Is A Drone With A Thermal Camera As Good At Locating Grassland Songbird Nests As Traditional Search Methods?

As captured in the phrase, “the canary in the coal mine,” songbirds are known to be indicators of ecosystem health. Detecting downward population trends in songbirds can be a clue to biologists that an ecosystem is declining in quality or imperiled. However, population trends can be difficult to assess because songbird abundance often fluctuates at particular locations for a variety of reasons which are not necessarily reflective of the overall population trend.

Quantifying reproductive output is a more direct technique for monitoring songbird population persistence. But doing so requires locating nests and observing the number of young successfully fledged. This can be quite challenging for researchers. For example, in the grasslands of North America, songbird nests are often difficult to locate because they are placed deep within tall vegetation — and adults often reach them by running under the cover of grasses.

The most common methods for nest searching involve time-consuming observation of behavior and plodding through the habitat on foot to scan visually for nests and flushing adults. These techniques often leave behind trampled vegetation, human scent trails that cue predators to nest locations, or even direct damage to the nest or fledglings themselves.

However, new technologies are changing the way ecological research is performed, and they could even improve our ability to locate songbird nests. Previous studies have used handheld thermal imaging devices successfully to locate and observe grassland nests. Thermal cameras have been paired with unmanned aerial vehicles (UAVs), or drones, in police search-and-rescue operations, home inspections, and animal population surveys. However, they have not previously been paired to execute nest searches in grassland habitats.

Fig 1. Field sparrow nest visualized from a) typical human perspective, b) close-up color imagery, c) from ~4 m overhead with UAV-mounted thermal imagery. Figure republished with permission from Elsevier from

This study sought to compare the effectiveness and efficiency of traditional (on-foot) searches and UAV-assisted thermal searches. Nest searches were conducted in three grasslands ranging from 22 to 30 acres at the Pierce Cedar Creek Institute in Southwest Michigan, USA. Two of these were restored prairies, with many warm-season grasses such as big bluestem (Andropogon gerardi) and switchgrass (Panicum virgatum), while the other was composed of both warm and cool-season grasses. The target species, the field sparrow (Spizella pusilla), was selected because it is abundant in these prairies and it is known to build cryptic nests on or near the ground.

Active sparrow territories were first marked by a scout, who placed a flag near the center of the area being used by a sparrow pair and then added four corner flags marking the edges of the observed territorial activity. Each territory was then searched on consecutive mornings by two teams consisting of two individuals; one team used traditional methods to locate sparrow nests while the other team used a DJI Inspire 1 drone fitted with a FLIR XT Zenmuse thermal camera. Information was not relayed between the two teams or the scout to ensure that researches were always blind to the results of the other team.

Fig 2. Territory search layout. The yellow flag marks the center of activity with pink flags marking the edge of sparrow activity at the four cardinal directions. Searches were performed in the numbered order, consisting of an observation period where bird behavior was observed, followed by an active search either on foot or using the UAV. Figure republished with permission from Elsevier from

Both search methods were largely successful in locating field sparrow nests. UAV-assisted searches located the target nest in 17 of 20 territories, while traditional searches located the target nest in 18 of 20. The paired searches were equally successful in searches that started at 6 am (13/15). A smaller number of searches were conducted at 9 am (n=5). The UAV-assisted method failed to locate one nest during this later time period. UAV searches were found to be more difficult later in the morning, as the hot spots on the real-time thermal feed that indicated potential nests were less distinguishable amid warmer vegetation. Overall, UAV-assisted searches were 14.25 +/- 17.98 minutes faster than traditional methods, although the result was not statistically significant.

Fig 3. A) Mean search times by treatment and time block, including the percentage of nests successfully located and sample size (n). B) Pairwise comparison between tradition and UAV searches. Search times were capped at 120 minutes. Figure republished with permission from Elsevier from

Since UAV-assisted searches were statistically similar in efficiency to traditional search methods, they do present potential advantages. Primarily, UAV-assisted searches greatly minimize the need for people to enter the habitat on foot. Reducing the level of disturbance associated with research techniques is always valuable. In addition, future advances in drone technology, such as a dual thermal and visual video feed, could make nest monitoring on foot entirely obsolete because the image from a drone could be used to decipher between eggs and hatchlings and even count the number of individuals in a nest. UAV-assisted searches may also prove to be valuable in otherwise inaccessible habitats, such as wetlands or cliff-dwelling colonies.

