The effects of enhancing cage complexity on the behaviour and welfare of laboratory rats

Behavioural Processes 85 (2010) 172–180

Contents lists available at ScienceDirect

Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc

The effects of enhancing cage complexity on the behaviour and welfare of laboratory rats Usama A. Abou-Ismail a,b,∗ , Oliver H.P. Burman a , Christine J. Nicol a , Michael Mendl a a b

Centre for Behavioural Biology, University of Bristol, Langford House, Langford, Bristol BS40 5DU, UK Department of Animal Husbandry, Faculty of Veterinary Medicine, Mansoura University, Gomhoria St., Mansoura, P.O. box 35516, Egypt

a r t i c l e

i n f o

Article history: Received 18 January 2010 Received in revised form 20 May 2010 Accepted 5 July 2010 Keywords: Behaviour Enrichment Laboratory rat Welfare

a b s t r a c t This experiment was carried out to investigate the long-term effects of enhancing cage complexity on behavioural measures of welfare in laboratory rats. We housed 72 rats in groups of four in either ‘enriched’ or ‘unenriched’ cages for six weeks. Scan and focal animal sampling were conducted in both the light and dark phase of the second, fourth and sixth weeks. Results revealed that rats in the ‘enriched’ cages showed longer durations of sleep behaviour, and low levels of agonistic behaviour compared to rats in the ‘unenriched’ cages. Results importantly demonstrated that the behavioural changes observed in the enriched environment were due to the presence of the enrichments themselves in the cages (indirect effects) and not due merely to rats interacting with the enrichment items in their environment. Thus, enhancing the complexity of conventional laboratory cages can promote behaviour such as longer bouts of sleep that is likely to be indicative of good welfare, and diminish levels of behaviour such as aggression that is likely to lead to poor welfare. © 2010 Elsevier B.V. All rights reserved.

1. Introduction It is important that the housing environment of laboratory rodents should be studied and improved where possible, not only because laboratory rodents constitute a large proportion of the animals used in scientific studies (Home Office, 2002), but also to ensure that good quality and reliable scientific data are obtained (Sherwin, 2004; Würbel, 2001). One approach to improving the housing environment that has been the focus for a large number of studies, is environmental enrichment. Ideally, the term ‘environmental enrichment’ might be reserved for modifications to the housing environment that result in significant welfare improvement (Newberry, 1995). However, the term is often used to denote ‘modification of the environment’ without knowledge of the effects of such modification. In this paper, we use the term in this way as a short-hand label for the experimental treatments that researchers have imposed. ‘Enrichment’ may also involve the social and/or physical environment and, where relevant, we distinguish between these. Experiments on laboratory rodents have demonstrated wide effects of environmental modifications, including physiological (Belz et al., 2003; Chamove, 1989; Roy et al., 2001), psychologi-

∗ Corresponding author at: Animal Behaviour and Welfare Group, Department of Animal Husbandry, Faculty of Veterinary Medicine, Mansoura University, Mansoura, Egypt. Tel.: +20 162557996; fax: +20 502379952. E-mail address: [email protected] (U.A. Abou-Ismail). 0376-6357/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.beproc.2010.07.002

cal (Chamove, 1989; Patterson-Kane et al., 1999), developmental (Davenport et al., 1976), therapeutic (Hockly et al., 2002; Passineau et al., 2001), and behavioural effects (Armstrong et al., 1998; Chamove, 1989; Orok-Edem and Key, 1994; Van Loo et al., 2002). Furthermore, and importantly, it has been recently shown that beneficial environmental enrichment effects in improving laboratory rodent welfare may be achieved without reducing the precision and reproducibility of the experimental data derived from the animals (Wolfer et al., 2004). Reduced external validity of the research has also been shown to arise when laboratory rodents are housed in standard laboratory cages (Sherwin, 2004). In addition to the physiological, psychological, developmental and therapeutic advantages that housing in enriched conditions may provide, research has elucidated behavioural benefits from being housed in enriched environments. An important potential effect of environmental enrichment is on agonistic behaviour. Damaging social behaviour between conspecifics, such as excessive agonistic behaviour, is a common problem related to housing male laboratory rodents in captivity (e.g. Hurst et al., 1999; Van Loo et al., 2002), leading to physical damage and associated social stress and poor welfare (e.g. Hurst et al., 1996, 1999). The addition of environmental enrichment can, sometimes, reduce excessive aggression between animals kept under standard unenriched housing conditions. Orok-Edem and Key (1994) demonstrated that provision of some chewable items to the cages of laboratory rats such as gnawing blocks and wooden tongue depressors resulted in a dramatic reduction in the frequency of fighting. Similarly, in laboratory mice, Chamove (1989) indicated

