Water Deer Behaviour - Social Behaviour

---

Table of contents

• Group size and aggregations
• Solitary animals and small groups
• Pairs and bucks accompanying does
• Contact, grooming and sniffing
• Does with fawns, siblings and the effect of age
• Interaction rates
• Captive vs. wild data
• The bottom line
• References

---

Group size and aggregations

It's not uncommon to see substantial numbers of water deer in close proximity and, in his 1940 The Mammals of China and Mongolia, Glover Allen recounted Robert Swinhoe's early observation that "a herd of twenty was put up in the great marsh beyond the Hangchow Bay sea-wall, near Fenghsien", about 32 km (20 miles) south of Shanghai. Here in the UK, larger groupings tend to be reported in open habitats, most often in winter (particularly during the rut) and in early spring, as the days warm and plant growth resumes.

Arnold Cooke observed one, in his words, "remarkable aggregation" of 16 deer grazing on a recently mown mixed fen area of about a hectare (2.5 acres) inside the nature reserve at Woodwalton Fen at 16:30 on 14th February 2011, and during his Ph.D. research at Whipsnade Stefan Stadler found the largest group there contained 15 animals.

During an early afternoon in December 2022, I counted 35 in a single field on an estate in Bedfordshire, and in April 2025 Paul Childerley, who manages this estate, mentioned that there may be 60 to 70 deer in a single field. The largest aggregation I'm aware of at this location -- and the largest I've found reported anywhere, although the literature from China and Korea is sparse -- occurred in October 2019, when Stephen Plummer counted 112 deer in a single field. It should be noted, though, that this population is managed for stalking and the deer live at what might be considered an artificially high density (see: Population & Abundance – England).

We don't know how these large aggregations form – whether individuals or pairs move to the same favoured feeding and resting areas, for example, or whether larger groups coalesce. In my experience, though, a water deer doesn't seem particularly inclined to approach another simply because it has seen it, unlike in herding species (although see Contact, grooming and sniffing below), which hints at independent arrival.

However they form, such aggregations appear mostly to be associations of convenience, the animals using good feeding and/or resting areas simultaneously, and water deer show no tendency towards the herding behaviour we see in red (Cervus elaphus) or fallow deer (Dama dama). Neighbouring deer seem to pay no more than cursory attention to an alarm-barking or fleeing individual, for example. Nor does the fate of one deer necessarily trouble its neighbours. I've seen footage of one deer watching from a treeline no more than four metres (13 ft) away as a buck in the open was shot, the "watcher" making no attempt to flee until the stalkers approached to retrieve the carcass. In another video, a buck remained lying in the grass only a couple of metres (about 6.5 ft) from a doe that was shot and, once again, showed no inclination to flee until the stalker approached. The same independence is apparent when an aggregation is disturbed: the deer scatter in all directions rather than fleeing together as a cohesive unit. Such aggregations may also be transient over the course of a day, as Norfolk-based wildlife photographer Gary Rugless has described to me from his patch:

"If you spot a group they stay together for a while, then seem to go in different directions for most of the day then regroup early morning, sometimes afternoon."

Lixing Sun observed that water deer at Poyang Lake National Nature Reserve in Jiangxi, China tended to form larger groups during the breeding season (particularly early on, in November) and were most likely to be solitary during parturition (May–July). In a 2002 paper in Acta Zoologica Sinica, Sun explained:

"The reason why the group size fell into the lowest point in the parturition season was that each female actively sought a quiet location to give birth and care for her fawns. Males that used to be grouped with all-female groups became solitary, too. As fawns grew and could follow their mothers, females with fawns could group with other deer and consequently, the overall group size rose again."

Sun also noted that the deer were more likely to form groups on short grass than in tall vegetation, but the time spent feeding or engaged in vigilance behaviours didn't vary according to whether they were in a group or solitary, suggesting that an anti-predator response was not the reason for grouping. Earlier, in his 1991 thesis, Stadler had proposed a different benefit, at least for young animals: that the seasonal and age-specific associations at Whipsnade may help with social development:

"In the autumn, female fawns associated increasingly with other female fawns and male fawns did increasingly with other male fawns. Thus, associating with like-sex individuals of the same age class may be actively sought by fawns and this may promote sex-specific behavioural development."

Overall, there's little indication of sustained group cohesion, although the pair bonds, small groups travelling together, and friendly encounters described below suggest that it's not entirely absent.

↑ table of contents

Solitary animals and small groups

Despite water deer sometimes forming large aggregations, most sightings are of solitary individuals or small groups of two or three animals, typically all females, a female with this year's fawns, or (in my experience) a buck with one or two does.

