Listening for Clues to Coyote Well-Being

Across the Duke Forest’s 7,100 acres, sounds tell stories. A breeze rustles the leaves and insects scatter to safety. Water trickles through a nearby creek, filling a pool where a crayfish swims. Squirrels chatter as they leap from limb to limb, birds exchange songs across the canopy, and frogs call at dusk. Overhead, the rumble of a jet engine joins the distant hum of traffic on nearby roads.

Together, these sounds form the Forest’s soundscape. But what might we learn by listening closely to its animal inhabitants? Could a call reveal more than an animal’s presence? Could it offer clues about how that animal is faring?

Caroline Rowley, a PhD candidate in the Environment Program at Duke, is exploring those questions through acoustic monitoring. Rowley has established a network of recording devices throughout the Duke Forest. The devices capture ambient sounds, which Rowley then analyzes for the vocalizations of one particularly adaptable resident: the coyote.

Over the past century, coyotes (Canis latrans) have expanded beyond their historic range in the western United States and Mexico into much of the rest of North America and parts of Central America. The decline of larger predators, changes in land use, and coyotes’ behavioral flexibility have helped enable that expansion.

Their success raises a question: Does a thriving species necessarily mean thriving individuals?

Coyotes, particularly those living in urban environments, can face challenges that affect their quality of life even as their populations persist or grow. Rowley’s research asks whether their calls, known as “yip-howls,” could help researchers understand those experiences.

Audio: A recording of coyote yip-howls captured by Rowley’s equipment in the Duke Forest.

Using the Five Domains Model of animal welfare, the project will investigate whether acoustic features of coyote vocalizations can provide clues about nutrition, physical environment, health, behavioral interactions, and mental state. The goal is to explore whether listening can help researchers assess the well-being of individual animals.

“Being able to quantify an individual’s animal welfare allows us to understand how coyotes inhabiting different landscapes and levels of human-influenced activity are impacted by their home ranges,” says Rowley. A coyote living in a fragmented landscape, for example, may need to travel farther between habitat patches, increasing its energy demands. Because yip-howls are loud, prolonged, and physically demanding, the quality of an animal’s call might signal difficulties in meeting those demands. Understanding these connections could help guide interventions that support coyote welfare and prevent negative interactions between people and wildlife.

Advances in bioacoustics—the study of sounds produced by living organisms—and machine learning have also shown that researchers can distinguish individual coyotes by their howls. That opens the possibility of using repeated recordings to paint a fuller picture of an individual animal’s health and wellness. 

Identifying who is calling, however, is different from understanding how that animal is doing. Rowley’s project will investigate whether measurable features of calls can serve as reliable indicators of welfare. Establishing those connections will be essential before researchers can interpret a vocalization as evidence of an animal’s condition.

Previous research has shown that territorial calls can serve as “honest signals,” conveying reliable information about an animal’s intentions, emotional state, or physical condition. “My goal is to focus on parts of their calls that have been salient in previous studies, such as fundamental frequency and length of their calls,” Rowley says. To help interpret those acoustic features, they will also collect complementary data from camera traps, scat samples, and collared coyotes, connecting the recordings with information about the animals’ health, movement, nutrition, and social behavior.

Pictured: A coyote carrying a pup in its mouth as it moves through the Duke Forest.

With recording devices at five sites, the Duke Forest provides a setting for Rowley to investigate these questions across multiple habitat types. The recordings capture coyote calls within a broader soundscape of wildlife, weather, and human activity.

“Having Duke Forest sites for part of my research enriches my work tremendously; the forest enables me to ask questions about Durham and Orange County’s anthropogenic disturbance gradient, given its size and varying levels of human impact,” says Rowley. “Additionally, the large body of work done in the Duke Forest allows me to explore more nuanced questions about those field sites and have a greater understanding of the landscape.”

The project’s potential reaches beyond identifying coyotes or documenting where they occur. If vocalizations prove to be reliable indicators of welfare, acoustic monitoring could give researchers another way to understand how animals experience their surroundings and respond to environmental pressures.

For now, the work begins with listening. Among the rustling leaves, birdsong, and distant traffic near the Duke Forest, coyote calls may hold clues to a question that population numbers alone cannot answer: How are these animals doing?