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

that newly weaned laboratory mice reared in groups in complex cages were less aggressive relative to their littermates reared in groups in standard laboratory cages. Supplementation of cornhusk nesting materials decreased aggressive behaviours between group-housed mice during the first four days after grouping as compared to those housed in groups in standard laboratory cages (Armstrong et al., 1998). Prolonged group housing of male laboratory mice in enriched cages (nesting materials) resulted in a reduction of inter-male aggression (Van Loo et al., 2002). However, in contrast to these findings there are data that also reported an increase in agonistic behaviours between animals (particularly mice) housed in groups in “enriched” housing conditions (Haemisch and Gartner, 1997; Haemisch et al., 1994; McGregor and Ayling, 1999; Nevison et al., 1999; Van Loo et al., 2002), either due to the encouragement of territorial behaviours or due to exposure of vulnerable body parts such as tails to biting wounds. Another possible indicator of reduced welfare in confinementreared animals is the expression of abnormal behaviours such as stereotypies (Fraser and Broom, 1997; Mench and Mason, 1997). Research has shown that environmental enrichment has achieved some success in reducing abnormal behaviour. Laboratory and wild rodents kept in enriched housed conditions display less stereotypic behaviour as compared to those housed in standard laboratory cages (Callard et al., 2000; Powell et al., 2000; Turner et al., 2003; Würbel et al., 1998). However, there are also data that suggest that the provision of environmental enrichment has no effect on the expression of stereotypic behaviours (e.g. Line and Morgan, 1991; Spring et al., 1997). Although the evidence generally favours the conclusion that environmental enrichment can minimise both aggression and abnormal behaviour in laboratory rodents, it is clear that not all studies support this conclusion. Aspects of experimental design, may partly explain some of these contradictions. For example, in some environmental enrichment studies, the effects of environmental enrichment were studied using singly housed animals in enriched versus non-enriched environments (e.g. Belz et al., 2003; Denny, 1975; Townsend, 1997; Watson, 1993), despite the fact that single housing has been shown to be stressful (e.g. Gamallo et al., 1986). Group size has often been confounded with physical enrichment of the environment making it difficult to determine whether the social or physical environment is responsible for any effects observed. Often, animals have been group-housed under physically enriched conditions and compared to singly housed conspecifics in unenriched cages (e.g. Davenport et al., 1976; Greenough et al., 1978; Rosenzweig et al., 1969; Spangenberg et al., 2005), with these latter animals sometimes being housed in complete visual and tactile isolation from others (e.g. Fiala et al., 1977; Tagney, 1973). Therefore, as positive effects of enrichment are observed, in some studies, where physical enrichment is confounded with group size, more work is needed to investigate whether enrichment in its own is actually effective. Cage-size is another frequently encountered confounding variable, with enriched groups typically provided with not just more cage furniture, but also more space (e.g. Bennett et al., 1969; Black et al., 1989; Henderson, 1970; Huck and Price, 1975; Kiyono et al., 1981; Mirmiran et al., 1982; Moncek et al., 2004; Patterson-Kane et al., 1999; Van Gool and Mirmiran, 1986; Volkmar and Greenough, 1972). It is therefore hard to determine whether findings are the result of supplying standard laboratory cages with additional items, and/or a consequence of the additional floor space and increased cage height. Again, as positive effects of enrichment are observed, in some studies, where physical enrichment is confounded with cage size, more work is needed to investigate whether enrichment in its own is actually effective.

