A Novel Method for Rearing Orphaned Donkey Foals with Behaviour in Mind

Written by Corinne McCafferty, MSc, CHBC


Summary: The suckling behavior of donkey foals has rarely been explored, with most orphans being reared on protocols based on the assumption that their behavior and physiology do not differ from horses. An orphaned donkey foal at The Donkey Sanctuary was reared on an artificial suckling system, enabling the analysis of suckling behavior in an orphaned donkey. Milk formula was accessible 24-hours per day via the artificial system and suckling behaviour was monitored, revealing that the foal suckled at a variable rate, with suckling durations, frequency and quantities differing from what was outlined in the standard bottle-feeding protocol for equines.


 While rare, orphaned foals present some of the greatest challenges in relation to equid care and welfare. People responsible for their care tend to adopt bottle-feeding protocols, feeding formula little and often during the first few days or weeks. Quickly, due to the demanding nature of hand-rearing, nocturnal feeds are phased out and, after the first few weeks, it is usually the case that foals are fed less frequently and in larger quantities. The implications to the overall well-being of the foal from this change have rarely been explored.

Hand-rearing is often the only possible option for equine sanctuaries when an orphan arrives, as other long-term alternatives, such as matching the foal with a lactating foster mare or inducing lactation in a barren mare, are not normally possible, or are unsatisfactory in terms of success.1,2 These techniques can be especially difficult within a rescue centre, where many of the animals have complex physical and emotional needs.3 It is the normal procedure for a foster mother with no milk yield to be placed with the foal, thereby providing for the orphan’s social needs.4 However, the responsibility of bucket- or bottle-feeding the foal with formula falls to the human handlers.

Foals are dependent on a milk-based diet for at least the first three to four months of life5 and this has significant cost and resource implications. Hand-rearing has an additional drawback in that the foals can receive too much inappropriate handling from humans6 and can often result in expressed behavioural problems leading to boundary issues that can become more dangerous as the foal grows. There are also concerns that, due to maternal deprivation, the young raised in this way are disadvantaged in terms of growth and behavioural development when compared to naturally raised peers.7

Suckling is a key facet of maternal nurturing8 and is now recognized as more than just a means of obtaining milk, for its frequency does not actually correlate with milk intake.9,10 Duncan et al. (1984)11 identified that in the first eight weeks, male foals were likely to spend 40% more time suckling than females; however, this significant finding is rarely accommodated in hand-rearing protocols where the demands of bottle-feeding rarely afford adaptations for individual differences. Moreover, any attempt at bespoke feeding requires making assumptions about that individual foal’s preferences and is thereby imposed, rather than self-determined. The apparent lack of research and published literature on the natural suckling behaviour of donkeys has resulted in rearing protocols being often founded on the natural behaviour of horses, under the fallacious assumption that their behaviour is the same. Given the importance of this period in the foals’ overall development12 and how this time may be expressed in future behaviour, research into the potential of alternative protocols is perhaps overdue.

Placing the foal on an automatic feeding system, and thereby eliminating the considerable human resource element associated with hand-feeding, allows the foal some degree of freedom to self-administer and, potentially, to feed more naturally. In turn, this approach may not only yield a resolution to many of the challenges associated with bottle-feeding but may have a positive effect on the foal’s overall welfare. De Palo et al. (2016)13 established that, compared to naturally suckled foals that were fasted during the day, a cohort of foals that were artificially suckled during separation from the mare maintained higher weekly weight gains. Furthermore, in a separate study, De Palo et al. (2018)14 considered the biochemical and oxidative blood profiles of these donkeys that were partially artificially suckled to those who were entirely naturally suckled with periods of fasting. They established that almost all of the analyses varied between the groups and concluded that the suckling technique affects biochemical profiles and lipid peroxidation patterns in donkey foals. These findings provide some grounding for trialing an ad lib feeding system for an orphan, as this may promote healthy growth and development.

The primary aim of this case study was to identify and evaluate the effects on an orphan donkey foal from the use of an automatic suckling system. The quantities consumed, along with records of duration and frequency, were collected daily and a daily qualitative behavior assessment (QBA)15 of the emotional state of the animal was carried out along with ethogram analyses. It was predicted that the foal would choose to drink little and often if given the chance, and that this would allow more freedom to exhibit natural behaviour, thus reducing frustration and behavioural problems. Furthermore, separating the feeding from the human handling would further diminish the opportunity for the foal to develop behavioural problems often associated with hand-rearing.

Case study


The subject of the case study was a grey standard donkey colt. He was orphaned at 1 week old and admitted to the New Arrivals Unit at The Donkey Sanctuary, Sidmouth, U.K. Prior to relinquishment to the sanctuary he had been with his mother for one week and was able to feed from her; however, once arrangements had been made to bring him into the sanctuary, the mare was euthanised due to complications during her labour that had rendered her unable to stand. Upon arrival, he was bottle-fed every two to three hours and stabled alone for one week until a foster mare could be selected and relocated from another site at The Donkey Sanctuary. After this time he was housed with the foster mare and had access to a field during the day. He was bottle-fed by hand for four weeks and then transferred onto the artificial system.

Image 1. The foal drinking from the teat on the back of the stable door

The artificial suckling system

Grooms simulated hand bottle-feeding to initially prompt nursing from the teat and the foal picked it up within a couple of days, being closely monitored for signs of stress throughout the transition period and receiving regular checks around the clock. The product used was the Heatwave Milk Warmer. A lamb bottle teat was attached to the stable door at an appropriate height and a hole was drilled in the door. The milk formula was passed through the machine via plastic tubing leading from an enclosed bucket to the teat when the foal suckled. The Heatwave Milk Warmer works by heating the formula in the plastic tubing as it passes through the machine. It heated the milk formula to donkey body temperature (36.2 – 37.8 degrees C). A motion-activated wildlife trail camera was fitted directly above the teat on the roof of the stable and was used to record the foal every time he suckled from the teat. Once the foal moved away, the camera would go to standby. In this way, the camera timed and recorded all of the nursing events, providing a detailed video record that was then reviewed and analysed at a later stage.

