such a test and rejected the idea that competition has caused acoustic differences among songs. The researchers analyzed sounds from samples of the dawn chorus recorded in more than ninety locations, comprising the songs of more than three hundred bird species. Amazonian birds, the study found, tend to sing at frequencies and speeds that allow for the best transmission of their songs through dense vegetation, a somewhat lower frequency and slower pace than those of temperate birds. With so many species singing in this fairly tight range of sounds, we might expect intense competition for acoustic space, perhaps driving apart the frequencies of species that sing together. If so, birds singing at the same time and place should have lower overlap in their songs than stimulated “communities” of birds randomly picked by the experimenters from their database. But the birds’ songs showed the opposite. Species that sing together have songs that are more similar to one another than we’d expect by chance. This was true for all measures of acoustic structure: pace of delivery, highest frequency in the call, and range of frequencies or bandwidth of each call.
Individual birds singing in the Amazon sometimes adjust the second-by-second timing of their songs to avoid overlap, but at a larger scale there is no evidence that competition has caused species to diverge in the structure of their calls. Instead, the forms of the songs seem drawn together into clusters. Two factors may cause this sonic grouping. First, closely related species often share both habitat preferences and the structure of their songs. Small flycatchers are drawn to forest patches rich in flying insects. Large parrots gather in fruit-rich forests and antwrens forage where insects are abundant. Closely related hummingbirds feed from flowers growing in the same trees. Shared pedigree, and thus tastes for food and habitat, brings the sounds of closely related species into the same places. Second, birds of different species may be linked in a communicative web. When competing species share and understand one another’s sounds, they can communicate swiftly and unambiguously. This allows them to efficiently mediate their jostling for food and space, and quickly alerts them to incursions from outsiders. Thus, shared song characteristics paradoxically link competitors in a cooperative network. In the Amazon it seems that the more intense the territorial competition among bird species, the more similar are the structures and timings of the birds’ acoustic signals. The need for mutually shared communication channels among competitors is not unique to birds. Governments in Moscow and Washington are connected by hotline, commercial competitors agree on aesthetic conventions for branding and the form of retail spaces, and competition within professions is mediated through the use of shared jargon.
We have mixed results from studies of vocal competition in animal species other than Amazonian birds. The eighteen most abundant cricket species in a forest in Panama do seem to have divided up sound frequencies to avoid overlap. A survey of eleven frog choruses found three in which competition appeared to have led to divergence of frequencies, but the others showed no such evidence. Birds in the temperate forests have widely overlapping song frequencies, although they separate their songs through their timing and spacing. Acoustic competition, then, seems at best only an occasional factor in the diversity of sound frequencies that we hear in natural settings. And in what is arguably the most acoustically crowded place on Earth, the dawn chorus of birds in the Amazon rain forest, sounds of species that sing together have converged, not diverged.