Out of 2,195 sightings logged at Woodwalton Fen between June 1976 and May 1979, Arnold Cooke and Lynne Farrell noted that 61% were of solitary deer. At Poyang Lake, Sun recorded that 90% of all groups were composed of only two or three individuals, and during the mating season, when bucks become aggressive towards one another and try to round up females, 70% of single deer for which sex could be attributed were bucks, compared with only 30% of females seen alone.

Since January 2019, I've been camera-trapping on a farm in Buckinghamshire during the summer and winter. Of 1,589 trail camera videos featuring water deer collected up until February 2026, the vast majority (94%) showed a solitary individual, while just over 4% contained two deer in close proximity; only 20 videos showed three or more. In two videos, captured about a month apart in 2025, five deer (all appeared to me to be females) were travelling together; the deer moved into a clearing, following the same general track, and exited the clearing together. The largest grouping I have captured at the site was (at least) seven animals during December 2025; they were together for about a minute, some running around, others sniffing noses, before the group dispersed, though it's unclear from the video whether this was rut-related. (It's important to note that these are percentages of videos rather than of independent encounters, so they are not directly comparable with the sighting-based figures above.)

The largest wild group I've personally observed (i.e., not captured on a remote camera) that appeared cohesive was at Ouse Fen in July 2024 and comprised four individuals: a buck, an adult female with a young fawn, and an adult of undetermined sex. The animals walked together into a patch of long grass and lay down, remaining there for the 15 minutes or so I was able to observe the area.

In captivity, Stadler sometimes encountered groups of seven or more during his Ph.D. studies at Whipsnade but, of nearly 17,000 records of 175 individuals, 66% were of deer either singly or in pairs, and only 5% of all the groupings recorded consisted of five animals or more.

↑ table of contents

Pairs and bucks accompanying does

While the stable group formations observed in many other deer species do not appear to be characteristic of Hydropotes, Cooke and Farrell documented a stable pair bond at Woodwalton, where the same buck and doe occupied a territory throughout the year. In his 1982 account of the species, Bob Lawrence noted that barren does may remain paired with a buck outside the breeding season. Consistent with this, my own observations suggest that deer travelling or feeding in pairs outside the principal fawn-rearing period are most often either two does or a buck-doe pair. I have encountered buck-doe pairs on multiple occasions during the summer in wild populations in Buckinghamshire, Bedfordshire, and Norfolk, as well as among free-range captive individuals at Whipsnade and Woburn.

I have also observed bucks accompanying does during the summer months, a behaviour that, until recently, I assumed indicated that the female was not lactating. In June 2024, however, I watched a buck and doe feeding together at Woburn when the doe returned to attend to twin fawns concealed at the edge of a field. The buck followed her, briefly sniffing and licking one of the fawns that approached it, before sniffing the doe's hindquarters and accompanying her away, leaving the fawns unattended. This was the only occasion on which I've seen a doe return to her concealed fawns accompanied by a buck; my experience has typically been that does chase other deer away from their fawns' lying-up sites. The following month, as noted above, I observed a mature buck accompanying a female with a young fawn at Ouse Fen, and in July 2025 one of my trail cameras in Buckinghamshire recorded a buck and doe apparently travelling together with a young fawn in tow. These observations suggest that bucks may, on occasion, remain in association with does that have dependent young, which contrasts with Sun's finding at Poyang Lake that males became solitary around parturition.

At Branféré Zoological Park in France, Gérard Dubost and colleagues found no lasting bonds, even between mothers and their fawns; no notable interactions occurred once the youngsters were more than about five months old. In their 2011 paper in Acta Theriologica, Dubost and his team wrote:

"There was no durable bond between mothers and own young. Four young born to focal females and still alive in the mating season had less contact with their mothers after the first period of the mating season than did unrelated young."

There were also no special bonds apparent between the focal females and other mature individuals of the population, which tallies with the findings of Stadler at Whipsnade and of Lixing Sun and Nianhua Dai at Poyang Lake. Stadler noted in his thesis that although females sought other females outside the fawning period, the associations were temporary, and no long-lasting associations were observed between does at Whipsnade. Though they may be transient, there are a few reports of what appear to be congenial interactions between adults, most notably does.