A table outside a stable. On the table there is a milk warmer and a bucket of milk replacer. A plastic tube leads to the left of the milk warmer, to an artificial teat. A wildlife camera is set up to view behind the door.

Image 2. Artificial suckling system setup and operation

The four main behavioural parameters of suckling behaviour: initiation of suckling bout, suckling bout duration, suckling bout frequency and suckling bout termination,16 were recorded 24 hours a day over a 16-week period. The collected data included totals and averages of the frequency of suckling bouts, the duration of the suckling bouts and the quantities of formula consumed. The latter were calculated daily over 16 weeks, with “day” being defined as 0700-1900 and “night” as 1900-0700. Emergent trends identified within these data were then compared to the figures outlined in The Donkey Sanctuary’s bottle-feeding protocol..

When ad lib milk formula was available via the Heatwave Milk Warmer, the foal adopted a highly variable pattern of drinking (Figure 1). At times, he exceeded the recommended daily amount of 10% of body weight6 and yet, at other times, he suckled well below this level. He consistently alternated between drinking large and small quantities in stark contrast to his behaviour during the weaning period (days 85 – 111, Figure 1) when quantities were heavily controlled. This could imply, therefore, that foals are capable of some degree of self-regulation over feeding quantities. There was a significant positive correlation between the volume of formula consumed and the time spent suckling, r=0.32, p<0.01, a result that seemingly runs counter to the findings of previous studies concerning suckling durations and milk consumption10,17 that could not establish such a correlation.

A graph illustrating amount of formula consumed daily. The graph trends downwards.

Figure 1. Quantities of formula consumed daily (18/01/19 – 17/05/19)

For the first six weeks of the study period, which commenced when the foal was 1 month old, he consumed an average of 6 litres of formula per day (Figure 2). From this point, consumption steadily grew to 8 litres on Week 9 and then immediately fell rapidly to 5 litres during the following week. Gradual weaning commenced at Week 12, from which point the system was capped at a limit of 4 litres. It was then further reduced to 3 litres on Week 15, and these falls are evident in the averages for these weeks in Figure 2. These data identify that during weaning the foal would usually consume all of the available milk. This is in accordance with the observed behaviour of foals during weaning from the artificial system that has been identified by De Palo et al. (2018).14

A graph showing the daily average of formula consumed. The graph trends downwards with a marked peak around 5 weeks.

Figure 2. Daily averages of formula consumed over 16 weeks (18/01/19 – 17/05/19)

Nocturnal and diurnal variations in the quantities consumed were clearly observed (Figure 3). It was often the case that when large quantities were consumed in one half of a 24-hour cycle, it tended to be offset by the foal drinking less in the second half of the same cycle. This is demonstrated in the data for Day 24 (Figure 3). At other times, where more moderate quantities were consumed, the suckling quantities were quite evenly distributed across both time periods. It was perhaps not surprising that larger quantities were consumed during the day, rather than at night, as this was a far more active period in general since the mare and foal had access to a field and staff were present, handling the donkeys and completing daily routines.

Graph showing difference in day vs night consumption of formula.

Figure 3. Quantities suckled from 0700-1900 (day) vs. 1900-0700 (night)

Throughout the period of the study, the foal drank more than the outlined reported norms, consistently drinking between 3 and 10 litres, and often drinking more than 10% of his total body weight. Additionally, he consistently took more feeds per day than would have been offered on the bottle-feeding protocol. Although showing some variability, the number of feeds ranged from 10 to 35 per day (Figure 4). This suggests that, when bottle-feeding, a foal is constrained to only perform their natural suckling behaviour at specific times of the day and they are unable to nurse at their preferred frequency.

Figure 4. Average total number of feeds when bottle-fed and when artificially suckled

The volume of individual feeds taken was smaller when the foal was artificially suckling, usually staying within the 100-500 ml average range. This was in contrast to the bottle-feeding protocol, where fewer, larger feeds of around 1 litre were delivered. It seemed that, given the choice, maintaining a variable pattern of consumption was preferable (Figure 5).

Figure 5. Average feed sizes per suckle when bottle fed and when artificially suckled.

Ethogram analysis

An ethogram for equines nursing on an artificial system was developed by collating the relevant nursing behaviours from both McDonnell (2003)18 and Tateo et al. (2009).19 The primary behavioural focus was on those observed behaviours relevant to the artificial system. This ethogram was therefore used to record and measure the range of natural behaviours that the foal was able to perform while using the system.

Suckling duration – number of minutes spent suckling
Suckling frequency – number of suckling bouts per day
Average suckling bout (in minutes)


Total ml of formula consumed
Average ml of formula consumed per suckling bout
Non-nutritive suckling frequency – suckling for purposes other than receiving milk formula
Licking and chewing nearby structures frequency (number per day)
Rubbing face on nearby structures (number per day)
Parallel nursing frequency (number per day) – foal standing parallel to teat
Perpendicular nursing frequency (number per day) – foal standing at 90 degree angle to teat
Conflict with foster mare frequency (number per day) – frustration, resource guarding, kicking, biting, ears back, aggressive/defensive facial expression from foster mare or foal or both
Conflict with teat frequency (number per day) – frustration, aggressive/defensive facial expression, inability to obtain milk though effort made from foal
Butting the teat frequency (number per day) – bumping teat with nose intentionally
Standing beside teat without suckling frequency (number per day)
Immediately walking away after suckle frequency (number per day)
Nurse after rest frequency observed (number per day)
Nurse after separation from the teat frequency observed (number per day)
Nurse after play behaviour frequency observed (number per day) – attempts to play with toys, people or the foster mare