173

A variety of behavioural measures have been taken in previous studies of environmental enrichment in laboratory rodents, but very few studies have investigated its effects on sleep behaviour. Sleep behaviour has been shown to be a potentially useful noninvasive tool for assessing welfare in laboratory rats (Abou-Ismail et al., 2007). Abou-Ismail et al. (2007) found that high sleep frequency and duration correlated with increased bodyweight and weight gain, and decreased adrenal weight suggesting that they may indicate good welfare. It would therefore be interesting to investigate the effects of enrichment on simply recorded behavioural measures of sleep. In addition, few studies have investigated the effects of environmental enrichment on the welfare of laboratory rodents over the longer term. It is possible that the animals habituate to the presence of a physical enrichment object, or that the enrichment produces a long-term change in behaviour, and this important question should receive further investigation. This experiment was therefore carried out to provide a detailed picture of the effects of enhancing cage complexity, by supplying commonly used additional cage structures while controlling for both physical and social aspects of the environment, on behavioural measures of laboratory rat welfare. We also attempted to take into consideration whether any behavioural changes as a consequence of the enrichment were due to animals redirecting their behaviours towards the enrichment objects, or due to other reasons such as the way animals perceive their environment when it is changed. 2. Methods 2.1. General animal housing and husbandry This experiment was carried out using three batches of rats. Within each batch there were two replicates of the two experimental treatments used (see later). Each batch, comprised twenty-four outbred male Wistar rats (Harlan, UK). The rats were six weeks of age on arrival, and were, immediately, arbitrarily allocated in four large metal cages (70 cm length × 52 cm width × 36 cm height). Sawdust was provided as a bedding material. These cages had different dimensions to those in which the rats were housed during the study and allowed time, after transportation, for rats to settle in to their new environment and adjust to the lighting regime without experiencing the conditions they would be housed in during the study. Food and water were provided ad-libitum and were checked daily. Rats were maintained under a 12:12 h light:dark schedule, with white light on between 1200 and 2400 and dim red lighting (two 60 W bulbs) to facilitate dark phase observation, at a temperature of 20 ± 2 ◦ C. Once a week, all rats were removed from their cages and rehoused in clean cages with new bedding material. The fur of each rat was marked with hair dye (Nice ‘n’ Easy Natural Black 122, Bristol Myers Ltd., Uxbridge, UK) in one of four different patterns on the day of arrival. These marks were refreshed after three weeks, allowing sufficient time after dyeing before behavioural observation to reduce any possible effects on behaviour of the dyeing process (e.g. Hurst et al., 1999). Tails were also marked, in one of four distinguishable manners, with a permanent marker pen (N50, Pentel Pen, Pentel Co, LTD., Japan), to provide an additional means of identification. Tail marks were renewed every week. 2.2. Housing conditions Rats were arbitrarily allocated to one of the following two conditions for six consecutive weeks: 1) “Standard”: polypropylene cages (48.5 cm length × 33 cm width × 21 cm height) without any additional cages structures.

174

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

Table 1 Description of the additional cage structures used to enhance the cage complexity. Structure

Number supplied

Provision and renewal

Shredded paper Nestlets Aspen wood block Rodent retreat (shelter)

Handful 4 1 1

Percher

1

Crawl ball

1

Throughout the experiment, renewed every week. Throughout the experiment, renewed every week. Throughout the experiment, renewed every week. Throughout the experiment, re-introduced uncleaned to the cage. Throughout the experiment, re-introduced uncleaned to the cage. For one week only throughout the experiment.

Rope Nylabone Ladder

1 1 1

For one week only throughout the experiment. For one week only throughout the experiment. For one week only throughout the experiment.

Wood ball

1

For one week only throughout the experiment.

