Preliminary Observations Using Thermal Aerial Surveillance

Derrick Dixon
Whitetail Research
Southern Missouri, USA

Abstract

Direct observation of interactions between coyotes (Canis latrans) and white-tailed deer (Odocoileus virginianus) is difficult, particularly at night when many interactions occur outside the range of conventional observation. From July 10 through August 23, 2026, I used a thermal-equipped unmanned aerial vehicle to observe coyotes and white-tailed deer across approximately 2,000 acres in southern Missouri. During the 45-day observation period, I recorded 623 fawn detections. Because individual fawns were not physically marked or otherwise consistently identifiable, these detections do not represent 623 unique animals.

A mature buck was located within 50 yards of a fawn during 62 of the 623 detections (9.95%). Of those mature buck–fawn associations, approximately 83% occurred at distances of less than 10 yards. In many of the closest associations, the mature buck and fawn were bedded in immediate proximity. Approximately seven visually distinguishable mature bucks appeared to account for the observed associations.

Coyotes directly documented three fawn predation events during the observation period. Two involved coyotes classified as transients and one involved a resident breeding group associated with pups. A mature buck accompanied none of the three fawns at the time of the documented predation event.

Repeated observations of the resident coyote group also showed coyotes moving near aggregations of deer without consistently initiating direct pursuit. Deer frequently changed position as coyotes moved through occupied areas, resulting in gradual displacement of deer across portions of the fields. Adult females occasionally approached or attempted to strike coyotes at close range, whereas comparable close approaches toward mature males were not documented during the observed encounters.

These observations generated four working hypotheses: (1) coyotes may use persistent, low-intensity movement around groups of deer in ways that increase opportunities to expose or isolate fawns; (2) deer may alter their spatial organization in response to predictable coyote movement; (3) the presence of mature male deer may influence coyote movement or access to areas occupied by fawns; and (4) mature male white-tailed deer may occur in very close proximity to fawns more frequently than would be expected from independent use of the same habitat. These hypotheses remain preliminary and require prospective, quantitative testing.

Introduction

Coyotes are an established source of mortality for white-tailed deer fawns, although their contribution to fawn mortality varies substantially among landscapes and predator communities. Research has demonstrated that habitat composition and landscape structure can influence the probability of coyote predation, while work in multi-predator systems has further illustrated that the importance of coyotes depends partly on predator abundance and ecological context (Gulsby et al., 2017; Kautz et al., 2019). (USFS Research & Development)

Successful predation, however, represents only one possible outcome of an interaction between predator and prey. Predator presence can influence movement, habitat selection, vigilance, group structure, and other behaviors even when no kill occurs. These nonconsumptive effects are well documented among ungulates and constitute an important component of predator–prey ecology (Chitwood et al., 2022). (PubMed Central (PMC))

White-tailed deer have demonstrated behavioral responses specifically associated with coyote presence and abundance. Gulsby et al. (2018) found that coyote abundance was associated with changes in foraging and vigilance among several deer sex and age classes. Clipp et al. (2024) subsequently found greater deer detection intensity at sites occupied by coyotes and identified increased movement as one possible explanation. Adult female white-tailed deer have also been documented increasing vigilance when fawns are present, indicating that reproductive and social conditions can influence antipredator behavior (Lashley et al., 2014). (The Ecological Society of America)

Most studies of these relationships necessarily rely on telemetry, mortality investigations, camera traps, predator abundance estimates, or other indirect measurements. These approaches provide important information but offer limited opportunity to observe predator and prey behavior during individual nocturnal encounters continuously.

Thermal-equipped unmanned aerial vehicles provide another means of observing free-ranging mammals after dark. Thermal UAV systems have been successfully evaluated for detecting and monitoring terrestrial mammals, although flight altitude, aircraft disturbance, species identification, and other methodological considerations remain important (Larsen et al., 2023). (DOI)

The present project began as an effort to document nighttime use of the landscape by white-tailed deer fawns during summer 2026. Repeated observations of coyotes occupying the same fields subsequently provided an opportunity to observe interactions among coyotes, fawns, adult females, and mature males.

As the observation period progressed, several spatial and behavioral patterns occurred repeatedly enough to justify further investigation.

The objective of this work was exploratory. I did not attempt to estimate population-level fawn survival or establish causal relationships between the animals being observed. Instead, I sought to document recurring interactions and spatial associations and use those observations to develop measurable hypotheses for future research.