The foal performed a wide range of established natural behaviours relating to nursing when placed on the artificial suckling system. Examples included the choice to stay close to the teat without drinking, a behaviour regularly observed at the beginning of the study period. Such behaviour may be predicated on feelings of security that bottle-feeding does not provide. It certainly closely mirrors the observed behaviour of foals and their mothers in natural circumstances, where foals stay relatively close to their mother in the first five weeks of life.20 The foal was also observed immediately departing from the teat upon cessation of nursing. The number of these occurrences increased drastically in the final weeks, which happened to correlate with the teat becoming less dependable as a source of food due to the restricted quantities being administered at weaning. This could imply that the foal became increasingly independent and thereby less reliant on the teat as a food source, potentially demonstrating behaviour that also mirrors the natural development of independence at this time.20

Figure 6 shows the foal butting the teat with his nose, licking or chewing it, and rubbing his head on the teat and on nearby structures, all behaviours that are considered to be non-nutritive suckling.21 Indeed, non-nutritive suckling behaviour was observed throughout the study period, even at times when milk was restricted. Furthermore, it increased towards weaning, indicating perhaps that although nutritive suckling decreased as milk became unavailable, general suckling still formed a large portion of his time budget. An increase in butting around weaning (Figure 6, Weeks 13 and 15), which is a behaviour recognized in most ungulate species and performed to encourage the mothers’ milk to flow,22 was a natural response to the reduction in milk availability during these weeks. While such behaviour could indicate frustration, this is a natural behaviour and it is likely that in this instance it does not exceed any normal levels of frustration experienced in natural circumstances as mothers do not produce infinite amounts of milk and they will sometimes deny suckling attempts.22 However, this expected behaviour would not be possible to perform on a bottle-feeding program, and an absence of the behaviour during bottle feeding does not indicate that the feelings of frustration and stress are not present when milk is reduced, merely that frustration cannot be expressed naturally. The concern is that this suppression could result in the development of abnormal oral behaviour in future.23 It appears that non-nutritive suckling can also be a form of comfort-seeking24 and therefore implies that artificially suckled foals may self-sooth in a way that bottle-fed orphans cannot.

These data were limited with regards to witnessing events that preceded suckling, such as rest, separation from the teat, and play, due to the camera placement being above the teat. Future studies on the topic should consider the use of multiple cameras to capture this more fully. However, all three behaviours were observed to be preceding events at some point during the 16-weeks period. Where periods of recumbence were captured on camera it was often followed by a suckling bout. Periods where the foal was separated from the teat due to being taken for walks, being shut out of the stable for mucking out, and the removal of the teat for machine cleaning were also regularly followed by suckling bouts. Bouts of play were hard to capture on camera, as these most often occurred in the field. However, when they were captured there were times when they too were followed by post-activity suckling. This demonstrated that at times when the foal felt a natural drive to feed, such as immediately following periods of rest, activity, or deprivation of milk, he could fulfil this drive when a bottle-fed foal may not be able to.

Two different nursing positions were observed during this case. Firstly, parallel nursing, when the body was positioned parallel to the door/wall where the teat is fixed. Secondly, perpendicular nursing, where the body is positioned at right angles to the door/wall. Often, the foal adopted a position somewhere in between parallel and perpendicular. A “reverse” parallel position was not observed, because the foal could only stand parallel to the stable wall in one direction, forcing him to either nurse from the left side or from straight on. In this instance, therefore, it cannot be concluded that artificial suckling will always allow full flexibility of body position, as this study was constrained by stable space and the operative setup of the system. Due to this restriction on space, the previously observed natural side preference of foals25 could not be observed even if it had been present. However, even in such limited space, more autonomy over body positioning was observed, and it was certainly more than bottle-feeding provides in cases where the human feeder maintains control over the foal’s body positioning.

Figure 6. Four types of non-nutritive nursing behaviour recorded over 16-weeks (18/01/19 – 17/05/19)

Qualitative behaviour assessment

The Qualitative Behavioural Assessment (QBA) protocol developed by Wemelsfelder (2001)15 was adopted to monitor the foal’s emotional state in this case study. The extant descriptors were adapted specifically for donkeys, and particularly for those being held in equine facilities, through collaboration between Wemelsfelder and The Donkey Sanctuary. For example, the additional term “bored” was subsequently included.  During QBA, a score between 0 and 125 on a visual analogue scale was given for each descriptor (outlined in Table 2).