2) “Complex”: standard cages that were supplied with a number of additional cage structures. In this study, different objects were used to enhance the complexity of the standard laboratory rat cages such as gnawing objects (aspen wood blocks, wood balls and nylabones), shelter (rodent retreat), devices for climbing (ladders and ropes) and other objects to stimulate general activity and gnawing such as crawl balls and wood balls. The description of these additional cage structures and their frequency of provision is presented in Table 1. In addition to enhancing cage complexity we also provided increased foraging opportunities. This was done by mixing four intact food pellets, and four pellets that had been broken into numerous small pieces with the bedding material after the cages were cleaned every week. Foraging devices have been shown to be an effective way of enriching laboratory rat cages (e.g. Johnson et al., 2004). Some of these cage structures were introduced to the cage from the first week and remained throughout the 6-week experiment, such as the shredded paper, nestlets, aspen wood block, percher and the rodent retreat. Other structures were supplied as ‘novel’ cage structures; the item was introduced to the cage for one week only starting from the second week, and was replaced by another different novel item every week after cage cleaning such as the crawl ball, rope, nylabone, ladder and wood ball. The order in which these ‘novel’ items were introduced to the cage was counterbalanced between replicates to control for any possible effect of age upon reaction to the objects. 2.3. Behavioural observation All behavioural observations were carried out by the same experienced researcher (UA-I) throughout the study. The observer entered the experimental room at least five minutes before the scheduled start time of the observation to allow the rats to habituate to his presence (e.g. Hurst et al., 1999). 2.4. Scan and focal sampling sessions Observation was carried out in two sessions per day for each of the four experimental groups (two ‘standard’ and two ‘complex’) in each batch. The first session took place during the light phase (white light was on); starting at 1000 h and ending at 1045 h. The second session was carried out while the white light was off (during the dark phase of the day); starting at 1200 h and ending at 1245 h. Each experimental group was observed for one day per week for three weeks throughout the experiment, every other week starting from the second week of the experiment (i.e. week two, four and six). The day of observation was counterbalanced between different

Description 5 cm × 5 cm sterilized cotton fibre pads, Lillico. 34 mm × 70 mm long aspen wood block, Lillico. 20.5 cm L × 15.7 cm W × 11.5 cm H Guinea pig huts, red-tinted, Lillico. Pure cotton rope and natural wood, Percher Spider, 4260, PetLove. 115 mm, with 3 mm × 58 mm holes, red-tinted polycarbonate, Lillico. 60 cm length of 1.5 cm diameter hemp rope. Regular size, original flavour, (36 g), Nylabone. 9 step wooden ladder 35.5 cm, Premium Bird Toys, Platinum Pet Products. Roll ‘N’ chew (small), item #62067, natural wood 9 cm diameter, Super Pet.

treatment groups and replicates to control for possible effects of observation day or age on behaviour. Two methods of behaviour sampling were used throughout the experiment. Scan sampling and focal sampling (Martin and Bateson, 1993). 2.4.1. Scan sampling The whole group was scanned once every eight minutes throughout each observation session to record specific behaviour patterns such as spatial distribution within the cage, and the use of the additional cage structures by the rats (for the ‘complex’ cage groups only). Each session lasted 32 min (the length of the focal sampling session, thus yielding four scans per session, a total of eight scans per group per day (observation week), and a total of 24 scans per group per replicate. Behaviour patterns that were collected by scan sampling are listed in Table 2. Behaviours were manually recorded onto check sheets. 2.4.2. Focal sampling Focal sampling was used to record the frequencies or durations, or both, of all elements of laboratory rat behaviour (time Table 2 Behavioural elements recorded by scan sampling. Behaviour pattern

Definition

Active

The term active includes all behaviour except sleep and stationary. Eating food from food hopper. Drinking water from waterspouts. The presence of at least two animals, either resting or sleeping, while in direct bodily contact (with any part of their bodies in contact). When the whole body of the rat, excluding its tail, is entirely underneath the food hopper or waterspouts at the moment of the scan. The total number of rats nesting inside the shelter; there is more than one half of the rat’s body (apart from the tail) inside the nest box. The total number of rats present on the top of the shelter; (either in an active or inactive state). The total number of rats interact with any of the enrichment objects in form of sniffing, gnawing, resting upon, playing with (pushing or pulling), sleeping inside or even being in direct physical contact with the object (sleeping, resting or setting in close contact with any of the objects).