Study Area and Methods

Study Area and Observation Period

Observations were conducted from July 10 through August 23, 2026, encompassing 45 days across approximately 2,000 acres in southern Missouri.

The study landscape included large open fields, wooded habitat, edge cover, and associated vegetation. A thermal-equipped unmanned aerial vehicle was used primarily after sunset to locate and observe coyotes and white-tailed deer.

When practical, fawns were initially identified near dusk. Once identified, their locations could be monitored using thermal imagery after dark, reducing the need for repeated low-altitude approaches or visible illumination.

All field observations, UAV operation, animal identification, and subsequent classification of observations were conducted by a single observer.

Fawn Detections

A total of 623 fawn detections were recorded during the 45-day observation period, averaging approximately 13.84 detections per day.

The term detection is used intentionally.

Individual fawns were not captured, physically marked, fitted with telemetry devices, or otherwise consistently identifiable throughout the study. Some fawns were therefore observed repeatedly, and the number of unique individuals represented within the 623 detections is unknown.

Accordingly, the 623 detections should not be treated as 623 independent animals or 623 independent biological samples.

Mature Buck–Fawn Association

For this study, a fawn was classified as being associated with a mature buck when a mature male white-tailed deer was located less than 50 yards from the fawn during an observation.

Distance between animals was estimated using the UAV's rangefinding and mapping capability. A coordinate pin was established at the location of the fawn and another at the location of the mature buck. We then measured the linear distance between the two mapped positions.

The <50-yard threshold was applied consistently when classifying mature buck–fawn associations.

The position of the mature buck relative to coyotes was not used to determine whether an association occurred. A buck was classified as associated solely according to its measured distance from the fawn. This distinction was maintained so that interpretation of a buck's apparent behavior would not influence classification.

A mature buck was located within 50 yards during 62 of 623 fawn detections (9.95%).

No mature buck was documented within the 50-yard threshold during the remaining 561 detections (90.05%).

Distance within the association category was strongly concentrated near the fawn. Approximately 83% of mature buck–fawn associations occurred at distances of less than 10 yards. The remaining observations occurred between approximately 10 and 50 yards.

Approximately seven visually distinguishable mature bucks appeared to account for these associations. Because individual bucks were not physically marked, this number represents an observational estimate rather than a confirmed number of unique animals.

Coyote Observations

Fewer than 20 coyotes were observed during the broader project.

As the study progressed, observations became concentrated on an apparently resident family group consisting of approximately eight coyotes. This group repeatedly used several large fields that deer also heavily occupied after dark.

When possible, I recorded coyote position relative to deer, general movement and foraging behavior, deer responses to approaching coyotes, and direct interactions between coyotes and adult deer.

Coyotes were not captured, marked, or equipped with telemetry devices. Classification of animals as resident or transient was therefore based on repeated spatial and behavioral observations rather than confirmed home-range data.

Results

Spatial Association Between Fawns and Mature Bucks

Mature bucks occurred within 50 yards of a fawn during 62 of 623 detections (9.95%). The remaining 561 detections (90.05%) occurred without a mature buck documented within the predefined association distance.

The majority of associated bucks were substantially closer than the 50-yard classification threshold.

Approximately 83% of the mature buck–fawn associations occurred at distances of less than 10 yards.

In many of these close observations, the mature buck and fawn were bedded in immediate proximity to one another.

The remaining approximately 17% of associations occurred at distances between 10 and 50 yards.

Some of the more distant observations displayed a different spatial arrangement. In several instances involving a mature buck approximately 30–50 yards from a fawn, the buck occupied a position between the fawn and areas of ongoing coyote activity.

This positioning was recorded descriptively but played no role in determining whether an observation qualified as a mature buck–fawn association.

Approximately seven visually distinguishable mature bucks were represented among the observations, suggesting that the pattern was not attributable exclusively to repeated observations of a single mature male. However, individual identity could not be independently confirmed.

Documented Predation Events

Three fawn predation events by coyotes were directly documented during the observation period.

Two involved coyotes classified as transients.

One involved the resident breeding group associated with pups.

A mature buck accompanied none of the three fawns at the time of the documented predation event.

Because the true number of individual fawns represented within the 623 detections is unknown, these events cannot be used to calculate a mortality rate or survival estimate. They are reported only as directly observed predation events.