Descriptor Definition
At Ease Calm with other animals, interested but with appropriate calm reactions to changes, carefree, resting in close proximity to other animals. Maintains normal relaxed behaviour around people or other animals.
Curious Inquisitive, interested in surroundings, interacts with new or novel objects, people or animals,  investigates, explores environment objects or people.
Friendly Companionable, affectionate, sociable; not hostile. Positive reactions towards another person or animal, seeks attention from human (i.e., the donkey approaches another animal/person and expresses grooming behaviour, may sniff or interact in some way).
Happy Content; feeling, showing or expressing joy, pleased, lively, playing, satisfied, relaxed alertness to environment, interested in surroundings.
Playful Very active on own or with others, actively seeking play opportunities, frisky/frolicsome, playing with toys/objects, reciprocal play fighting. Trying to encourage other animals in play.
Relaxed Not tense or rigid, easy-going, calm, tranquil, standing still while resting a foot, normal level head position, minimal relaxed tail swishing, calm awareness/interest in surroundings.
Responsive Reacts appropriately to behaviour of other animals and people, active, acknowledging, receptive, aware of the environment, shows appropriate responses to stimuli, vocalising.
Apathetic Having or showing little or no emotion, uninterested, indifferent, not responsive to stimuli, little or no ear movement or ears low, little or no interaction with other donkeys, depressed, not moving or moving slowly, low head position, appearing to behave stoically, shutdown.
Agitated A negative emotional state, highly active, restless, fidgety, excited, worried/upset, disturbed, in a bad mood, annoyed, stereotypy – weaving, fence pacing, irritated by flies, head shaking, muscle twitch, rapid and rigid tail swishing, persistent ear movements.
Aggressive Hostile, attacking, defensive, disruptive, angry, wants to fight/attack another donkey, intention to harm, displays kick threats, bite threats, head tossing, ears flat-back against neck, vigorous tail swishing towards other animals or people, physical contact, biting, chasing, kicking.
Anxious General assessment of character of individual or group. Worried/tense, troubled, apprehensive, distressed, jumpy, nervous, watchful, responsive to a possible threat/danger (i.e., looking around/vigilant, moving ears, opposite of at ease).
Distressed Upset, afflicted, distraught, worried, defensive behaviours, high levels of movement and resistance to restraint (i.e., high resistance to handling, attempts to escape, abnormal defecation, rearing up).
Fearful Startled, afraid, hesitant, timid, uneasy, may often be linked to something going on in the environment, flight response, move or run away, back up, refuse to move further, may show defensive behaviour, bite threats, kick threats, or avoidance behaviour, such as pushing handler.
Pushy Assertive or forceful, gaining access to physical resource such as food, water, or human contact by physical means, displacement of another donkey, head butt out the way.
Uncomfortable Painful, irritated, repetitive behaviour such as rubbing, biting, or scratching particular spots of the body, (e.g., shooing away flies, trying to remove a too-tight head-collar) foot stamping, weight shifting, showing conflict or uncertainty with tense muzzle, back or tucked tail.
Withdrawn Unsociable, introverted, reclusive, not wanting to communicate with other animals or people, uncommunicative, reserved, unresponsive, not searching for contact with others, solitary.
Bored Wandering aimlessly without direction or intention, but not relaxed or resting, behaviour such as wood chewing, unnatural repetitive behaviour, general lethargy not caused by excessive workload,

Five consecutive minutes of video were selected at random and watched and scored twice per day — one during the day and one during the night. Dr. Stuart L. Norris, The Donkey Sanctuary’s senior statistician, ran the QBA analyses. Initial analysis was performed using principal component analysis for the FactoMineR26 R package, data was log 10 transformed. Principal Component Axis 1 and principal Component Axis 2 were showed to understand the mood and energy of the foal respectively. Once the principal component scores for each of the axes were calculated, these were displayed using biplots produced using the ‘ggplot2” R package.27 After principal component analysis had been completed the two QBA axes were separated out for further analysis. The mean and standard error for the PC1 (Mood) and PC2 (Energy) was calculated, analysis of variance was then applied to the raw scores for PC1 and PC2 separately to test for significance with further Tukey post hoc tests to test where significance lay, in either the “Mood” or “Energy,” using the R package “car”.28 Further analysis to understand difference in the scores from week to week was preformed using similarity percentage analysis from the R package “vegan”.29 All statistical analysis was performed using R v5.2 and RStudio v1.2.30,31

Many positive emotions in both high- and low-energy quadrants were observed during the 16-week period including relaxed, at ease, responsive, curious, playful, and happy. These results can be seen in the red clusters in Figure 7. At times, there was some degree of agitated behaviour observed, especially where the foal ran into conflict with the foster mare, or the teat was not relinquishing milk (which happened occasionally when an air bubble got into the tubing). Boredom was also observed at times when he was in view of the camera but not suckling, suggesting that suckling could serve as environmental enrichment for orphans when there is little else to do. This manifested as him butting the teat for prolonged periods even when milk was available and there were no blockages in the tubing. These results appeared to support that the artificial suckling system on the foal’s mental wellbeing.

Figure 7. QBA of the orphan using an artificial suckling system

Clinical parameters

Throughout the duration of the study, the clinical health of the foal was monitored and evaluated by The Donkey Sanctuary’s veterinarians. The daily management of the foal and the machinery was the duty of The Donkey Sanctuary employees, mainly grooms. The sanctuary’s behaviour team monitored the foal’s behavioural management and training. The foal’s weight increased steadily over the course of 16 weeks, with the biggest weight gain observed between Week 2 and Week 3 (Figure 8). Weight remained constant from Week 12 to 13, the point at which the milk supply was capped at 4 litres. Furthermore, there were no changes from Week 15 to 16 when the milk available was reduced still further to 3 litres. This could imply some degree of weaning stress or, more simply, could be a reaction to a reduction in milk. Any stress suffered was unlikely to exceed the levels generally associated with weaning from a mare’s milk; however, where stress is an issue, decreasing milk more gradually may be a preferred option.

In comparison to other foals born at The Donkey Sanctuary in 2018 and 2019, the foal in this study (Sam) was, at the first recorded weight, one of the lightest in the cohort, with only one foal (Casey) weighing less (Figure 9). However, by 4 months of age, Sam had become heaviest amongst his contemporaries, demonstrating a substantial increase in weight while being reared on the artificial system. All of the other foals were naturally suckled except foal Ben K, who was hand-reared on The Donkey Sanctuary bottle-feeding protocol after being rejected at birth. Though Ben K had a higher starting weight than Sam, his weight increased at a slower rate than all other foals (Figure 9). This could indicate that foals are not fed enough under bottle-feeding protocols, but further study is needed with a wider subject pool to establish whether the trend witnessed here holds any significance. As these foals are coming into a rescue setting there is also the potential that the mother is not always in full health, and it is often the case that the full medical history of the mother is unknown. Therefore, tailoring a bottle-feeding protocol based upon these factors is not possible. The development of a standardized donkey foal growth chart would also be of benefit to further investigations, as none currently have been developed.