Feeding Drinking Clump

Underneath food hopper or waterspouts

Number of animals under the shelter

Number of animals on the shelter Number of animals interacting with the enrichment objects

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

175

Table 3 Ethogram for behavioural elements recorded by focal sampling. Category

Component

Description

Measure

A. General activities:

1. Intake maintenance 2. Non-intake maintenance 3. Movement activities 4. Exploratory behaviour

Feeding, eating faeces and drinking Self-grooming and pandiculation (stretching and yawning) Movement and/or climbing the cage lid Sniffing cage wall, cage top, and sniffing air outside the cage

Duration Duration Duration Duration

B. Bedding material manipulation

1. Bedding material manipulation 2. Burrowing

Digging and pushing or pulling bedding material

Duration Duration

1. Sleep

The animal tries to bury its body, or part of it, as to be, partly inside the bedding material, or completely under it Sitting or lying flat, with eyes normally opened or nearly closed Alert (eyes opened) but with no directed attention while sitting, standing or leaning against a cage side, mate, food pot or a cage structure The neck is stretched and the animal sniffs the air inside the cage while in sitting, standing position or leaning against a cage wall or cage structure Lying unalert with both eyes closed, apparently asleep

2. Self-generated sleep interruptions 3. Sleep disruption by others 1. Social investigation 2. Allogrooming

Sleep grooming, sleep stretching, sleep sniffing, position change during sleep and shivering Sleep disruption by other cage mates vocalizations or movement Social investigation and received social investigation Allogrooming given and Allogrooming received

1. Aggression 2. Defence 3. Dominant over

Aggression, biting given and chasing given Defence, biting received and chasing received A rat positioned over another one with its forepaws placed on the latter, regardless of who initiated or received the aggressive act A rat lies on its back, fully exposing its ventral surface to another rat, regardless of who initiated or received the aggressive act Aggressive grooming given and aggressive grooming received

Frequency Frequency Duration

Mounting of cage mate, given and received Chewing cage bars Standing on the hind paws, stretching the body and taking the forepaws off the ground, while sniffing the air inside the cage Sniffing, chewing, climbing and manipulating the enrichment objects

Frequency Duration Duration

C. Stationary:

1. Rest 2. Awake non-active

3. Sniffing air inside the cage D. Sleep:

E. Non-aggressive social interaction: F. Agonism:

4. On-back-posture

5. Aggressive grooming G. Other behaviour:

1. Mounting 2. Bar-chewing 3. Rearing

H. Enrichment directed behaviours:

1. Enrichment directed behaviours

budget). Each rat was observed continuously for eight minutes per session, and for two sessions every observation day, thus yielding 16 min per rat per day (observation week) and a total 48 min per rat per replicate. During that time all instances of its behaviour were recorded. Behaviours were recorded in real time using Observer software (version 3.0, Noldus Information Technology B.V. 1994–1996) run on a PSION “Work About” (hand held computer). The elements of behaviour patterns recorded by focal sampling method are listed in the Table 3. 2.5. Statistical analyses SPSS (version 12.0 for windows) was used for all statistical analyses. Data were checked for normality and homogeneity of variances to test for the suitability of using parametric tests. Data were transformed if they did not meet the requirements of the parametric tests. In those cases where data failed to show normal distribution or homogeneity even after transformation, equivalent non-parametric tests were used. All data presented are estimated marginal means (EMM) ± SE. Different types of data were treated as follows: 2.5.1. Scan sample data The average % of rats performing the behaviour patterns recorded by scan sampling was calculated for each session by dividing the total number of rats observed performing each behaviour category across all 4 scans by the total number of rats in the cage.

Duration Duration

Duration

Frequency and duration Frequency Frequency Duration Duration

Duration

Duration

Duration

We used a General Linear Model (GLM) to test for the effects of between-subjects factors (housing treatment (standard and complex) and batch (1–3)) and within-subjects factors (repeated measures of session (light and dark), observation week (weeks 2,4,6)) and their interactions. 2.5.2. Focal sample data For each behaviour category we calculated the observation session group average for the four focal rats within a cage because the rats within the same group were unlikely to behave independently. Either the total frequency or the total duration of the behavioural variables, or both were measured. The same design of GLM as for scan sampling was used to investigate the effects of treatment, batch, session and observation week. 2.5.3. Enrichment-adjusted focal sample data To answer the question whether any potential effects of enrichment provision were directly due to increased time spent in interaction with the enrichment items or whether there were effects not solely due to this reallocation of time, data for rats in complex housing were adjusted to account for the time spent interacting with the enrichment items in each observation session. All elements of enrichment-directed behaviour that involved active physical interaction with the enrichment items including sniffing, gnawing, climbing and manipulating the enrichment were combined. Then, for each animal in the enriched housing conditions, the enrichment-adjusted observation time was calculated by sub-

