Sound Tools for Wildlife Research

By Matthew Underwood

In 1665, English scientist Robert Hooke wrote of the new worlds he glimpsed through his microscope. In every speck of earth, he claimed, “we now behold almost as great a variety of Creatures, as we were able before to reckon up in the whole Universe itself.” A tireless promoter of technologies that could extend the power of the human senses, Hooke believed that investigators armed with proper tools could “excel others in their Observations, and Deductions, almost as much as they do Beasts.” Today, the Lab’s Bioacoustics Research Program (BRP) is developing technologies and methods that are providing startling new insights into animal behavior and more effective techniques for tracking and counting wildlife populations.

One such technology—as simple in concept as Hooke’s microscope—is the use of strategically placed microphone arrays to study wildlife. The microphone arrays in use today grew from an effort begun in the early 1980s to count migrating bowhead whales in northern Alaska. These massive cetaceans are difficult to count by sight alone, because they often swim beneath the Arctic pack ice and can slip unseen past whale census points. BRP director Christopher Clark came up with a means of counting these whales by recording their vocalizations with a series of underwater microphones (called hydrophones) suspended through holes in the ice, spaced one kilometer apart. Each vocalizing whale was recorded simultaneously at varying angles and distances by three to five different hydrophones. The data were later digitized for computer analysis with a program called Canary, developed by BRP. Canary calculated each whale’s location by determining the difference in time (measured in milliseconds) that it took for a vocalization to reach each hydrophone, providing precise information so researchers could plot each whale’s position and find out how many were present.

Times have changed, and the technology has flourished. BRP researchers have developed a new, even more effective sound-analysis program called Raven, and they now make digital recordings right in the field. And the cumbersome cables are disappearing. Today BRP whale researchers use arrays of ocean-bottom Autonomous Recording Units (ARUs) called “pop-ups,” engineered on the principle of “set it and forget it.” Each unit contains batteries, a microphone, and a computer hard drive—all encased in a 17-inch, pressurized sphere attached to an anchor. Dropped in the ocean and left for months, they record the noises of the deep directly to disk. When prompted, each pop-up detaches from its anchor and floats to the surface, where it can be picked up by researchers.

Photo credit: Jon Reis

Kurt Fristrup, assistant director of the Bioacoustics Research Program, helped design “pop-up” units that record the sounds of whales from the bottom of the ocean. The units pop up to the surface when they are finished recording.

Currently, BRP researchers are using pop-ups in Cape Cod Bay to monitor northern right whales, which are listed as endangered species and are prone to fatal encounters with various ships and fishing gear. The data the ARUs gather reveal where the whales have traveled and when. According to BRP engineer Tom Calupca, researchers may eventually develop an auto-detection program for the whales, using ARUs mounted on buoys. The ARU would pick up a right whale’s vocalizations, pinpoint the animal’s location, and transmit the information instantly to the Lab through a cell-phone modem. The Lab could then provide a live-action map to ship traffic controllers, indicating the present location of every northern right whale around Cape Cod and taking the guesswork out of the effort to keep boats and behemoths apart in the bay.

Photo credit: Melissa Groo

Led by the matriarch, a forest elephant family emerges into the Dzanga clearing in the Central African Republic. Microphone arrays have helped show that adult females vocalize extensively to keep their families together.

Microphone arrays may also be the key to keeping forest elephants and farmers from clashing in West Africa. In Ghana’s Kakum National Park, the Lab’s Elephant Listening Project recently tested its innovative method of using land-based ARUs to census the park’s elephant population. (To census elephants in this densely forested area, researchers previously had to count the number of fresh elephant dung piles along forest transects.) An outgrowth of that project may help to solve the problem of elephants raiding crops along the small park’s borders. The Elephant Listening Project is currently testing an “acoustic net”—a string of microphones that could detect the vocalizations of approaching elephants and alert game managers before conflicts flare up.

Photo credit: Melissa Groo

Engineer Eric Spaulding climbs a tree to check on an Autonomous Recording Unit. The units must be attached to trees to prevent the elephants from damaging them.

As microphone array technology improves, the complexity of animal communication is becoming ever more apparent. Until recently, studying bird communication meant recording up to two birds—usually males fighting over territory—with one microphone. BRP biologist Sandra Vehrencamp has been using microphone arrays to record entire neighborhoods of Banded Wrens in Costa Rica simultaneously and to investigate the social function of the birds’ raucous, multiplayer dawn chorus.

Each male wren employs around 20 song types in his personal repertoire, many in common with his neighbors. Beginning around 5:00 every morning, the males begin an intense, 45-minute jam session, passing riff after riff from bird to bird to bird. Vehrencamp can record up to 1,000 songs in one morning from four or five birds whose territories she surrounds with an array. Location analyses reveal the territory where each song originated, identifying the singer. Vehrencamp’s goal is to amass a complete record of every song those four or five birds sang, so that the song type, the bird who sang it, and the next song are all arranged in chronological order. The work is time-intensive, first requiring months to build baseline reference recordings of each individual’s entire repertoire. Verifying song-type matches is currently done manually (automated software is in the works), so indexing four mornings’ choruses in this tiny study area has taken nearly a year.

Yet even this small amount of data offers a view into wren social organization. Already, Vehrencamp has noticed that the birds listen to all of their neighbors simultaneously, “making incredibly complex decisions about what song to sing at any moment in time, depending on who they’re listening to and who they want to answer.” The wrens nest in acacia trees, which grow sparsely in the forests they inhabit, so nesting real estate is a hotly contested commodity. In order to find a suitable nesting site, Vehrencamp explains, a wren sometimes has to move in on another bird’s territory. “The territories are dynamic, and that’s why we think these are dawn chorus singers. We know it’s a male-male thing, and every morning they’re getting up and sorting out where their boundaries are.” Researchers will pay close attention to this relationship as they analyze more data, looking to see whether the dynamics of the dawn chorus change as the territory boundaries shift.

Marc Dantzker (the Macaulay Library’s visual curator) and Gail Patricelli hope that using microphone arrays will help them to understand more fully what function a male sage-grouse’s ability to project a sound toward a particular female plays in the mating display. Dantzker has used a microphone array for mapping acoustic radiation patterns of male grouse to see where the birds project their various mating display sounds, some of which are directed with remarkable accuracy. He and Patricelli hypothesize that the most directional portion of the display, a high-pitched whistle that can be heard loud and clear only off the bird’s shoulder, is also the most important—a sort of pick-up line intended for the ears of one female only. Dantzker and Patricelli plan to test this by measuring the male’s display from the female’s perspective using a microphone mounted on a robotic female sage-grouse. By measuring the male’s acoustic radiation pattern with a microphone array while also measuring what the female hears, they will be able to test whether some males are better than others at tracking females and directing sound to them, and how that affects their mating success.

Understanding sage-grouse mating habits, determining the territorial dynamics at work in a neighborhood of Banded Wrens, and keeping elephants out of farm fields and right whales out of fishing nets seem to be fundamentally different scientific efforts, each requiring a window onto a different animal’s world. Yet in each case, that window is provided by a single, elegantly simple technology, pioneered by the Lab’s Bioacoustics Research Program. Microphone arrays help researchers to discover the complex worlds that these animals experience and also to preserve those species that are in peril before they disappear forever.