Figure 9. Foal growth rates at The Donkey Sanctuary 2018-2019

There were several issues that were highlighted by this case study. The foal exhibited excessive scouring (diarrhea) while being fed milk formula ad lib. This scouring was present from birth, four weeks before being offered ad lib milk, with the vet noting:

“Check as loose faeces. Temp 38.5c, BAR, grooms report eating well, no faecal staining of hind legs. Seemed hungry between feeds, drank 100mls warm water – grooms can give this between feeds if needed to ensure not dehydrated. Urinated well, normal colour. HR WNL’s.”

This scouring was not abated by the use of gastrointestinal medication and only got better after the milk was reduced. All clinical parameters remained healthy throughout the rearing period. The supervising vet was not concerned,, noting:

“Reported continued loose faeces HR 60, RR N + T 37.8c mm pink NAD on clinical exam. Not much faeces seen in stable, clinically normal, bright, alert, strong very likely to be caused by over feeding – as milk ad lib – as not showing any signs of disease.”

The milk was reduced and the vet reported that the “consistency of faeces had improved noticeably.”  This corroborates findings that high milk replacer intake prolongs diarrhoea.32 Therefore, it seems likely that the problem was caused by overconsumption of formula, but as the issue was present for four weeks prior to artificial suckling, other causes cannot be ruled out. The Aintree foal milk replacer is designed for consumption by horse foals, and thus the formula closely mirrors that of mares’ milk. Previous research has shown that the milks of different equid species are highly similar.33 However, more recently, a 2015 study found that horse milk had a higher fat, protein and energy content, and lower lactose, than donkey milk.34 Thus, it is possible that this type of milk replacer could be the source of the upset due to these compositional differences. Unfortunately, milk replacer formula made specifically for donkeys is expensive and hard to find; most milk replacer is marketed for human consumption rather than orphaned foals.


The foal exhibited natural behaviour for a donkey colt. Male donkeys can be extremely playful, often engaging in aggressive play well into adulthood.35 Foals learn a lot about the world by biting things and interacting with them in this manner. There has also been evidence that bottle-reared foals exhibit more inappropriate behaviour toward people due to “over-handling,” or rather, incorrect handling.6 Behaviours such as rearing and biting are often reinforced by accident when people push the foal away, as the foal perceives this as positive, playful attention. People will often encourage or ignore this behaviour when foals are small because they find the behaviour cute, but then come to realise their mistake as the donkey grows and the behaviour becomes more dangerous. There are also concerns that orphaned foals are at a disadvantage compared to naturally raised peers in terms of physical and behavioural development due to lack of maternal investment.7

To address the foal’s aggressive play behaviour, the behaviour team (comprising Ben Hart, CHCB, and the author) did a couple of sessions with the foal and the grooms responsible for his care to demonstrate how to use positive reinforcement to modify the behaviour. A protocol was put in place to reward the foal when he stood still and calmly beside people with his head facing the front, instead of turning to nibble. Grooms were asked to carry out dedicated sessions to teach him to stand still with a relaxed, forward head carriage in exchange for scratches and human attention. They were asked to ignore his attempts to bite or jump up at them and to wait patiently for him to offer standing still, as retreating would only cause him to escalate the behaviour and chase them. Once the foal got the hang of this, his handlers were asked to stretch the duration of his patience by pausing for 1 second, then 2 seconds, and so on, using a shaped approach until he could wait patiently while standing still for his scratches. Once they’d mastered this, the handlers were asked to systematically desensitise him to their leaving/walking away from him. We asked them to start by leaning/turning away from the foal, and then stretching this so they were taking 1, then 2, then 3 steps away from him and so on, rewarding with scratches for standing still at each step. We explained that they could then stretch the duration by taking some steps away, pausing for 1 second, and then approaching again.

Image 3. Foal in training session with Ben Hart, CHBC

This plan worked to an extent, and the foal’s behaviour improved, especially on a head collar. However, with over ten grooms responsible for his care, it was extremely challenging to maintain consistency between handlers. Moreover, the foal was still fairly socially deprived with only his foster mother for company, and so his natural drive to play socially was not being fulfilled. In order to improve consistency and minimise risk, grooms were encouraged to work in protected contact, shutting the foal away during mucking out and poo-picking, as his rearing and biting behaviour became more disruptive as he grew larger.

Image 4. Training using the figure-of-eight rope, with Ben Hart, CHBC

Sessions with the grooms in charge of the foal’s care continued, and they were taught how to use the figure-of-eight rope in order to train the foal to lead and wear a head collar. The figure-of-eight rope is a safer way to train a foal to respond to light pressure and release cues. Foals will instinctively push against pressure, which can be dangerous and cause the foal to rear up and flip over when you begin to apply pressure to their head using a head collar and lead rope. The figure-of-eight rope allows the handler to apply pressure to the chest and back end, teaching the foal to move off pressure in different directions. It also allows the handler to restrain the foal safely and securely for husbandry procedures, such as vet or farrier visits that must take place before the full training process can occur. From here, the handler can start getting the donkey used to having their body and legs touched, and can begin to get the foal used to wearing a head collar without rushing to use it to apply pressure or restrain the animal.

With the support of the behaviour team, the sanctuary grooms were able to make progress with the foal’s routine handling and training for husbandry procedures, getting him to the point where he was safe to handle and lead out for walks to explore the wider world. This training had a positive impact on his overall behaviour, as he was less frustrated in the presence of people and his life was enriched by being able to go for walks. He did continue to show some inappropriate behaviour when loose in the field or stable. After being weaned from the artificial system, the foal was able to be moved out of the New Arrivals Unit and onto a farm where he could learn to interact with other donkeys. Here he was introduced to another colt around his age, a 2-year-old gelding, a 3-year-old mare and a 5-year-old mare (mother of the other colt). The foal thrived with more opportunity to play and interact with other donkeys and eventually was weaned from his foster mother when his bond with her diminished, which enabled her to rejoin her old group with more appropriate companions.