176

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

Fig. 1. EMM ± SE ‘% rats underneath the food hopper and water spouts’ in the two housing conditions (data combined for both observation week and session).

tracting the total time spent interacting with the enrichment items from the total observation time. The frequency and duration of all other (non-enrichment directed) behavioural variables were then divided by the enrichment-adjusted observation time and multiplied by 480 (total length of observation session in seconds. This meant that although each behavioural variable was adjusted to account for time spent interacting with the enrichment items, the total observation time for each session remained at eight minutes (480 s) for each rat. The same GLM analysis as for scan and focal sample data was used. 3. Results 3.1. Scan sample data Housing laboratory rats in enriched versus unenriched cages did not significantly change any of the general activities recorded by scan sampling, including: ‘% rats active’; ‘% rats feeding’; and ‘% rats drinking’ (F1,6 = 0.89, NS; F1,6 = 0.14, NS and F1,6 = 3.81, NS, respectively). Similarly, neither of the two measures of laboratory rat clumping behaviour; ‘mean number of clumps’ and ‘mean number of rats in the largest clump’ was changed by adding complexity to the standard cages of laboratory rats (F1,6 = 2.57, NS; F1,6 = 2.22, NS, respectively). However, enriching the laboratory cages had a significant effect on the rats’ spatial distribution within the cages. ‘% rats underneath the food hopper and water spouts’ was much lower in the enriched cages as compared to the unenriched cages (F1,6 = 80.07, P < 0.01) (see Fig. 1). 3.2. Focal sample data 3.2.1. General activities Unenriched housed rats spent more time in exploration (F1,6 = 12.07, P < 0.05; Fig. 2), and bedding-directed activities (F1,6 = 9.29, P < 0.05) than enriched rats. There was a session*housing interaction effect (F1,6 = 63.25, P < 0.01), and an observation week*housing interaction effect (F2,12 = 8.60, P < 0.01) on the total time spent in movement activity. Post hoc t-tests showed that the unenriched rats were moving for a longer time in the dark phase (session 2) of the light/dark cycle (t10 = −3.13, P < 0.05), and in the 2nd observation week (t10 = −3.42, P < 0.01) as compared to enriched housed rats. 3.2.2. Sleep behaviour Enriched rats slept more frequently (had more bouts of sleep behaviour) than rats housed in the unenriched conditions (F1,6 = 8.83, P < 0.05). There was a session*housing interaction effect on the total time spent asleep by the laboratory rats (F1,6 = 7.09, P < 0.05). Post hoc t-tests revealed that the enriched rats slept for longer in the light phase (session 1) of the light/dark cycle

Fig. 2. EMM ± SE ‘total time of exploration and bedding-directed behaviour’ by the rats in the two housing conditions (data combined for both housing condition and observation week). * P < 0.05; ** P < 0.01.

(t10 = 3.40, P < 0.01) compared to the unenriched rats (see Fig. 3). On the other hand, there was no difference between the housing conditions on the frequency of self-generated sleep interruptions (F1,6 = 4.43, NS). 3.2.3. Agonistic behaviour There was a significant and a consistent session*housing interaction for all of the various components of agonistic activities, including: frequency of aggression (F1,6 = 5.71, P < 0.05); frequency of defence (F1,6 = 8.07, P < 0.05); duration of dominant over (F1,6 = 6.60, P < 0.05); duration of on-back-posture (F1,6 = 16.20, P < 0.01); and duration of aggressive grooming (F1,6 = 4.10, P < 0.05). Post hoc t-tests revealed that rats in the unenriched housing conditions showed a higher frequency of aggression (t10 = −2.66, P < 0.05), defence (t10 = −2.93, P < 0.05), longer duration of dominant over position (t10 = −2.18, P < 0.05), longer duration of on-backposture position (t10 = −2.43, P < 0.05) and longer duration of aggressive grooming (t10 = −2.44, P < 0.05; Fig. 4) in the dark phase of the light/dark cycle as compared to rats housed in the enriched conditions. 3.2.4. Other behaviours Housing laboratory rats in enriched or unenriched conditions did not significantly affect the amount of the time they spent awake but non-active (F1,6 = 0.42, NS), the frequency of mounting behaviour (U = 18, N1 = N2 = 6, NS), or the duration of bar-chewing (U = 17.50, N1 = N2 = 6, NS). The latter two behaviours had a very low incidence.

