Article Text


Assessing reptile welfare using behavioural criteria
  1. Clifford Warwick,
  2. Phillip Arena,
  3. Samantha Lindley,
  4. Mike Jessop and
  5. Catrina Steedman


While clinical reptile medicine as a science is in its ascendancy among veterinary surgeons and other interested groups, familiarity with the often related issue of reptilian behavioural and psychological health appears less common. Behavioural change in reptiles, as in other animals, is often the primary indicator of disturbance, injury or disease. Just as a behavioural sign may be an indicator of stress or a physical problem, a physical sign may be an indicator of a behavioural problem, and abnormal behaviour may result in injury and disease. This article focuses on abnormal behaviour in reptiles, including signs of captivity-stress, injury and disease and their aetiologies, and takes a fresh look at some old and established biological and husbandry problems. Concise diagnostic guidance on behaviour issues is also included. The article might serve to prompt questions that may be asked of reptile keepers when evaluating animal and husbandry background.

Statistics from

Clifford Warwick graduated in biology in 1990 and has undertaken research on reptiles, captive-reptile behaviour and biological strategies. In 2004, he obtained a primary health care qualification at Leeds University Medical School, specialising in zoonotic disease. He now works as a private consultant.

Phillip Arena graduated in biology in 1986 having majored in reptile ecology. He was awarded a PhD by Murdoch University, Australia, in 1991 for his work on reptile and amphibian anatomy and physiology. He is a lecturer in academic skills, biology and veterinary studies at Murdoch.

Samantha Lindley qualified in veterinary science at the University of Bristol in 1988 and completed a large animal internship at Glasgow Veterinary School in 1989. She then embarked on a career in large animal practice, before focusing her attention on animal therapy, behaviour, acupuncture, chronic pain management and the welfare of captive wild animals.

Mike Jessop qualified from the Royal Veterinary College in 1984 and is a partner in a small animal practice in South Wales. He developed a keen interest in exotic animal medicine and surgery and has a particular interest in chelonians.

Catrina Steedman graduated in psychology in 1987. She has undertaken research on the ecological effects of human harvesting of freshwater terrapin populations in North America, crocodilian slaughter methods in ranching operations, and reptile behavioural problems in zoos. She currently works for the Emergent Disease Foundation, a charity which aims to raise awareness of zoonoses.

Biological and behavioural considerations

Significant differences exist between domesticated companion animal species such as dogs and cats and exotic non-domesticated pet species, which include all reptiles. Biologically, dogs and cats, along with other commonly domesticated animals including cattle and horses, possess essentially genetically ‘pre-adapted’ and ‘soft-wired’ traits allowing them to coexist (‘life-share’) with other species and the captive context. By contrast, reptiles possess few pre-adaptive features and are hard-wired with innate biological, behavioural and psychological needs that preset them to life in nature.

An inescapable factor that dramatically and negatively impacts on the biological suitability of reptiles to captivity is that, unlike dogs and cats, reptiles will almost universally be ‘life-restricted’ in small, arbitrarily and poorly conceived vivariums maintained by non-professionals. These major biological considerations and management deficiencies imbue the veterinary surgeon with onerous duties to look for overt and emerging, as well as occult, health (physical, behavioural, psychological and husbandry) issues associated with reptiles that are presented to them, irrespective of the reason for consultation.

Behavioural assessment as a tool

Contrary to common perceptions, reptiles manifest an array of abnormal behaviours that indicate stress.

Behavioural assessment of animals (including reptiles) is an essential method for evaluating their condition and welfare. While physiological stress measurements are available through blood and, less invasively, faecal sampling, numerous confounding factors are involved with this method, including a dearth of pure baseline data and the limitations of focused interpretation. For example, human studies report that, while cortisol may be mediated by certain factors such as agitation-related stimuli, states including perceived stress, anxiety and depression may not increase cortisol (van Eck and others 1996). Extrapolation from these findings suggests that conditions such as understimulation in animals may not be revealed through physiological measurement (eg, corticosterone in reptiles).