The behaviour team visited this new location on two occasions to discuss the foal’s management and training, talking the new team of grooms through the same procedure for training self-control and patience. In this location with other donkeys around, he became much easier to train, though remained naturally playful and people-oriented. Importantly, he did stop rearing up onto his handlers and there were significantly fewer instances of biting. The foal was castrated once he was old enough and his testicles had descended, after which he settled further. The new team responsible for his care were able to take over his training without further input from the behaviour team and the foal has now moved on to live with a foster family in a private home as part of The Donkey Sanctuary’s Guardianship Scheme. You can read about him in his new home here.

The Human Element

The grooms in charge of the foal’s care were much happier that he was able to feed without the need for a bottle, as this reduced their workload and stress significantly. It also meant they could focus on training without confusing him by offering bottle feeds, which may have served to reinforce his existing behavioural issues.

Thus, the onsite observations of the grooms were recognized as an important further source of data for this case study and they were subsequently interviewed about their experience of implementing the artificial suckling system. Given the timeframe and the varying degrees of involvement across the duration of the case, the interviews were unstructured, with no specific pre-determined questions.

When asked if using the system to rear the foal made their job easier, one response fairly common to all was:

“Yes, great that we didn’t have to do night feeds – or bottle-feeding full stop once the system was up and running.”

They were then asked if they had any practical issues in using the machine, to which they responded:

“No, other than setting it up the first day. We had some trouble replacing the correct cleaner for the machine as it isn’t stocked in stores and has to be ordered in specially.”

In spite of these issues, it does seem that employing an artificial system reduced workload for those involved in the rearing process and had a positive effect on overall wellbeing of the people in this case. The system does come with its own set of challenges and concerns, but as this was the first time the system had been utilized in this setting it was a learning curve for all involved.

Since this case, the system has been implemented for a second time at The Donkey Sanctuary Ireland, who were responsible for the care of a rejected donkey foal. The Irish team built a structure in which to house the system, with space for the foal to enter to reach the teat. This feeding-system can be seen in Image 5, containing two pictures, one during the first few weeks on the system, and the other after several months. You can see here that foal was much bigger and near the point of weaning in the second picture. This structure used to house the machine and the teat improved upon the previous system in some ways, as this foal was able to live with other similarly aged companions, without them disrupting the system. This was extremely beneficial as the foal was less socially isolated. The team reported no issues with other individuals attempting to drink from the system. One drawback was that the foal had less autonomy over his body position, due to the structure limiting him to a perpendicular nursing position.

Image 5. Feeding system built for feeding and to house Heatwave Milk Warmer

Having now successfully reared two foals on this system, we were able to put together some helpful tips for any others who wish to try it:

  • Monitor the quantities of formula consumed on a regular basis. Use a milk container labelled with ml measurements to store the formula and record the quantities drunk every couple of hours.
  • Decide on the maximum amount of milk you wish the foal to consume per day (consult your vet about this) and don’t exceed this volume within a 24-hour period. You can top up the machine at set times, with set amounts, every day so that the foal doesn’t finish all of the milk at once and is then left with nothing for long periods of time. If the foal is regularly allowed to exceed the ideal amount it may result in excessive scouring, so monitor the foal’s droppings closely.
  • Allow access to milk both day and night; the foal will feed through the night if given the chance to.
  • You must cover the milk container to prevent other animals (such as cats) getting into it. It’s also necessary to weigh the container down in some way so that it is not knocked/blown over in the night and to protect it from the elements in general.
  • Keep the machine clean and sterile by rinsing it through every day and priming it, following the instructions.
  • Regularly check that milk is coming through the tubes easily as it can become air locked, in which case the machine will need to be cleaned/primed or the air bubble will need to be sucked through.
  • Weight the foal regularly and keep a record of their weight gain.
  • Foals like to suckle after bouts of play or rest, or after being separated from the teat for any period of time, so plan routine milk top-ups around your daily schedule to allow for this.
  • Wean gradually, lowering the quantities at each routine milk top-up rather than just cutting out whole feeds altogether. This will allow for normal consumption little and often, instead of a few large feeds.
  • Don’t remove the teat for long periods during weaning as the foal will become frustrated and may lick the milk traces where the teat used to be. Foals will suckle for comfort/something to do even if there is no milk. You can replace the milk with water instead to provide something of lower nutritional value and to ensure they don’t become dehydrated.
  • Provide plenty of environmental enrichment and make sure the foal is with good company; otherwise they may suckle excessively out of boredom.
  • It is essential to keep up the training with your foal as you normally would. Using the system to reduce problems associated with over-handling is not enough to produce a foal without behaviour problems.
  • If you are rearing multiple foals on the system, ensure you provide enough milk for all of them. Monitor them carefully, ideally with a camera, so that you can ensure each foal has access. Some systems may allow for multiple teats to be attached, and these would be preferable in cases where multiple foals are to be fed (for example, in the donkey-milk industry).

Final thoughts

The observations made in this case study were used to explore and test the use of the artificial suckling system and demonstrated that it allows for greater autonomy over feeding behaviour than a bottle-feeding system provides. Not only did it allow the foal to drink little and often, but it also allowed him to express a greater range of natural behaviours. The importance of such findings is enhanced when the Five Domains Model of animal welfare are considered.36 In line with the framework’s goals, the foal in this case study had a “life worth living” thanks to the artificial suckling system, which had positive impacts on his emotional wellbeing as well as on his growth and physical development.

