Urban Wildlife on Camera

By Susan Dirr, MA Student in Project Dragonfly, Miami University

Camera traps are cameras which are remotely activated by sensors to take still or video images. These tools have become increasingly popular in wildlife research over last twenty years, and their applications continue to expand as both technology and methodologies for camera trap research improve. All camera traps are equipped with a sensor, typically an infrared sensor which recognizes the temperature difference of an animal against its background. The area covered by the sensor is called the detection zone, which may be smaller than the camera's field of view. In order to illuminate the subject, the camera is equipped with a flash, which may include a white flash and/or an infrared flash for night images (8).

Camera trap (CT) technology has several advantages compared to direct observations of animal behavior, including avoiding the influence of a human observer on behavior and reducing observer bias. In addition, camera traps can gather data for multiple studies simultaneously and the data can be independently analyzed by multiple researchers (3).

However, CT studies are limited by the ability to process the often vast amounts of data, which can be a time and labor intensive process. Advances in machine learning to automatically categorize images as well as use of citizen scientist volunteers to process images may reduce this limitation. Another concern expressed by researchers is that the light and sounds emitted by cameras traps may interrupt wildlife behavior, although this effect has been reduced with newer models that use less detectible flash technology. Finally, camera failure or theft can cause data loss, especially when cameras are checked infrequently (3).

For many researchers, the benefits of CT outweigh the drawbacks, opening up exciting new possibilities in animal behavior research. Camera traps have been used to explore a wide array of questions including:

• • Presence of rare or hard to find species

  • Abundance and density estimates

  • Patterns of activity

  • Disease monitoring

  • Diet

  • Effectiveness of conservation measures

  • Nest predation

  • Response to human disturbance

More than half of the world's population lives in urban areas, with this proportion expected to increase in the next decade. Understanding the ecology of urban areas is essential to maximize their conservation potential, minimize environmental harm, and mitigate human/wildlife conflicts. In addition, urban ecology provides an opportunity to connect people with the nature, which can lead to the development of pro-environmental behaviors. Camera trap technology is an essential tool in meeting these goals.

With a population of 2.7 million people, Chicago is the largest city in the Midwest and the third largest in the country. Built in a region formerly dominated by prairie and oak/hickory forests, the land within the city limits has long since been converted for residential, commercial, and industrial uses. While Chicago is a dense urban region, there is a large park system within the city as well as a forest preserve system with properties throughout the county that provide wildlife habitat. Camera traps placed throughout the region have documented wildlife including coyotes, racoons, opossums, fox, skunks, woodchucks, white-tailed deer, mink, weasels, chipmunks, muskrats, beavers, tree squirrels, and flying squirrels.

One project that has captured the attention of Chicago's public is the Chicago River Otter Project. Using camera traps with scent lures, researchers have documented the return of the elusive river otter to the Chicago region. The camera traps have been able to identify the specific locations used by river otters. Researchers publicized their efforts on social media, sharing images from the camera traps. These images of otters help build public awareness and increase care for the Chicago River and its tributaries.

Camera traps may also be used to understand how wildlife respond to human-altered landscapes. For example, researchers in Chicago combine camera trap data with satellite images to investigate how wildlife activity levels change in response to artificial lighting in the urban environment. Researchers found that urban lighting is connected to altered activity patterns in both nocturnal and diurnal species, and concluded that artificial lighting is contributing to the degradation of wildlife habitat in urban areas (9).

How does camera trap technology help you answer your research questions?

"Camera traps allow us to observe animals in ways we have never been able to before. Newer versions have no flash and make no detectable sound, thus taking photos without disturbing animals. From these photos we can learn about how species interact with one another, how they share space, what times of day they are active, and what types of habitat they prefer and don’t prefer (just to name a few). Our colleagues have even learned about the presence of disease by assessing body condition from photos."

What are some of the challenges you have encountered in using camera traps?

"The main challenge lies in reviewing the photos. Even if you do a careful job of setting up the camera, there will often be false triggers—a blade of grass moving back and forth by the wind, sunlight hitting the camera just right, and occasionally animals moving too quickly through the frame to be identified or a camera simply malfunctioning. Although there are many ongoing attempts to use machine learning to identify animals in photos, or at least sort through the photos that are empty, they aren’t widely useful across environments yet, so there is still a lot of time dedicated to reviewing photos. But, we are getting very close to being able to effectively use machine learning to assist in the photo review process, which is exciting!"

How would you like to see camera trap technology used in urban ecology research in the future?

"I’d love to see ways for cameras to be used across more taxa (right now mainly limited to terrestrial mammals). It would be really useful for cameras to better differentiate between blowing leaves/vegetation and animals and to reduce the over triggering (admittedly this would probably be reduced with more expensive cameras). It would also be really neat for cameras to include the data processing internally so you would simply download the processed data rather than the photos themselves (hopefully with some cool highlight photos). I’d also love to see cameras used for population estimates (right now that is difficult with mammals that are hard to identify individually, but some smart stats folks are working on it)."