While this study reports on the benefits of using drones in bird research, biologists must also continue to examine the negative impacts of drones. UAVs may introduce a new source of disturbance due to the noise and wind generated from the blades. Indeed, field sparrows in the current study generally flushed upon the approach of the drone. But, adults always returned to the nest site shortly after and never abandoned nests after drone surveys were completed. In summary, the results from this study indicate that UAVs are at least an equally efficient alternative to traditional nest searches that may reduce the impacts on nesting habitats. It is also quite likely that this technique will continue to improve with coming advances in drone and thermal technology.

These findings are described in the article entitled Real-time thermal imagery from an unmanned aerial vehicle can locate ground nests of a grassland songbird at rates similar to traditional methods, recently published in the journal Biological Conservation.

About The Author

Chace Scholten

Chace Scholten is a research scientist at Calvin College in Grand Rapids, Michigan.

Darren Proppe

Dr. Proppe is interested in how animals respond to novel environments – especially those altered by human development. His research focuses primarily on the responses of songbirds to the increasing levels of anthropogenic (human caused) noise that overlaps their vocal signals. Since these overlapped signals are less detectable, he is investigating mechanisms that species employ to increase the effectiveness of their signals in areas of high noise. He is also examining the negative impacts of noise on species diversity and fitness, and searching for potential tools to mitigate these negative effects.

Speak Your Mind!


How Tall Was Jesus: What Science Says

Many people wonder how tall was Jesus, while this is impossible to be 100 percent sure, below we estimate that Jesus could have been much taller than the average male today. Here’s the deal: No one knows exactly how tall Jesus was, but with the details below, we can get a good sense of how […]

The Caspian Sea–Hindu Kush Index (CasHKI): A Climatic Index That Affects Dust Activity Over Southwest Asia

A comprehensive investigation of the dust-storm characteristics over the Sistan Basin in eastern Iran revealed that the dust-storm days during the summer season were associated with an atmospheric teleconnection pattern consisted of high-pressure values over the Caspian Sea and lower than normal pressure over the Hindu Kush – Pamir Mountains. During the summer season, the […]

How Sound Spreads Is Reflected In Large Open Spaces

In a room, we typically hear that a voice is louder than in an open area. This is the effect of reflected sounds which comes out of the mouth and is reflected on the walls and finally comes back into the receiver’s ear. Those sound reflections enhance the voice of people so that louder sound […]

B. infantis EVC001 Reduces The Genes In Bacteria That Cause Antibiotic Resistance By 90 Percent

The World Health Organization (WHO) calls antibiotic resistance “one of the biggest threats to global health, food security, and development today,” and the Centers for Disease Control and Prevention (CDC) calls it “one of the biggest public health challenges of our time.” It’s estimated that it costs the US healthcare system alone an estimated $5 […]

NDH-2: An Alternative Respiratory Enzyme

Living cells rely on the transmembrane difference of electrochemical potential to store the energy needed for their biochemical reactions. Membrane potential is generated by respiratory chain proteins, which use the energy released during the oxidation of energy-rich substrates to translocate ions across the membrane against their electrochemical gradient. The transmembrane difference in electrochemical potential is […]

How To Write A Cursive Capital J

Are you interested in learning how to write a cursive capital J? Read on to find out not only how to write a cursive capital J, but learn some interesting facts about cursive. You can write a cursive capital J by carrying out the following steps. Pick a point to start drawing the J at […]

Bread Mold: How To Identify Types Of Mold

” Bread mold is a type of fungus and is therefore eukaryotic (contains membrane-bound organelles) and multi-cellular. Mold cells are present in a long filamentous structure called a hypha. The mold cells are connected via pores in the septa between cells and are surrounded by a tube-shaped cell wall. A collection of hyphae is called […]