The findings support that rearing orphaned foals on the artificial suckling system should be explored as an option in the future, as this will allow us to monitor emerging behavioural trends and arrive at conclusions about what is natural for domestic donkeys. Suckling quantities and bout frequencies and durations will continue to be of particular interest as they will allow equine organisations, such as The Donkey Sanctuary, and individual owners to refine the orphan rearing protocols to best fit the needs of donkeys and treat each one as an individual. The current bottle-feeding protocol differed drastically from the foal’s natural inclinations and the influence of this on foals in their early stages of development and subsequent behaviour will require further monitoring to ensure that orphans are being raised using the best practices available. This means improve existing bottle-feeding protocols with the Five Domains Model in mind in cases where the artificial suckling system is unable to be adopted.

The ethogram developed in this case study should be regarded as an active and systemic development to be utilized and expanded upon as more foals are reared on the system, but it was able to measure a wide range of natural behaviours nonetheless. Indeed, ad lib, self-directed feeding seemed to increase behavioural expression. With further research, it will be possible to devise a more accurate list of natural behaviours surrounding suckling for donkeys and make more direct comparisons between those being reared via bottle-feeding and their naturally suckled peers. The potential existence of any individual differences, including sex and breed differences, is also an avenue in need of exploration. The QBA can continue to be utilized to identify areas where frustration and boredom cause problems in orphan foals. Such recognition will allow those caring for donkey orphans to anticipate periods of negative emotion, for example weaning, and help to manage them effectively so as to reduce the risk of hyperlipemia.37 It is also important that we recognize positive experiences and capture data that will support the adoption of protocols that foster positive emotion in donkeys in line with the Five Domains Model.

The variable drinking pattern observed provided a fascinating insight into the natural self-regulation of the foal’s drinking, something that has not yet been explored in detail. When a foal is feeding from a mare, it is extremely challenging to measure the quantities of milk consumed, so this case study allowed us to accurately record the quantities a foal would drink given the choice. It shows that the quantity and regularity of suckles has been underestimated and preferred meal sizes have been overestimated in bottle-feeding protocols. This would suggest that current protocols are based on what is most convenient for humans, not what is best for foals. Such an approach results in the imposition of what are regarded as the optimum quantities of milk, rather than allowing the foal to dictate the feeding regimen. Current estimates of how much donkey foals should consume are largely based on research conducted on horses, so there is no guarantee that these are accurate for donkey orphans. Therefore, the opportunity to observe this process has provided some species-specific insights, enabling donkey foals to be reared on species-appropriate protocols in the future.

This opportunity to study an orphaned donkey so rigorously was seized as the best attempts were made to give the foal the best start in life, with his welfare being of the upmost importance. The welfare of donkey orphans will continue to improve as species-specific adaptations are made to protocols and systems that allow for more behavioural autonomy. However, with such implementation there also needs to be an allied recognition amongst those who care for donkeys that offering freedom of choice should take primacy over simple logistical convenience.