Normal and abnormal behaviour in context

Reptile behavioural diversity approaches, parallels and sometimes surpasses that seen in birds and mammals (Gillingham 2004). Normal behaviour implies not only natural behaviours, but also their appropriate range and context. For example, in nature, it may be normal and healthy for an animal to spend hours of exploratory locomotor activity in order to hunt for food, whereas it may be abnormal and unhealthy for an animal to spend less than an hour pacing a small enclosure in a captive scenario with plentiful food.

There is a shortage of field-based behavioural observations for the vast range of species held in captivity and this deprives captive animal observers of good comparative information. Relatedly, the inherent possible conundrum exists that some animals that may be stressed by field observations may change their habits. For example, eye contact alone between observers and free-living iguanas resulted in significant disruption of normal hierarchical perching (Burger and others 1991). Accordingly, stress-related states in captive animals in some cases could be underrated as ‘normal’ because they have been mistakenly based on human-invoked responses in the field.

Of course, stressors are present in nature and occur commonly. However, in nature, animals (regardless of their eventual success) are adapted to cope with specific evolved challenges in an environment that has an overall balance and suitability.

By contrast, captive conditions typically replace many features of the natural world with artificial and frequently poorly matched alternatives that deprive animals of known normal behaviour and associated biological needs, such as hunting, spatial range, and macro-habitat investigation (Arena and Warwick 2004, Warwick 2004).

Misinterpreting common biological and behavioural signs

Reptile sellers, keepers and some professionals commonly interpret signs of ‘good feeding’, ‘good bodyweight’ and ‘active reproduction’ as positive indicators of welfare. However, in isolation, these signs are poor indicators of welfare and may be highly misleading (Broom and Johnson 1993, Warwick 2004). Indeed, the presence of some genuinely ‘positive’ signs may not convey good welfare where any concurrent negative health or welfare sign is identified.

Signs of abnormal behaviour and captivity-stress

Generally, captivity-related chronic stress behaviour may result in increased abnormal behaviour, behavioural inhibition, vigilance behaviour, hiding, fearfulness and frequency of startle, aggression, and freezing behaviour, and decreased exploratory behaviour, reproductive behaviour, and behavioural complexity, as summarised by Morgan and Tromborg (2007).

However, these criteria vary according to animal class and species. Table 1 presents behavioural signs of captivity-stress in reptiles and Table 2 presents sample signs of normal behaviour, quiescence and ‘comfort’. For these summaries, the authors have adopted established behavioural assessments as summarised by Arena and Warwick (2004), Warwick (2004) and Warwick and others (2011a).

Table 1:

Behavioural signs of captivity-stress

Table 2:

Behavioural signs of quiescence and ‘comfort’

As with other assessments, many problematic behavioural issues may be ‘masked’ by acute arousal states modified by a reptile's temporary presence and examination in the clinical environment, which confound or preclude accurate diagnosis.

Signs of normal behaviour, quiescence and comfort

Although totally normal reptile behaviour in the acute clinical environment is unlikely, Table 2 outlines some key signs that may assist in gauging animal condition. The presence of normal signs does not imply overall good health state where any abnormal sign is also observed. Relatedly, some behavioural signs may be ambiguous, for example, sleep may be associated with normal or abnormal behaviour or health.

Physical signs of behavioural problems

A physical sign (such as an injury or topical infection) may be an indicator of an environment-related behavioural problem. For example:

Rostral lesions may be associated with interaction with transparent boundaries (ITB). Fig 1 shows an example of ITB, which is a common and often persistent (up to 100 per cent activity period) captivity-stress-related behavioural problem involving attempts to push against, crawl up, dig under or round the transparent ‘invisible’ barriers (typically glass) of enclosures; Fig 2 shows friction lesions resulting from this abnormal behaviour;

Fig 1

Interaction with transparent boundaries (ITB). ITB is a captivity stress-related behavioural problem (Picture: Pro-Wildlife)

Fig 2

Interaction with transparent boundaries (ITB)-associated chronic friction lesion; rostral lesion on maxilla and mandible. (Picture copyright:

Ventro-mandibular lesions may be associated with arboreal species diving and/or dropping onto hard surfaces;

Thermal ‘contact’ burns (typically on the dorsal region) may be associated with deficient overall temperatures, thermal gradation or basking facilities;