Coyote Behavior Near Deer

The resident coyote group repeatedly occupied fields simultaneously used by relatively large numbers of deer.

Coyotes frequently moved through these areas at comparatively low speeds, with movement interrupted by short pounces, pauses, sitting, resting, and behavior consistent with small-prey foraging.

Much of this activity occurred while deer remained nearby.

Coyotes were repeatedly observed within approximately 10–50 yards of deer groups without immediately initiating direct pursuit.

Deer did not consistently respond by leaving the field rapidly. Instead, individuals and groups commonly moved short distances as coyotes changed position.

Over prolonged observation, repeated coyote movement could gradually displace deer across portions of a field.

No observed sequence conclusively demonstrated that this displacement culminated in the successful separation and capture of a fawn.

Interactions With Adult Deer

Coyotes occasionally approached adult female deer at short distances.

During some encounters, a doe responded by moving toward the coyote and attempting to strike or stomp it. Coyotes generally increased their distance following these confrontations and subsequently returned to slower movement or apparent foraging.

Comparable close approaches toward mature bucks were not documented during the encounters observed in this study.

This pattern remains descriptive. Encounter opportunity was not standardized between adult females and mature males, and the current dataset cannot demonstrate that coyotes systematically avoid mature bucks.

Discussion

Gradual Displacement of Deer by Coyotes

One of the more consistent behavioral observations involved coyotes remaining relatively close to groups of deer without immediately initiating pursuit.

The simplest explanation is shared habitat use. Coyotes may have been primarily hunting rodents, insects, or other prey while deer independently occupied the same productive fields. Under this explanation, deer displacement would reflect a response to incidental predator proximity rather than a deliberate coyote hunting strategy.

Another possibility is that persistent low-intensity movement through areas occupied by deer provides coyotes with an indirect advantage.

Repeated deer movement could change spacing within a group. If adults and fawns respond differently, or if repeated displacement causes a fawn to become temporarily separated from other deer, coyotes could gain an opportunity that did not exist before the interaction began.

The current observations cannot distinguish between these possibilities.

The first working hypothesis is therefore:

H1: Persistent, low-intensity movement by coyotes around deer aggregations increases the probability that a fawn becomes spatially separated or exposed.

Testing this hypothesis would require measuring deer spacing before, during, and after coyote approaches and comparing those changes with periods when coyotes are absent.

Deer Response to Coyote Movement

The deer response was equally notable.

Despite regular coyote activity, deer repeatedly continued using the same large fields. Even when coyotes approached relatively closely, deer frequently repositioned rather than abandoning the area completely.

This general pattern is consistent with previous research demonstrating that ungulates may respond to predation risk through changes in movement, habitat use, grouping, or vigilance rather than complete spatial avoidance (Chitwood et al., 2022). White-tailed deer have also demonstrated behavioral responses to coyote abundance and increased detection intensity at locations occupied by coyotes, potentially reflecting increased movement in response to predator presence (Gulsby et al., 2018; Clipp et al., 2024). (Wiley Online Library)

Large open fields may also affect the interaction itself. Visibility, group size, vegetation structure, escape opportunities, and predator detection could all influence the relative risk deer experience. Previous work has demonstrated that landscape heterogeneity can affect coyote predation risk to white-tailed deer fawns, although the mechanisms and magnitude of these effects are landscape dependent (Gulsby et al., 2017). (The Wildlife Society)

This produced a second working hypothesis:

H2: White-tailed deer modify their spatial organization and movement in response to predictable coyote behavior in ways that reduce predation risk.

The present observations do not establish that deer responses are coordinated or that the observed movements improve fawn survival.

Mature Bucks and Fawns

The most unexpected pattern observed during the project involved mature male deer.

Mature bucks occurred within 50 yards of fawns during 62 detections. More importantly, these observations were not distributed evenly throughout the 50-yard classification radius.

Approximately 83% of all mature buck–fawn associations occurred at distances of less than 10 yards.

In many of these observations, the buck and fawn were bedded in immediate proximity.

This distinction is important. A mature buck located 45 yards from a fawn provides relatively limited information by itself. Repeated observations of bucks within 10 yards, however, represent a more concentrated spatial relationship and provide a specific pattern that can be tested.

The current study cannot determine whether the relationship occurred more frequently than would be expected by chance.

Fawns and mature bucks may independently select the same locations because of forage availability, vegetation structure, thermal conditions, bedding characteristics, visibility, proximity to escape cover, or other unmeasured habitat variables.