  1. Grogan, E. & Mcdonnell, S. (2005). Mare and Foal Bonding and Problems. Clinical Techniques in Equine Practice 228-237.
  2. Stoneham, S.J., Morresey, P., & Ousey, J. (2017). Nutritional management and practical feeding of the orphan foal. Equine Veterinary Education, 29:3, 165-173.
  3. Holcomb, K.E., Stull, C.L., & Kass, P.H. (2010). Unwanted horses: The role of non-profit equine rescue and sanctuary organizations. Journal of Animal Science, 88:12, 4142-4150.
  4. French, J. M. (1998). Mother–offspring relationships in donkeys. Applied Animal Behavior Science, 60:2-3, 253-258.
  5. McKenzie III, H.C., & Geor, R.J. (2009). Feeding management of sick neonatal foals. Veterinary Clinics of North America: Equine Practice, 25:1, 109-119.
  6. Naylor, J.M., & Bell, R. (1985). Raising the orphan foal. Veterinary Clinics of North America: Equine Practice, 1:1, 169-178.
  7. Latham, N.R., & Mason, G.J. (2008). Maternal deprivation and the development of stereotypic behaviour. Applied Animal Behaviour Science, 110:1-2, 84-108.
  8. Crowell-Davis, S.L., & Weeks, J.W. (2005). Maternal behaviour and mare-foal interaction. In: The Domestic Horse: The Origins, Management and Development of Its Behaviour. New York: Cambridge University Press, 126-138.
  9. Cameron, E.Z. et al (1999). Suckling behaviour does not measure milk intake in horses, Equus caballus. Animal Behaviour, 57:3, 673-678.
  10. Cameron, E.Z. (1998). Is suckling behaviour a useful predictor of milk intake? A review. Animal Behaviour, 56:3, 521-532.
  11. Duncan, P., Harvey, P.H., & Wells, S.M. (1984). On lactation and associated behaviour in a natural herd of horses. Animal Behaviour, 32:1, 255-263.
  12. Henry, S., Richard‐Yris, M.A., & Hausberger, M. (2006). Influence of various early human–foal interferences on subsequent human–foal relationship. Developmental Psychobiology: The Journal of the International Society for Developmental Psychobiology, 48:8,
  13. De Palo, P. et al (2016). Artificial suckling in Martina Franca donkey foals: effect on in vivo performances and carcass composition. Tropical Animal Health and Production, 48:1, 167-173.
  14. De Palo, P. et al (2018). Survey of biochemical and oxidative profile in donkey foals suckled with one natural and one semi-artificial technique. PLoS One, 13:6, e0198774.
  15. Wemelsfelder, F. et al (2001). Assessing the ‘whole animal’: a free choice profiling approach. Animal Behavior, 62:2, 209-220.
  16. King, S.R. et al (2016). Behavior of horses, zebras, and asses. In: Wild Equids: Ecology, Management, and Conservation, P. Kaczensky & J.I. Rensom (eds) MD: Johns Hopkins University Press, 23-40.
  17. Cameron, E.Z. et al (2008). Maternal investment results in better foal condition through increased play behaviour in horses. Animal Behaviour, 76:5, 1511-1518.
  18. McDonnell, S. M. (2003). The Equid Ethogram: A Practical Field Guide to Horse Behaviour. KY: Eclipse Press.
  19. Tateo, A. et al (2009). Artificially suckled IHDH (Italian Heavy Draught Horse) foals: in vivo performances and ethograms. Italian Journal of Animal Science, 8(sup2), 724-726.
  20. Crowell-Davis, S.L. (1986). Spatial relations between mares and foals of the Welsh pony (Equus caballus). Animal Behaviour, 34:4, 1007-1015.
  21. Haley, D.B. et al (1998). Butting by calves, Bos taurus, and rate of milk flow. Animal Behaviour, 56:6, 1545-1551.
  22. Lent, P.C. (1974). Mother-infant relationships in ungulates. Symposium Reports: The Behaviour of Ungulates and its Relation to Management, 1, 14-55.
  23. Nicol, C.J., & Badnell-Waters, A.J. (2005). Suckling behaviour in domestic foals and the development of abnormal oral behaviour. Animal Behaviour, 70:1, 21-29.
  24. Hausberger, M., & Richard-Yris, A. (2005). Domestic horse, origins, development and stability. In: The Domestic Horse: The Origins, Development and Management of its Behaviour, 33.
  25. Pluháček, J. et al (2013). Laterality of suckling behaviour in three zebra species. Laterality: Asymmetries of Body, Brain and Cognition, 18:3, 349-364.
  26. Lê S, Josse J, Husson F (2008). FactoMineR: A Package for Multivariate Analysis. Journal of Statistical Software, 25:1, 1-18.
  27. Wickham H (2016). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4,
  28. Fox J, Weisberg S (2011). An R Companion to Applied Regression, Second edition. Sage, Thousand Oaks CA.
  29. Oksanen J, et al (2018). vegan: Community Ecology Package_. R package version 2.5-2.
  30. R Core Team (2018). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.
  31. RStudio Team (2018). RStudio: Integrated Development Environment for R. RStudio, Inc., Boston, MA.
  32. Cymbaluk, N.F. et al (1993). Importance of milk replacer intake and composition in rearing orphan foals. The Canadian Veterinary Journal, 34:8, 479.
  33. Oftedal, O.T., & Jenness, R. (1988). Interspecies variation in milk composition among horses, zebras and asses (Perissodactyla: Equidae). Journal of Dairy Research, 55:1, 57-66.
  34. Gantner, V. et al (2015). The overall and fat composition of milk of various species. Mljekarstvo/Dairy,65:4.
  35. Yılmaz, O. et al (2012) The domesticated donkey: I – species characteristics. Canadian Journal of Applied Sciences 4:2, 339-353;
  36. Mellor D.J. (2017) Operational details of the Five Domains Models and its key applications to the assessment and management of animal welfare. Animals 7:8, 60.
  37. Burden, F.A., et al (2011). Hyperlipemia in a population of aged donkeys: description, prevalence, and potential risk factors. Journal of Veterinary Internal Medicine, 25:6, 1420-1425.


This Case Study would not have been possible without the collaboration of many of my colleagues at The Donkey Sanctuary. Thank you to Dr Stuart Norris for running the statistical analysis for the Qualitative Behaviour Assessment and for reproducing my graphs in R. Prof Francoise Welmsfelder should be credited as the developer of Qualitative Behavior Assessment. Thank you to Dr Roger Cutting and Dr Fiona Cooke for reviewing this case study and helping me to prepare it for publication. Much appreciation to Senior Lead in Behaviour and IAABC CHBC Ben Hart for providing expert support and oversight of the foal’s training. Many thanks to veterinarian Dr Elena Barrio for clinical consultation and supervision throughout the implementation of the suckling system. None of this would have been possible without the support and maintenance of the foal and the artificial suckling system by manager Sarah Blair-Salter and her team at the New Arrivals Unit at The Donkey Sanctuary.


This case study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors, but it was supported by the hosting institution, The Donkey Sanctuary, through the provision of the equipment and access to the participant. The author states no conflict of interest. This work has not been published elsewhere, nor has it been accepted for publication elsewhere. It has been approved for publication by all contributors and the hosting institution, The Donkey Sanctuary. Data is available upon request by contacting the corresponding author. This study received ethical approval from Dr Faith Burden, Executive Director of Equine Operations and Dr Fiona Cooke, Head of Research (Europe) at The Donkey Sanctuary.

Corinne is an IAABC Certified Horse Behaviour Consultant with an MSc in Animal Behaviour and an MA (hons) in Psychology. She has her own equine behaviour consultancy, CM Equine, and works full time for The Donkey Sanctuary as an Equine Behaviourist as part of a larger behaviour team. She has extensive practical experience with donkeys in a shelter setting, seeing donkeys through their journey from their initial relinquishment to potential rehoming on The Donkey Sanctuary’s Guardianship Scheme.

Contact info

If you have any questions about donkey behaviour, please contact The Donkey Sanctuary’s behaviour team at equine.behaviour@thedonkeysanctuary.org.uk.

If you have any questions about methods of data collection or analysis in this case study, or on the artificial suckling system itself please contact corinne.mccafferty@thedonkeysanctuary.org.uk.

For Corinne’s private equine consulting services please contact info@cmequine.co.uk or visit https://www.cmequine.co.uk/.

TO CITE: McCafferty, C. (2022) A novel method for rearing donkey foals with behavior in mind. The IAABC Foundation Journal 24, doi: 10.55736/iaabcfj24.4