Head and extremity injuries and infections may be associated with co-occupant aggression or courtship behaviour in overly restrictive environments with insufficient escape zones. Fig 3 shows a maxillary abscess in a lizard caused by a co-occupant bite, and Fig 4 shows a limb bite abscess in a terrapin, also from co-occupant aggression;

Fig 3

Maxillary abscess in a lizard caused by co-occupant bite (Picture: Phillip Arena)

Fig 4

Limb bite abscess in a terrapin. Co-occupant aggression or courtship harassment can lead to bite-related infections (Picture: Clifford Warwick)

Ventral dermatoses may be associated with hypoactivity behaviour in overly restrictive environments;

Intestinal impactions may be associated with pica (substrate-eating) in under-stimulating environments. Fig 5 shows a radiograph of a severe ‘pica’ (lithophagy)-related gastrointestinal impaction in a tortoise. Fig 6 shows a sample of evacuated cage gravel from the same animal.

Fig 5

Radiograph showing pica (lithophagy)-related impaction in a tortoise. One cause is understimulating environments leading to this habit-forming behaviour (Picture: Emma Nicholas)

Fig 6

Sample evacuated stones/gravel from pica-related impaction (Picture: Clifford Warwick)


The components of artificial environments are notoriously complex, involving issues that include, but are not limited to, space, temperature, humidity, light, airflow and furnishings – each with its own array of known (and often more importantly unknown) key factors. Winning the National Lottery involves getting six correct variables out of 49 to beat a 1 in 14.5 million chance. In husbandry, the chances of six correct variables out of thousands of dynamic variables are very small and this may partly explain why many reptiles do not survive long in captivity.

Although this article is primarily about behaviour rather than general husbandry, it will include two physical environmental issues with close bio-behavioural associations – spatial and thermal considerations – and examine more closely particular issues within each subject.

Spatial considerations

Many reptiles are mistakenly and inhumanely kept in small cages due to erroneous advice handed down from one pet trader, hobbyist or ill-informed keeper to another. Common false understandings are that many reptiles ‘feel safer’ in small environments and that they are naturally ‘sedentary and don’t need space'. This rationale may suit the convenience of those seeking to promote reptiles as a ‘cage pet’, but it is scientifically and ethically wrong.

While reptiles, like other animals, require shelter to which they can voluntarily withdraw, the key elements are that the animal seeks a ‘hiding place’ when it senses the need for it and it does this voluntarily. Imposing a confined space on an animal is biologically equivalent to trapping it.

Home range studies of reptiles have frequently shown them to be highly active and that they travel either within local ranges of several hundreds of square meters or indefinite ranges measured in hundreds or thousands of kilometres. For example, arboreal monitors have been documented moving daily ranges greater than 186 m, home ranges for some skink lizards are 1 ha, box turtles 40 ha, indigo snakes 158 ha, and for sea turtles, travel can be measured in the thousands of kilometres.

Small species and juveniles commonly utilise as much, and sometimes more, total space than large species and adults. Smaller forms are often insectivorous and these may need to feed more frequently than larger forms and also require a great deal of activity to track and catch their highly active prey.

Regardless of these differences, all reptiles are active, including species such as pythons that are popularly, but wrongly, perceived as sedentary. Some species, in particular large carnivores such as monitor lizards and pythons, may adopt brief sedentary periods following consumption of large meals, but this is a transient phase and not one that should be used to judge an animal's general activity pattern or spatial needs.

Postural-positional orientation

Not only is significant free space essential for normal behaviour and welfare, but the spatial landscape is also important to accommodate an animal's postural-positional orientation.

All reptiles appear to seek out and occupy ‘angles’ and orientations (sometimes highly subtle). These postural-positional desires appear to play roles in delivering comfort and focused thermal needs, as well as in the amelioration of discomfort. Accordingly, a snake that needs to adopt a straight-line posture to relieve intestinal discomfort cannot do so in a cage that is shorter than the snake itself. Fig 7 presents an example of severe spatial restriction where the snake cannot voluntarily adopt a straight-line body posture.

Fig 7

Severe spatial restriction. Snakes need to adopt straight-line postures (Picture: Animal Protection Agency)

Therefore, merely providing a branch to bask on or a hide to retreat to is inadequate.

