Sounds Wild and Broken, page 29
This work is on the leading edge of a revolution in how scientists listen to the world. In 2000 and 2001, I tromped through the forests, marking off each bird as I heard it, much as thousands of other field biologists have done all over the world in our attempts to measure, understand, and ameliorate the many effects of humans on other species. But these surveys are time-consuming and sample a minuscule portion of the soundscape.
Extended sound recordings, analyzed by computer, offer a complement to more traditional forms of field study. In addition to larger time samples and greater statistical power, the recordings solve problems inherent in surveys that rely on field observers. We all differ in our hearing abilities and identification skills, adding variability to the quality of observations. Naturalists and scientists also have taxonomic biases. It is not hard to find people who can name every bird sound in their region. But few people can identify by ear all the insects, especially in tropical regions. In addition, not all tropical species sing at once in a narrow breeding season as they do in temperate regions, necessitating surveys over many months. Scientific study quickly hits limits of human capacity and knowledge.
By processing vast troves of digital sound data, algorithms extract patterns and trends unknowable by older scientific methods. In the last ten years, the price of recording devices with capacious memories has dropped. AudioMoth, for example, is smaller than a pack of cards and can record continuously for several days or, if programmed to record only a few hours each day, more than a month. The device and its supporting software are open-source, its blueprints and code freely available to all, and, for those who prefer not to solder their own gear, costs only seventy dollars premade.
These technological advances have led to thousands of research projects that typically fall into one of two categories, reflecting two different types of software analysis. Some software is programmed to sift through the recordings, plucking out specific sounds. Managers in Cameroon’s Korup National Park used a grid of recording devices to measure gunshots and the effectiveness of antipoaching patrols. Hydrophones in Massachusetts Bay have tracked cod-spawning aggregations using recordings of the fish’s mating chorus, pinpointing the most productive sites and uncovering declines. Rare and threatened species such as elephants in dense African rain forests, fish in tropical wetlands, and birds in Puerto Rican forests have all been studied with the help of electronic ears deployed throughout their habitats. Species like bats and insects whose sounds are too high for human ears are also readily tracked with electronic sound recorders. Once detected and classified by the software’s algorithms, sounds from multiple recorders can estimate changes in behavior and population size, or be compared with the data on other recorders to estimate the location of the animal.
The other approach is the one taken by Burivalova and her colleagues. Instead of picking out the identity of individual species, the software scans and analyzes the entire canvas of sound, measuring saturation, loudness, and frequency to find patterns across space and time.
No software can yet identify all the singing species in one place and thus dissect all the component parts of a soundscape, although some can simultaneously pick out a couple dozen voices. When I stand in a Tennessee forest and name all the birds, frogs, squirrels, and insects singing around me, and recognize the meaning and emotion in a human companion’s voice, I’m outperforming the most powerful “artificial intelligence.” Perhaps future technologies will surpass us, but for now, humans can still beat a computer in a contest over sonic pattern recognition. This is a reminder of the potential cost of listening through computers. As is true in so many parts of our lives, our time and attention are drawn by these new technologies inward into the human world of electronica, rather than outward into direct sensory experience of the living Earth. Even the name of the new technique, passive acoustic monitoring, suggests a withdrawal of active human senses.
In addition to their potential utility for researchers and managers today, soundscape recordings also create an archive for the future, digital memories of how Earth sounds today. Generations to come will listen with questions we cannot imagine. Every stored recording is a gift to tomorrow.
The soundscapes of years to come will be missing some of Earth’s voices. Part of what we record is thus a preamble to extinction. Digital sound files will help us to grieve. They will also partly inoculate against the problem of the “shifting baseline,” the diminishing expectations of each generation as it gets used to a world less filled with song. My grandfather told me how he missed the bird- and insect-rich sounds of the fields and towns of his youth in northern England. Without his story, I’d encounter modern soundscapes as “normal.” Each sound recording is an anchor against this tide of forgetting.
Most automated soundscape recordings to date have been short-term and focused on particular questions and regions. But larger-scale archival work is also commencing. The Australian Acoustic Observatory, for example, is installing recorders at one hundred sites across the continent, with the aim of recording continuously, initially for five years, and making the stored sound freely available. These electronic remembrances are a technological complement to the stories we must tell one another. Data needs accompanying narrative. Hopefully, if we act now, our legacies will convey not only loss but evidence of renewed flourishing in years to come.
Despite their utility as time capsules for the future, I was skeptical about whether these technologies could help to conserve forests. Another gadget, I thought, great for naturalists and academics in pursuit of new projects, but of little relevance to slowing the ravages of deforestation. After all, we know the nature of the problem: millions of hectares of tropical forests are lost annually to fire, saws, and bulldozers. A bleeding, fading patient needs immediate practical help, not an ever more precise and technologically sophisticated diagnosis.