↑ table of contents

Contact, grooming and sniffing

In a thesis on muntjac and water deer behaviour completed in 1971, Wolf-Peter Scherpe reported soziale Körperpflege (social grooming) between captive does. Similarly, writing in 1982, François Feer described six instances of léchage mutuel (mutual licking) between females in a 0.13 ha (0.3-acre) enclosure at Wuppertal Zoo in Germany, the two animals licking one another on the forehead, around the eye, on the back of the ears and on the neck. The females also rubbed their foreheads against each other's necks – Feer considered this was probably a mother-daughter pair. Feer also noted that his dominant buck had the most frequent contact with one female in particular and suggested that such associations could represent rudimentary coupling. Raymond Chaplin's video of the water deer he kept in his garden during the late 1960s and early 1970s shows one of the mature bucks ("Max") lying with an adult doe ("Pippa") while chewing cud, with Pippa even licking both of his preorbital glands. Chaplin told me that the two were virtually inseparable, spending a great deal of time "nuzzling" one another and always lying together at night while ruminating.

I have observed grooming between does twice during the summer of 2022, first at Whipsnade and then at Woburn. At Whipsnade, a group of three animals (a buck and two does) were feeding together when one doe walked over and licked at the hock of the other for several seconds while the subject grazed. The "licker" did not attempt to groom the buck, nor did the doe whose hock was being licked attempt to reciprocate. At Woburn, an adult doe rose from her couch along a fence line where she was ruminating and cantered out to meet a second doe approaching from across the field. The deer approached each other briskly and engaged in a bout of mutual grooming that lasted about 40 seconds, after which the first doe returned to her couch and resumed ruminating while the second moved further up the field to graze. During the grooming, each doe licked the other's face, rear end, ears and flank, and the second doe licked the first's right hock. I suspect that in both instances the two does were familial, possibly a doe and her daughter from the previous year (judging by a size difference), or siblings. Dubost's finding of no durable mother-offspring bonds means this can be no more than a presumption, however.

Grooming also occurs among younger animals. A video from South Korea that I have seen showed well-grown siblings grooming one another in the wild, one licking the preorbital glands of the other. At Whipsnade in mid-June 2024, I also witnessed a young fawn licking the preorbital gland of its mother, and on several occasions, both at Whipsnade and in the wild, I have seen twin fawns licking each other's faces, ears and flanks, and sniffing at each other's rumps.

At Branféré, Dubost and his team found licking to be very rare among water deer, recording only five bouts, all of which were a male licking the nose, body or rear of a female and none of which was reciprocated. This contrasts with the doe-to-doe licking that Feer reported and I have observed. The researchers also documented nose-to-nose sniffing between deer, as well as nose-to-rump sniffing when bucks were assessing a doe's oestrous phase during the rut, and I have observed both on several occasions in captivity and in the wild, including two instances of nose-to-rump sniffing by bucks outside the rut. Females touched noses with other females more often than they sniffed at their rears (178 vs. 68 encounters), while the opposite was true when males made contact with females (30 vs. 53 encounters). Females also touched noses with fawns on 214 occasions and sniffed their rears on 103; fawns did both with their mothers, with slightly more nose touching than rear sniffing. Stadler reported social sniffing to be a common behaviour at Whipsnade, although most animals avoided one another for a period before close-contact sniffing could occur. Sniffing at the tail was more likely to result in withdrawal than sniffing at the head.

↑ table of contents

Does with fawns, siblings and the effect of age

I'm aware of at least one example suggesting that adult females may associate even when they have young fawns in tow. On 25th June 2023, Sharon Scott observed two does walking across a field in Buckinghamshire in parallel -- i.e., about a metre (3 ft) apart, apparently travelling together -- each with a fawn following behind. Sharon wasn't sure whether the fawns were twins of one of the females, with the other doe simply tagging along, or whether each fawn was a singleton of the doe it was following, though Sharon noted how one fawn looked older than the other, which might suggest the second scenario.

In August 2025, one of my trail cameras recorded two does and two fawns in a bed of rushes beside a small pond, in the same area where Sharon made her observation, although all four were never in frame at once. An adult doe walked through the rushes from right to left, followed by a fawn. The doe left the frame and the fawn was almost out of shot (only its rump was visible) when it turned and came back into view. A second fawn had been sniffing around in the rushes at the same time, and was joined by a different adult female, which licked its back briefly before the pair exited the rushes onto the pond bank. The first fawn then joined them. Although two adult females and two fawns were in the same small area, the first doe was no longer in frame by this point, and the field of view was too narrow, and the recording too brief, to draw any conclusions about relatedness. This could simply be a doe with two fawns and a second female passing through, but the second female showed no aggression towards either fawn.