Good cage, bad cage

It may be unwise to attempt to set or promote ‘appropriate’ or ‘adequate’ spatial needs and cage sizes, as in the restrictive captive setting this arguably equates to ascribing a positive connotation to what is in effect a negative situation. Even well considered cage-size recommendations effectively amount to an acceptable safe minimum.

Overcrowding and crypto-overcrowding

Overcrowding is manifested in two ways, ‘overt overcrowding’ and ‘covert (or crypto) overcrowding’. Excessive numbers of animals occupying a certain amount of space determines overt overcrowding. Crypto-overcrowding essentially refers to the availability of all facilities to all animals when they require access to those facilities.

Accordingly, an enclosure that appears large and abundant but that lacks the ability to ‘service’ all animals' needs at any time is capable of being overcrowded by way of other inherent deficiencies. Therefore, in order that a space is not overcrowded, it must allow both space to roam as well as possess sufficient facilities – for example, a water bowl or basking site – that all animals can use together at any one time (Warwick and others 2011b).

Fig 8 presents an example of severe overt overcrowding where the baby turtles depicted obviously cannot occupy land space or utilise basking heat facilities. Fig 9 shows crypto-overcrowding where the baby turtles can occupy limited land space, but cannot all utilise the basking heat facility at one time.

Fig 8

Severe overt overcrowding. Obvious overcrowding of all facilities (Picture: PETA)

Fig 9

Crypto-overcrowding. All turtles cannot utilise all of the facilities (eg, basking) at any one time (Picture: Mike Jessop)

Thermoregulatory issues

Reptiles naturally use behaviour to select and occupy niche temperatures and attain precise thermal conditions. Inability to thermoregulate within precise, self-perceived (by the animal) needs and even with regard to a single event may result in the exacerbation of acute stress as well as chronic debilitation.

Behavioural fever and stress

In reptiles, the fever response is primarily behavioural rather than physiological. ‘Behavioural fever’ is manifested by the compromised animal directly seeking higher than usual temperatures by selecting warmer zones.

Similarly, healthy but ‘stressed’ reptiles, such as those that react poorly to handling or to intraspecies competition, may show ‘emotional fever’ and seek out higher temperatures until they ‘settle’. It is probably important that they can seek raised temperatures after even minor stressor events. Handling may equate to capture and predation.

Voluntary hypothermia

In addition to voluntary hyperthermia, there is also voluntary hypothermia, a state where some injured or diseased animals appear deliberately to seek out lower optimum or very low temperatures. Clearly, climatic factors may stimulate hypothermia such as animals that seek to hibernate following the onset of a naturally cooler environment. Where injured or diseased reptiles are concerned, it may be helpful to take into consideration that voluntary hypothermia may be a requirement for some animals to survive a problem, most likely because reduced microbial growth, physiological disturbance and ‘quiescence’ and healing may in some cases result from a compromised animal's ‘biological shutdown’. However, the strategy should be regarded cautiously in sick individuals and should not always be presumed or accepted as positive. Identifying this state is difficult, but may be a useful differential for the clinician.

Appropriate thermal gradients are essential for health maintenance. The need to alter body temperature correctly may be determined by highly subtle physiological cues only perceptible by the individual animal, rather than by a human ‘guesstimate’. Although it is common practice to raise the body temperature of injured or sick reptiles artificially, in the authors' view this intervention requires caution; while it may promote healing in some cases, rapid microbial overgrowth and toxaemic effects may be triggered – potentially raising catastrophic physiological demands in an unprepared and compensated animal. Furthermore, a compromised individual may not physically be able to escape hyperthermia once it has reached a preferred body temperature.


Informed behavioural assessment of reptiles offers an underused window into health and welfare that can reveal problems as clearly as overt clinical signs such as lameness. Assessing reptile behaviour requires progressive familiarity with the subject and species. However, the diagnostic and health state rewards are often surprisingly positive, not least because behaviour frequently constitutes a messenger not only of immediate and transient issues, but also as an enduring holistic conveyer of all that is ‘self’.


Further reading

  1. 10.
  2. 11.
View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.