Conversations with Burivalova, the project’s leader, and one of her coauthors, Eddie Game, lead scientist for The Nature Conservancy’s Asia Pacific region, showed me otherwise. They explained how extended field recordings and computer analysis of large acoustic datasets could both guide on-the-ground conservation and attract more funding for this work. With other researchers, Burivalova and Game have also deployed recording devices to help people in Papua New Guinea to track biodiversity in forests and agricultural areas, information that then informs local decision-making about future uses of the land.
“It worked out better than I thought it would,” Eddie told me. “In Borneo, the recordings are more sensitive to differences in forests than I had expected. . . . We know from our previous work and research by others that well-managed logged forests can have about the same gross biodiversity as protected forests. But this masked local differences and uniqueness in protected forests. Previous research mostly used field surveys of birds and mammals and missed these fine-grained differences. In Papua New Guinea, the sound recordings give local people a powerful and relatively cheap way of monitoring their forests.
“We’re an organization that prides itself on having evidence that what we’re doing is effective. When we speak to academics, they see this work as really boring science, but for us it is very meaningful to know that what we consider to be better land management practices result in richer soundscapes.” He explained that the variable sonic texture and local differentiation of unlogged forests suggest that dividing logging into several small areas rather than one large one could have a lower impact, allowing local differences to persist.
“How can we help the logging industry be more friendly to biodiversity?” Burivalova asked. “Even companies that are interested in being more environmentally friendly cannot do much in terms of biodiversity monitoring. It’s too expensive and difficult. Acoustic recordings could give them an easier way to gauge how they are doing.”
To those steeped in the antilogging rhetoric from parts of the environmental movement, talk of conservationists working with timber companies in Bornean rain forests might seem wrongheaded. In the United States, the ravenous excesses of the timber industry have provoked a strong counterreaction. The Sierra Club, for example, opposes commercial logging on federally owned lands, even those expressly created with the aim of supporting public oversight of forestry. Loggers reliably appear as villains in North American forest-infused fiction and nonfiction alike.
But the chainsaw can paradoxically be the salvation of the forest. In Borneo, selective logging removes large, commercially valuable trees. The rest are left in place because they are either too small, or not valuable, or legally protected. These “secondary” forests—those that have been logged, often two or three times—harbor many of the same species as the primary forest. Such logging certainly has ecological costs. Some species are lost, especially those like woodpeckers and fruit-eating birds that specialize on the largest trees. Logging roads can increase erosion and serve as conduits for the arrival of people seeking land to clear for smallholdings. But if done right, logged forests regrow. Four hundred million years of evolution has taught forests resilience. Given a chance, biodiversity surges back. In Tennessee, selectively logged forests have high bird diversity, but monoculture plantations do not. In Borneo, secondary forests are havens for indigenous species when compared with industrial-scale oil palm and pulpwood plantations. Replicated bird surveys in Malaysian Borneo, for example, found that the numbers of threatened bird species were two hundred times lower in oil palm plantations than in selectively logged forests. Even in the “wildlife friendly” plantations that included fragments of remnant forest within the plantation, the abundance of these birds was sixty times lower. Plantations are also poor habitat for frogs and insects.
Both Burivalova and Game also emphasized in our conversations that the wider context of surrounding land is very important. A secondary forest ringed with plantations is biologically impoverished compared with one embedded in a forested landscape. A primary forest in a sea of secondary forests likely has a more thriving ecological community than one hemmed in by plantations.
Logging provides livelihood for local communities, work and income rooted in the regenerative power of soil and trees. Oil palm plantations and mines also provide income, but they do so at a greater cost to the fertility and diversity of the land.
We are not creatures disembodied from needs for food, energy, and shelter. Wood can be renewable. Fossil fuels, steel, plastic, and concrete generally are not. To lock up many forests in “protected areas” free of human use, then, is to exile ourselves from the community of life, forcing us deeper into unsustainable relationship with synthetic materials or forest products shipped in from elsewhere, imposing the costs of our consumption on people and forests out of range of our senses. The question should not be whether we cut trees, but where and how we should do so. We certainly need extensive areas set aside, free from the saw. We also need policies and on-the-ground enforcement to bar rapacious cutting that degrades the land. But a thriving future also requires that we participate in the forest community as all other animal species do, as consumers. This is a matter of ecological and economic realism. Our lifeblood is drawn from the Earth. People need work. The oft-cited alternative to extractive use of forest products—ecotourism by wealthy foreigners flying in from overseas—helps in some areas but spurs an increase in deforestation rates in others, is not a viable source of income for local people in most parts of the tropics, and assumes that ever-increasing international travel by the wealthy is sustainable.
In the future, sound recordings might also serve to strengthen monitoring by governments, local communities, companies, and organizations that try to monitor and “certify” the ecological soundness of wood and other products.
