Water Deer Interaction with Humans - Response to Disturbance

Water deer are behaviourally similar to hares in that, when faced with danger, they will duck low in cover and remain quiet and motionless, before erupting away at the last minute. John Heathcote, who manages the Claxton Manor estate in Norfolk, told me that water deer are the only species he knows that can be disturbed standing on the edge of a copse and flee away from the trees rather than into them, something I have also observed. On two occasions, both with does, I observed a behaviour I had previously only seen in roe deer (Capreolus capreolus): "pseudo-grazing", in which a deer that has spotted something suspicious lowers its head as if feeding -- perhaps convincing a predator it hasn't been noticed and prompting it to move -- while keeping its eyes fixed on the threat. In common with other deer, although much less conspicuous owing to its size and lack of rump markings, a fleeing water deer will sometimes hold its tail upright (tail flagging). I have observed this in adults and fawns as young as four or five weeks old, sometimes accompanied by rear-kicking with the hind feet, an consider that it's likely an anti-predator behaviour.

Humans appear to be perceived as a threat in most instances, and this can make water deer very difficult to approach. In their 1981 Nature Conservancy Council report, Arnold Cooke and Lynne Farrell described how, during the autumn of 1976, they attempted to assess how close the deer at Woodwalton Fen in Cambridgeshire would tolerate being approached. Across 22 encounters, the mean flight initiation distance (FID) was 36 m (118 ft.) -- i.e., the observer could get to within 36 m before the deer ran -- although the range was from 8 to 79 m (26-260 ft.) depending on age and the habitat in which they were approached. Younger deer tended to be less flighty than adults, and those within the reserve could often be approached more closely than those in the surrounding arable fields. Reflecting on this in June 2025, Arnold told me: "I remember it was quite difficult to get a deer in a situation where it could be recorded properly, i.e. in the open where I could approach it in full view without it dashing off immediately or even before I'd seen it."

The overall timid nature of water deer is evident even in captivity, unless they have been hand-reared, which can make them strongly imprinted on their human "parent". At Whipsnade, fawns hand-reared by keepers could be called over like dogs; I once watched a four-week-old fawn scream and hurl itself at a fence as its keeper walked away. At Whipsnade Zoo, where the deer see people throughout the year, my experience has been that they remain relatively skittish and can be difficult to observe at close quarters without the cover of a vehicle, although they're still easier to approach than wild individuals. Richard Champion quantified this in his 1996 B.Sc. thesis at Whipsnade, finding that water deer in public areas of the zoo had adapted to human disturbance by significantly reducing their average FID. Deer in public areas could be approached to within 17.5 m (56 ft.), on average, while those in private areas fled when an observer came to within 35 m (115 ft.). There was also considerable individual variation in the distance run when disturbed: deer in public areas fled between 2 and 110 m (6.5–361 ft.), while those in private areas moved between 4 and 120 m (13–400 ft.). Perhaps more interestingly, Champion observed that FIDs were longer -- meaning the deer were more skittish -- during the morning than the evening in public areas, while those in private areas were much more consistent throughout the day. Females were generally more sensitive to disturbance than males.

Champion's findings were broadly replicated by Xin-Xin Tian in his M.Sc. studies at East China Normal University, during which he compared the response to human approach among free-ranging captive deer kept at the Songjiang Punan forest park to that of wild deer living on the Yancheng Nature Reserve in Jiangsu Province. Alert behaviour consisted of five elements: listening, scanning, staring (which may or may not be accompanied by neck stretching), walking or running away, and barking. Notably, aggressive responses to human approach were never observed. (This is consistent with my experience, and I know of no examples of water deer behaving aggressively towards people.) Songjiang deer stopped what they were doing and stared at an intruder at 23 m (75 ft.), on average, and ran away when the person got to within about 19 m (62 ft.). The longest recorded FID at Songjiang was 150 m (492 ft.), while that at Yancheng was nearly 180 m (590 ft.), suggesting wild deer were significantly more skittish than free-ranging captive animals. Tian concluded that, as a direct approach did not alter the FID for semi-captive water deer, humans were perceived as a non-lethal predator. He also noted that an anti-predator response develops early (at around a week old) and that the deer appeared to risk-assess prior to fleeing, running away sooner in open habitat than woodland, with their FID varying according to previous experience with humans, the size of the area, population density, and the presence of fences.

