For clarity, within this article, ‘lifting’ refers to lifting of the rabbit from the floor or hutch in the hands of a person, ‘holding’ refers to the restraint of the rabbit off the floor in the arms of a person and ‘carrying’ refers to both ‘lifting’ or ‘holding’. The word ‘handling’ has previously been used indiscriminately. In this article, ‘handling’ is reserved for non‐specific tactile interaction unless in a direct quotation from previous studies.

Many rabbits show fear behaviours when lifted from the ground. Estimates from owner surveys suggest that between 57 (Schepers et al . 2009 ) and 61% (Rooney et al . 2014 ) of pet rabbits struggle if lifted and fear‐related aggression is common: Normando & Gelli ( 2011 ) showed in 1 study that 24% of pet rabbits had bitten the owner at least once. This constitutes a significant behavioural and welfare problem and needs to be addressed.

There has been increasing interest in the welfare of the UK's 1.3 million pet rabbits (PDSA 2015 ). Previous work towards improving rabbit welfare has focussed on the importance of a good diet (Prebble & Meredith 2014 ), exercise (Dixon et al . 2010 ) and companionship (PDSA 2011 ), but the welfare implications of lifting and holding this nervous prey species have been less discussed. Of these 4 areas, it is lifting and holding that is most likely to damage the human–animal bond. This can worsen the situation further because a strong bond is a powerful motivator to improve pet welfare (Wensley 2008 ).

EFFECTS OF CARRYING RABBITS

There have been various surveys of pet rabbit behaviour and welfare. Mullan & Main (2007) explored the ability of owners to lift their rabbits – comparing the owner's assessment of their own level of confidence with the behaviour of their rabbits. Unsurprisingly, they found that owners who were confident in lifting their rabbits were more likely to do so. But they also found that rabbits belonging to owners who self‐labelled as ‘unconfident’ were more likely to willingly approach. In addition, rabbits that were housed singly were less likely to struggle while being lifted (this has been previously noted in laboratory rabbits (Podberscek et al. 1991) and attributed to the need for some degree of social contact in a socially deprived individual; however, this trend was not observed by Schepers et al. (2009)). Mullan and Main hypothesise that the less‐lifted rabbits might have been more likely to approach an observer as they had less reason to fear interaction. It is surprising, then, that the authors go on to conclude that ‘the bond between them and therefore the welfare of the rabbit will be strengthened by confident handling’.

Schepers et al. (2009) discussed lifting rabbits in the context of reducing the induction of fear by humans and management procedures. A survey of rabbit owners (n=899) suggested that 57% of rabbits struggled at least slightly when being lifted. The data from a direct experimental sample (n=66) demonstrated that 49% of rabbits struggled when being lifted. Again, the authors observed that singly housed rabbits spent longer sniffing and touching the observer's hand and suggested that this was due to their social deprivation.

Rooney et al. (2014) reported that 61% of rabbits showed signs of fear when ‘handled’ by the owner, rising to 75% when ‘handled’ by an unfamiliar adult. A weakness in this study was the lack of specificity in the term ‘handling’ – this was assumed to mean lifting but might also have included other interactions on the ground including stroking and play. The authors discuss the need for differentiation in future studies between ‘picking up’ and ‘handling’ – the latter being conducted at floor level and less stressful for the rabbit. They raise concerns that the majority of rabbits are lifted at least weekly, which could potentially represent a significant source of stress, causing further sensitisation of rabbits to being lifted.

The above studies are attempts to quantify behaviour patterns in the pet rabbit population. In addition, there have been several papers published on the incidence and treatment of behavioural problems in rabbits. Normando & Gelli (2011) found that 24% of rabbits had bitten the owner at least once (from a survey of owners), while only 7% were said to be ‘aggressive towards the owner’. Yet despite this incidence of aggression, rabbits are rarely relinquished to rescue centres explicitly because of behavioural problems (Cook & Mccobb 2012, Ulfsdotter 2013). This suggests that their behavioural problems are seen to be less problematic than those of larger pet species (possibly because of close confinement, low risk of serious injury to humans or low expectations of rabbits as pets), therefore less attention is paid when they are showing signs of fear and distress. However, in this study, behavioural problems may be masked in the data: relinquishment because ‘the owner has lost interest in the rabbit’ may cover a range of behavioural problems, as inactivity and lack of play behaviours make for an uninteresting pet.

Magnus (2005) describes fear aggression related to being lifted as one of the most common behavioural problems seen in rabbits, a product of repeated aversive lifting experiences resulting in escalation of defence mechanisms. Crowell‐Davis (2007) discusses desensitisation and counter‐conditioning for fear aggressive rabbits in the context of being lifted. She states that ‘from the rabbit's point of view, being picked up by a giant human can be very frightening. Multiple experiences of being picked up and carried around securely and painlessly are essential for a rabbit to become comfortable with human handling’ (p. 41). This approach fails to acknowledge that infliction of such fear is not in the best interests of the animal, and that frequent lifting can instead cause sensitisation and decreased willingness to interact with the owners. Although positive reinforcement can desensitise rabbits to being lifted to a certain extent (Bradbury 2013), it is still an aversive procedure (Buseth & Saunders 2015) – whereas floor‐based interactions, especially those involving head contact, are reinforcing in their own right (Swennes et al. 2011). Gentle stroking by a familiar person to reduce fear and stress may help, but there is no evidence that interaction with a person unfamiliar to the rabbit will have this effect.

Lifting rabbits is demonstrably stressful; lifting them and inducing tonic immobility is even more stressful. Tonic immobility in rabbits is a state of motor inhibition induced by inverting the rabbit onto its back. It is thought to have evolved as an antipredation mechanism – movement being the primary motivator for attack by canids (Fox 1969). McBride (2011) reports that some pet owners believe that tonic immobility (‘trancing’ or ‘hypnotising’) is caused by relaxation rather than fear, and popular internet images and memes reinforce this belief (Mati 2013). In addition, the action is still recommended in veterinary journals for procedures such as nail clipping (Malley 2007), which may be performed by owners. McBride et al. (2006) tested an admittedly small sample, but nevertheless found significant elevations in respiration, heart rate and plasma corticosterone after induction of tonic immobility. They also found behavioural changes indicative of fear during induction and emotional responses after the incident, suggesting that tonic immobility is perceived as unpleasant. Oxley & Ellis (2015) recommend that more research is done on the reasons for tonic immobility use by owners. Varga (2013) suggested that it was admissible for veterinary surgeons to continue to use tonic immobility for procedures such as radiography as it poses less risk than anaesthesia, but that it should only be used when necessary. There are alternative methods that do not rely on tonic immobility (House Rabbit Network Writers 2007).

If veterinary surgeons still choose to use tonic immobility for such procedures, it is important that they are aware of factors that may decrease the likelihood of inducing tonic immobility. Illustrations in articles depicting the state often show rabbits in tonic immobility without any human contact (Malley 2007) – this should not be encouraged and is not commonly achieved. It is harder to induce tonic immobility if the rabbit is near the home cage (Ewell et al. 1981) or if it is accustomed to being ‘handled’ (Verwer et al. 2009), whereas rabbits with high corticosterone levels are more likely to show tonic immobility (Carli et al. 1979). Encouraging more humane ways of keeping rabbits may lead to a population where tonic immobility is simply less effective. Therefore, educating owners to avoid tonic immobility alongside education to improve welfare is essential.