Consequently, these observations should not presently be interpreted as evidence of a social relationship between mature bucks and fawns.

They do, however, justify testing one.

This produces the fourth working hypothesis:

H4: Mature male white-tailed deer occur within very close proximity (<10 yards) of fawns more frequently than would be expected from independent spatial use of the same habitat.

A future study could test this by comparing observed buck–fawn distances with distances expected from simultaneously recorded locations of mature bucks and fawns across the landscape.

Potential Influence of Mature Bucks on Coyote Behavior

A separate question concerns what happens when a mature buck is present near a fawn.

Some of the less-proximate mature buck observations were especially noteworthy because the buck occupied a location between the fawn and an area of active coyote movement.

At the same time, close coyote approaches were documented toward adult females, including interactions in which does attempted to strike coyotes. Comparable approaches toward mature bucks were not observed during this project.

Neither observation demonstrates that mature bucks intentionally protect fawns.

Intentional protection is also not required for a biological effect to exist.

A mature male white-tailed deer represents a substantially different physical interaction for a coyote than a fawn. If coyotes perceive mature males differently and modify approach distance or travel direction because of their presence, a mature buck could influence predation opportunity without deliberately acting as a guardian.

The appropriate question is therefore narrower than whether mature bucks "protect" fawns.

The third working hypothesis is:

H3: The presence and proximity of a mature male white-tailed deer alters coyote approach behavior or spatial access to areas occupied by fawns.

This hypothesis can be evaluated quantitatively without assigning intention to either the buck or the coyote.

The observation that none of the three documented fawn predation events occurred while a mature buck was associated with the fawn provides an additional reason to examine the question, but three events are insufficient for inference.

Limitations

Several limitations substantially restrict the conclusions that can be drawn from this work.

First, the project was conducted on a single approximately 2,000-acre landscape in southern Missouri. Observations from this site should not be assumed to represent coyote or white-tailed deer behavior across other landscapes.

Second, neither fawns nor adult deer were physically marked. Repeated observations therefore included some of the same individuals, and the degree of pseudoreplication within the dataset is unknown.

Consequently, the 623 fawn detections cannot be treated as 623 independent samples.

Likewise, the estimate of approximately seven mature bucks was based on visual differentiation rather than independent identification.

Third, only three fawn predation events were directly documented. That sample is inadequate for evaluating whether mature-buck presence is associated with differences in fawn survival.

Fourth, the project was initially observational rather than a controlled study designed around the hypotheses that subsequently emerged. Observation effort was therefore not balanced across combinations such as fawns with mature bucks, fawns without mature bucks, coyotes present, coyotes absent, or differing coyote densities.

The <50-yard mature buck association threshold provides a consistent classification rule, but the study did not establish a comparable random expectation for buck–fawn proximity. Consequently, the observed 9.95% association rate and the concentration of 83% of associations within 10 yards cannot yet be interpreted as evidence that bucks and fawns occur together more often than expected by chance.

Future Research

The primary purpose of this preliminary project was to generate questions that could be converted into standardized measurements. A subsequent study should therefore be designed prospectively around the hypotheses identified here.

For each fawn encounter, future observations should record:

• fawn-to-nearest-coyote distance;

• fawn-to-nearest-adult-deer distance;

• fawn-to-nearest-mature-buck distance;

• adult sex and approximate age class;

• presence or absence of a mature buck within predetermined distance categories;

• number of coyotes present;

• deer group size and composition;

• coyote behavioral state;

• deer behavioral state;

• minimum coyote approach distance;

• direction and magnitude of deer displacement;

• habitat type;

• vegetation height;

• wind direction and speed;

• time of night; and

• encounter outcome.

Mature buck–fawn distances should be retained as continuous measurements rather than only being reduced to a binary <50-yard classification.

Distance categories such as <10 yards, 10–20 yards, 20–30 yards, 30–40 yards, and 40–50 yards could also be reported descriptively while retaining the raw distance for statistical analysis.

This would allow the apparent concentration of buck–fawn associations at very short distances to be tested directly.

Individual identification would substantially improve inference. GPS collars, physical marking, or reliable individual identification would allow repeated measurements from the same deer or coyotes to be incorporated appropriately into statistical models rather than treated as independent observations.