My experience has been that there is considerable individual variation in disturbance response, both in adults and fawns. Unlike in herding species such as red (Cervus elaphus) or fallow (Dama dama) deer, where alarm responses alert the whole herd, it's quite common for one water deer to bark and flee (or stand and bark) while other individuals feed nearby, apparently undisturbed (see: Vocalisation/Calls). Equally, on two occasions I have witnessed does respond to the barking of a Reeves' muntjac (Muntiacus reevesi); once by simply looking in that direction, and once by running at speed towards the source of the disturbance – I believe this was a dam with a fawn in the area. On several occasions I have seen a doe with a fawn at foot bark and flee in one direction while her fawn fled in another before lying down a few metres away. In one instance, I watched a doe with a four- or five-week-old fawn feeding shortly after sunrise for about ten minutes before she spotted me. After nearly a minute of staring and neck stretching she barked and fled; her call caused the fawn to immediately drop into the grass and sit tight. More recently, biologists at Shanghai's East China Normal University collected data on how anti-predator behaviours develop in fawns. Between June and October 2023, researchers led by Xiaorong and Min Chen observed how 30 ear-tagged fawns kept in a 0.5 ha (1.2 acre) enclosure in Huaxia Park -- composed of broadleaf woodland and grassland -- responded to human approach, with the aim of understanding at what age vigilance behaviours developed. In their 2025 paper in Behavioural Processes, the researchers documented five main vigilance behaviours, categorising them as:

Flight - The fawn turns and flees from the observer's approach.
Stare - The fawn stands or lies and looks in the direction of the observer for a varying period.
Stretch neck - The fawn bobs its head up and down, sometimes also side to side, while staring fixedly in the direction of the disturbance.
Vigilant paw - The fawn lifts the foreleg and suddenly strikes the ground with the hoof one or more times.
Roar - The fawn emits a short, repeated "er" sound, presumably a distress vocalisation to alert its mother.

The most frequently observed behaviours were staring and fleeing, recorded 942 and 915 times respectively, and together these were clearly the dominant vigilance strategy. FID ranged from 0 to 10.6 m (35 ft.), with a mean of 2.4 m (8 ft.). Broadly speaking, older fawns (9-18 weeks) had lower FIDs than younger ones (1-8 weeks), while those less than a week old did not flee, as is expected in newborns. The distress vocalisation was heard only twice, both times in fawns under three weeks old, while neck stretching and vigilant pawing didn't appear until around ten weeks. As the fawns aged, their vigilance behaviour became more complex: age accounted for just under 41% of the variation in vigilance behaviour, while the presence of another deer, the direction of the observer's approach, and the availability of shelter each accounted for no more than 5%. Older fawns tended to display longer, more varied sequences of responses during any given approach, presumably reflecting increasing experience of risk assessment. Interestingly, unlike some studies that have found animals to be sensitive to human eye contact, the researchers found no significant influence of observer gaze on the fawns' behaviour, noting that water deer appeared to perceive threat sources more holistically, sometimes fleeing without ever seeing the observer's eyes.

I know of no equivalent data on water deer behavioural response to humans in Korea, but they do seem more likely to live in and around human habitation there, and I have seen footage of deer in towns and cities in the south. The species' relative boldness in South Korea is anecdotally illustrated by an incident in April 2026, when Story Convenience Store in Yangju-si, Gyeonggi Province, posted CCTV footage of a man being knocked to the ground when a water deer ran into him in a busy metropolitan complex. The deer had been caught on camera elsewhere in the city running around for about ten minutes before the collision. The city has expanded in recent years into the borders of a forest park, and commentary in the video from Park Byeong-kown, Director of the Korean Urban Ecology Institute, suggests that a river channel and its peripheral vegetation running through the city to the forest provides excellent habitat for water deer, which is why animals end up in the city itself. In a 2011 paper in Landscape Ecology and Engineering, however, Baek-Jun Kim and colleagues reported that water deer were significantly less common in metropolitan areas with populations in excess of one million (present in 67% of samples) than in rural areas (83% of samples). Water deer nonetheless appear to have adapted to human presence to a greater degree in South Korea than elsewhere.