At present, forest certification schemes use crude measures of “sustainability” and “responsibility.” Inspectors spend limited time on the ground and check off relatively easily observed indicators: are roads built to minimize erosion, are workers wearing safety gear, does the map pinned to the supervisor’s office wall accord with the management plan, is land tenure clear, are known special areas like streams and wetlands protected, and does the written plan seek long-term viability of the forest? These are important questions, but they do not assess the presence of most forest species, let alone their well-being or changing fortunes. Soundscape recordings could, through the intermediaries of technology and statistics, elevate the voice of the living Earth community. The thunderous diversity of rain forest sound would then meet silent piles of human paperwork. Out of this incongruous union, a more vibrant future for all might grow.
Beyond their practical importance in land management, recordings can also spirit the forests’ voices up over the Bornean forest canopy—south across the Java Sea, north across the South China Sea, east across the Pacific Ocean—into the ears of those of us who need to hear. Donors, policy makers, and grant makers hear the unearthed sound and are moved to act. The rest of us, those without the leverage of unimaginable wealth or political power, also understand through these sounds: we are connected. One-third of the plant photosynthesis that supports life on land happens in tropical forests. The wood in our houses, paper, and furniture are often rooted in Southeast Asia. The palm oil in cosmetics, processed food, biodiesel, and farm animal feed is grown on former rain forest land. But we have broken all direct sensory connection to these forests that sustain us. Sound can bring us partway back to embodied sensory understanding. We might then make wiser choices about how and whether to use the products of forests far over the horizon rather than the materials and energies close at hand.
Eddie leans forward. “People really get that sound is linked to biodiversity. I’ve had more substantive conversations about forest monitoring with this sound data than with anything else. They experience the forest. The thing that blows their minds is how noisy it is, constantly.”
He pauses, eyes flicking upward as he searches for words.
“Through the sound, they get close to this almost undefinable property, ‘biodiversity,’ closer than any metric, or graph, or photo.”
* * *
—
There is another “algorithm” that can “process” thousands of hours of “data” about the changing forest: lived human experience. Almost all tropical regions are home to people whose ancestors have lived within the forest for centuries or millennia. Many of these cultures are now besieged. Forest conservation is therefore a matter of human rights.
In the Western tradition, forests are often seen as places of darkness, home to brigands and exiles. Wolves of all kinds. The edges of civilization. The forest is umbral, full of confusion. Dante lost the right road in a dark and savage forest. Children become disoriented in the forests of the Brothers Grimm. Ever since the Neolithic agricultural revolution, we have cleared trees to make way for pasture, crops, and towns. Even when Western cultures desire to manage land for wood or forest conservation, they usually do so as enterprises that exclude people from the land. In the United States, for example, national forests and national parks were established by expelling from within their boundaries every single human inhabitant, save for those who retained private “inholdings” or employees in park compound housing. Contemporary US state tax incentives for keeping land in “forest” often disappear if people live within the forest. In the official statistics of both the US government and the Food and Agriculture Organization of the United Nations, forests that have houses built among them or in which people grow food count as “lost” forest, but tree plantations and barren ground left after clear-cutting count as “forest.”
When this Western mindset meets tropical forest, human calamity often ensues. Governments declare forests terra nullius, empty land, opening a “frontier” to colonization of land that is home to people whose cultures have lived there for centuries or millennia. Corporations—both for-profit extractive industries and nonprofit conservation organizations—take title to land and drive out its human inhabitants. These are not only injustices of yesterday, enacted in the age of wooden ships, muskets, and disease-poisoned blankets. Indigenous cultures today are under sustained attack, their lands and lives taken by force and murder, and by the violence of the laws of nation-states and the global economy.
In Kalimantan, the Indonesian portion of Borneo, an alliance of fifteen organizations representing indigenous communities submitted an urgent appeal to the United Nations Committee on the Elimination of Racial Discrimination in 2020, stating that “vast encroachment on and takings of indigenous lands for road-building, plantations and mining” were underway, “all of which threatens to cause imminent, gross and irreparable harm to the Dayak and other indigenous peoples.” In Brazil, also in 2020, representatives of dozens of indigenous groups vigorously opposed new laws that would further “open indigenous lands up to exploitation.” Deforestation rates are now rapidly increasing after years of decline in Brazil, reaching their highest levels in a decade in 2020, more than eleven thousand square kilometers lost. Indigenous leader Célia Xakriabá of Brazil says, “I can hear the song of the birds now, but it’s also a song of misery, of sadness, because most of them, they are alone. They have lost their partners. . . . And we, the indigenous are becoming more alone, because they [miners, loggers, ranchers] are taking people from us.” In the Democratic Republic of the Congo, Rainforest Foundation UK found in 2019 that people “living around Central Africa’s largest national park have been subjected to murder, gang-rape and torture at the hands of park rangers.” The “widespread physical and sexual abuse being inflicted by ‘ecoguards’ ” occurred in conservation parks that were first established by driving indigenous peoples from the forests.