The mature-buck/coyote hypothesis could then be evaluated by comparing minimum coyote approach distances and encounter outcomes when mature bucks are present versus absent.

The buck–fawn association hypothesis could be evaluated by comparing observed distances with a null distribution representing the independent spatial use of the landscape by mature bucks and fawns.

Likewise, the coyote-displacement hypothesis could be tested by measuring whether fawn-to-adult distance, group cohesion, or fawn isolation changes after coyotes move through a deer aggregation.

With adequate replication and individual identification, mixed-effects models could account for repeated observations of individual animals, nights, locations, and coyote groups while evaluating the effects of mature-buck presence, coyote number, habitat, and other environmental variables.

Conclusion

During a 45-day period of thermal aerial observation in southern Missouri, I recorded 623 fawn detections.

Mature bucks were located within 50 yards of fawns during 62 detections (9.95%), and approximately 83% of those associations occurred at distances of less than 10 yards. In many of the closest observations, mature bucks and fawns were bedded in immediate proximity.

Approximately seven visually distinguishable mature bucks appeared to exhibit the association.

Three coyote predation events on fawns were directly documented. Two involved transient coyotes and one involved the resident breeding group associated with pups. A mature buck accompanied none of the three fawns at the time of predation.

Coyotes were also repeatedly observed moving near aggregations of deer without immediately initiating pursuit. Deer commonly repositioned as coyotes moved through occupied fields, and adult females occasionally confronted approaching coyotes. Similar close approaches toward mature bucks were not documented during the observation period.

These observations do not demonstrate that coyotes deliberately herd deer to expose fawns. They do not demonstrate that deer intentionally manipulate coyote movement. They do not demonstrate that mature bucks intentionally associate with or protect fawns.

They do identify several repeatable patterns that can now be tested.

The primary contribution of this preliminary work is therefore not the demonstration of a previously unknown behavior, but the development of four specific and falsifiable hypotheses arising from repeated direct observations:

H1: Persistent, low-intensity movement by coyotes around deer aggregations increases the probability that a fawn becomes spatially separated or exposed.

H2: White-tailed deer modify their spatial organization and movement in response to predictable coyote behavior in ways that reduce predation risk.

H3: The presence and proximity of a mature male white-tailed deer alters coyote approach behavior or spatial access to areas occupied by fawns.

H4: Mature male white-tailed deer occur within very close proximity (<10 yards) of fawns more frequently than expected based on independent spatial use of the same habitat.

The observations reported here provide the basis for moving those questions from exploratory field observation to controlled quantitative investigation.

References

Chitwood, M. C., Baruzzi, C., & Lashley, M. A. (2022). “Ecology of fear” in ungulates: Opportunities for improving conservation. Ecology and Evolution, 12(3), e8657. https://doi.org/10.1002/ece3.8657. (PubMed Central (PMC))

Clipp, H. L., Pesi, S. M., Miller, M. L., Gigliotti, L. C., Skelly, B. P., & Rota, C. T. (2024). White-tailed deer detection rates increase when coyotes are present. Ecology and Evolution, 14(3), e11149. https://doi.org/10.1002/ece3.11149. (USGS Publications)

Gulsby, W. D., Cherry, M. J., Johnson, J. T., Conner, L. M., & Miller, K. V. (2018). Behavioral response of white-tailed deer to coyote predation risk. Ecosphere, 9(3), e02141. https://doi.org/10.1002/ecs2.2141. (The Ecological Society of America)

Gulsby, W. D., Kilgo, J. C., Vukovich, M., & Martin, J. A. (2017). Landscape heterogeneity reduces coyote predation on white-tailed deer fawns. The Journal of Wildlife Management, 81(4), 601–609. https://doi.org/10.1002/jwmg.21240. (USFS Research & Development)

Kautz, T. M., Belant, J. L., Beyer, D. E., Jr., S. B. K., Petroelje, T. R., & Sollmann, R. (2019). Predator densities and white-tailed deer fawn survival. The Journal of Wildlife Management, 83(5), 1261–1270. https://doi.org/10.1002/jwmg.21681. (The Wildlife Society)

Lashley, M. A., Chitwood, M. C., Biggerstaff, M. T., Morina, D. L., Moorman, C. E., & DePerno, C. S. (2014). White-tailed deer vigilance: The influence of social and environmental factors. PLOS ONE, 9(3), e90652. https://doi.org/10.1371/journal.pone.0090652. (PLOS)