A researcher follows a giraffe to learn how it spends its day. The giraffe walks away. Its movement is faithfully recorded, but part of that movement may have been caused by the person doing the recording. The resulting data describe an animal responding to a researcher as well as an animal going about its ordinary life.
Ciska Scheijen and colleagues made that problem the subject of a study at a South African reserve. Combining GPS records, direct observations and hormone measurements from dung, they found that seven giraffes travelled farther when people were present. Repeated exposure reduced the movement difference, but the animals did not show a matching, clear decline in time spent watching their observers. Becoming accustomed to people was a process with several components, rather than a single switch. The focal study appeared in Animals in 2021.
Tracking the hours nobody was watching
The important comparison required knowing where the giraffes went when observers were absent. GPS devices made that possible. Eighteen giraffes were initially fitted with devices, but technical and fitting problems left sufficiently prolonged records for four adult males and three adult females. Their devices recorded locations hourly for between five and twelve months, producing more than 33,000 logged positions.
Those positions allowed the team to estimate movement between consecutive hourly fixes. This is displacement between recorded points, so it cannot capture every bend or detour along the route. Its advantage was consistency: the same measure was available during hours with and without people nearby.
On observation days, researchers located a giraffe by vehicle and then followed it on foot. They recorded when it watched them and which other giraffes were present. The study began in a population that was unaccustomed to close human observation. Hunting, including giraffe hunting on foot, had continued at the reserve until shortly before the project. That history makes this a particular encounter between people and animals, rather than a universal test of how every giraffe responds to every visitor.
The movement analysis accounted for individual differences, sex, time of day and wet or dry season. Those adjustments mattered because a comparison of all observed and unobserved hours would otherwise mix daytime human presence with the animals’ ordinary daily activity cycle.
A substantial movement difference, with signs of change
Observer presence was associated with a 66.6% increase in hourly distance travelled; the reported confidence interval ranged from about 50% to 85%. The result is a relative difference in the statistical model, rather than an extra fixed number of metres added to every hour.
As each giraffe accumulated more experience of observers, the additional movement declined. The study’s graph places these trends alongside the estimated movement without observers: the lines move closer together as observation hours accumulate. The pattern is consistent with habituation, meaning a reduced response after repeated exposure.
It does not establish a dependable timetable for producing a habituated giraffe. The paper extrapolated several hundred hours of exposure before the estimated human effect would disappear. Those projections extend beyond the average experience accumulated during the study and depend on this site, approach method and population. They are not a recommended number of hours to spend following a wild animal.
Other giraffes also mattered. With people present, a focal animal travelled farther when more group members were classified as unhabituated. Each additional such companion was associated with about 13% more hourly movement. An animal may therefore continue moving with its companions even as its own response to people changes. Treating each tracked giraffe as socially isolated would miss that possibility.
Looking at a person is not the same measurement as walking away
The vigilance observations complicated the picture. More observers, and more unhabituated giraffes in the group, were associated with more time spent watching the people. Larger groups were associated with less observer-directed watching by the focal animal.
But cumulative exposure did not significantly reduce this watching measure. That need not mean the giraffes experienced an identical emotional state throughout the year. Watching could accompany alarm, curiosity or other kinds of attention, and the researchers did not separate those possibilities experimentally. The secure result is that visible attention persisted while the movement response weakened.
The hormone evidence added a third timescale. Researchers collected 159 dung samples and measured faecal glucocorticoid metabolites, substances used to assess activity of the physiological stress-response system. Because hormones take time to pass into dung, samples were classified with reference to earlier observer presence, not merely whether a person stood nearby at defecation.
The simple observer-presence effect on metabolite concentration was uncertain by the study’s conventional significance threshold. Concentrations did decline with increasing accumulated exposure. This supports a changing physiological response over time, but it does not justify saying that every observed encounter caused a proven hormone spike or that the study measured lasting harm.
Habituation depends on what the animal encounters
A later sound-playback study compared giraffes at three reserves with different histories of human activity. In the reserve with least exposure, drone and vehicle sounds produced stronger immediate reactions than dove sounds in matched comparisons. That experiment isolated brief sound categories; Scheijen’s study followed the much broader experience of having people nearby for hours. Their results concern related questions, but the exposures and measurements cannot be substituted for one another.
Together, they suggest why a familiar vehicle and a person approaching on foot may belong to different categories in a giraffe’s experience. Experience with one situation should not automatically be treated as evidence of tolerance in another.
For research, the central implication is practical: observations need an account of the observer. GPS records supplied a comparison that direct following alone could not provide. Group composition helped explain why an animal’s movement might change even when its personal history seemed familiar. Multiple measures revealed that reduced movement did not mean all responses had disappeared.
The study did not measure reproductive losses, disease or survival caused by observation. Its contribution comes earlier in the chain of inference: it showed that the act of collecting behavioural data can change the behaviour being collected, and that recognising this effect requires more than deciding that a giraffe now looks calm.
Sources
- Ciska P. J. Scheijen, Sean van der Merwe, Andre Ganswindt and Francois Deacon (2021). “Anthropogenic Influences on Distance Traveled and Vigilance Behavior and Stress-Related Endocrine Correlates in Free-Roaming Giraffes.” Animals 11, 1239. Full paper.
- Kaitlyn Taylor, Sean van der Merwe, Francois Deacon, Andri Grobbelaar and Anton Baotic (2025). “Ecological Exposure History Shapes Giraffe Vigilance Responses to Anthropogenic Noise: A Multisite Playback Experiment.” Ecology and Evolution 15, e72763. Related playback study.