Age may also affect sociality. In his Whipsnade thesis, Stadler described yearling males roaming widely between May and July and avoiding social interaction where possible, such that interaction rates dropped to zero, or near zero, during this period. Male fawns beyond about three months old were also more likely to be found unaccompanied than females of the same age. Yearling females associated less with adult females during summer than winter, while they were found with adult males at a fairly continuous rate throughout the year. Overall (yearling males in early summer aside), fawns and yearlings of both sexes were more likely to be in pairs or small groups than adults of either sex. My experience has also been that it's quite common to see pairs of well-grown fawns together without any obvious adult affiliation during the late summer and autumn, suggesting siblings may remain in social contact after severing ties with their mother.

↑ table of contents

Interaction rates

In terms of frequency of interaction between individuals, the only data I have are again provided by Stadler. Monitoring of 23 animals during daylight showed that each interacted with another about 2.6 times per hour, on average, such that any given deer interacted every 23 minutes or so. Nearly half of all interactions (43%) happened during the rutting period, and only 10% during the fawning period. Average intervals between interactions were:

Water deer interaction rates (Stadler 1991)

Group	Duration between interactions
All animals	23 minutes
Males	18 minutes
Females	34 minutes
Winter (all animals)	22 minutes
Summer (all animals)	39 minutes
During rut (territory-holding males)	6 minutes
During rut (non-territorial males)	21 minutes

Purely in terms of interaction rates, males were the more socially active sex, albeit a substantial number of their interactions were aggressive and coincided with the rut.

↑ table of contents

Captive vs. wild data

Much of what we know about contact, grooming, sniffing and interaction rates in water deer comes from captive animals, whereas the wild data are largely limited to group sizes, pair bonds and a handful of anecdotes. In his thesis, Stadler suggested that the confines of captivity may force the manifestation of social behaviours that don't appear under more natural conditions. While this may be the case in some collections, the 50 or so deer at Whipsnade today have access to the 243 ha (600-acre) site, and the 40 or so at Woburn Abbey to the 1,214 ha (3,000-acre) deer park. Neither collection is crowded by comparison with some wild populations: more deer have been recorded in a single field on one Bedfordshire estate (see Group size and aggregations above) than both collections hold combined (see also: Population & Abundance – England), although conditions elsewhere were different -- Feer's animals at Wuppertal Zoo were kept in a 0.13 ha enclosure, for example -- and conditions at Branféré are less clear. Captive behaviour need not be a poor proxy for the wild, then, but nor can we assume it is a good one, and more wild observations are needed.

↑ table of contents

The bottom line

In essence, Chinese water deer cannot be considered a social species in the same way as the larger cervids, but nor are they strictly solitary. They appear broadly tolerant of other individuals in close proximity for much of the year, and in some circumstances small groups may form, possibly of related animals, although relatedness has not been established. Males at the peak of the rut, and females in the days before giving birth and while their fawns are largely immobile and concealed, are the broad exceptions to this tolerance. While bonds may endure for some time, we don't know for how long nor how commonplace their formation is in the wild. In captivity, persistent social bonds appear uncommon, but Stadler proposed, in his 1991 thesis, that the mild inclination of females to tolerate like-sex individuals may be interpreted as a first step in the evolution of group living in female deer. "Mildly sociable" therefore seems a fitting description of water deer, at least of females.

I would be very interested to hear from readers who have observations of, or thoughts about, social proclivities of water deer they'd like to share.

↑ table of contents

References

Allen, G.M. (1940). The Mammals of China and Mongolia. Natural History of Central Asia (Part 2). American Museum of Natural History, New York.

Cooke, A. & Farrell, L. (1983). Chinese Water Deer. The British Deer Society. No. 2 (Double Volume).

Dubost, G. et al. (2011). Social organization in the Chinese water deer, Hydropotes inermis. Acta Theriologica. 56: 189-198.

Feer, F. (1982). Quelques observations ethologiques sur l'Hydropote de Chine, Hydropotes inermis (Swinhoe, 1870) en captivite. Zeitschrift für Säugetierkunde. 47: 175-185. [In French]

Lawrence, R.P. (1982). Chinese water deer. Shooting Times and Country Magazine. (5174): 29.

Scherpe, W.P. (1971). Beobachtungen am Verhalten des chinesischen Wasserrehs (Hydropotes inermis, Swinhoe), des indischen Muntjaks (Muntiacus muntjak muntjak, Zimmermann) und des chinesischen Muntjaks (Muntiacus muntjak reevesi, Ogilby). Freie Universität. Berlin. [In German]

Stadler, S.G. (1991). Behaviour and social organisation of Chinese water deer (Hydropotes inermis) under semi-natural conditions. Ph.D. Thesis for Universitat Bielefeld, Germany.

Sun, L. (2002). Anti-predation does not fully explain grouping in the Chinese water deer (Hydropotes inermis). Acta Zoologica Sinica. 48: 302-308.