Fig. 3. EMM ± SE ‘total time spent sleep’ by the rats in the two housing conditions in the two observation sessions (S1 and S2) (data combined for observation week). ** P < 0.01.

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

177

While provision of additional cage structures may decrease the space allowance available to individual animals, and could therefore frustrate the quality and diversity of behaviours that can be expressed, this may be outweighed by the potential benefits discussed above. Moreover, certain objects can also, indirectly, increase the amount of absolute space in the cage. The RR shelter, for example, occupies part of the cage space, but also provides additional ‘floor’ space on top of it, hence increasing the absolute floor space available, and also compartmentalizing the cage. 4.2. Exploratory behaviour, movement and climbing

Fig. 4. EMM ± SE ‘total time of aggressive grooming’ by the rats in the two housing conditions in the two observation sessions (S1 and S2) (data combined for observation week). * P < 0.05.

3.3. Enrichment-adjusted focal sample data Following adjustment of the data for time spent interacting with enrichment objects, all the significant differences reported above between enriched and unenriched rats remained as they were for the unadjusted data. 4. Discussion The presence of enrichment objects in the cage was associated with a number of changes in rat behaviour. 4.1. Shelter use Rats in enriched cages were less frequently observed under the food hopper and water spouts than those in unenriched cages, indicating that they spent more time in the open part of the enriched cages (where the rodent retreat (RR) shelter was situated) than did rats in the unenriched cages. This may be because, when offered the opportunity, they chose to rest under the RR shelter than under the food hopper/water spouts. The lack of an age × housing treatment interaction effect suggests that any preference for the RR shelter was maintained over the six week period of the experiment. It is difficult to determine, from this experiment, the reasons for such a preference. Both food hopper and RR shelter offered some protection from bright light and conspecific attacks. However, in the RR shelter, unlike underneath the food hopper and water spouts, there was no disturbance by feeding or drinking cage mates, and this might be preferred by a resting rat. The RR shelters also provided additional wall contact (see Anzaldo et al., 1994, 1995; Lawlor, 2002) and a more enclosed space that may approximate the burrows and tunnels that rats inhabit in the wild. Manser et al. (1998) demonstrated that rats showed strong preferences for nest boxes in their cages, and that they used them to escape the light. Townsend (1997) reported that provision of standard polypropylene mouse cage (M2 type cages) reduced competition and agonistic confrontation between rats. Similarly, Chmiel and Noonan (1996) found that rats spent the majority of their time (78%) in the sheltered dark part of a specially designed cage, and therefore recommended provision of a sheltered area in the laboratory rats’ cages. Of course providing a rodent retreat in the cage does not necessarily mean that this is the perfect shelter for the rats, but it does provide them with choice. Even if two options are not the ideal ones, offering choices to animals may improve welfare by allowing animals to exert some control over the environment (e.g. Wiepkema and Koolhaas, 1993), particularly in confined laboratory conditions.

Rats in unenriched cages spent more time performing ‘exploratory’ (sniffing cage wall, lid and outside air) and beddingdirected behaviours than rats housed in enriched cages. This could simply be because they had no other cage structures (objects) to interact with, in contrast to rats in enriched cages who were able to spend time interacting with the various different enrichment objects in their environment. However, when adjusting for the time spent interacting directly with enrichment objects, this difference still remained, perhaps suggesting a more fundamental behavioural change, though it remains a possibility that ‘indirect interaction’ with enrichment objects (e.g. sitting quietly on the RR shelter) was also partly responsible for the observed results. In agreement with our findings, Orok-Edem and Key (1994) found a reduction in the frequency of bedding-directed activity (burrowing) in enriched housed rats. However in contrast to our findings, Orok-Edem and Key (1994) and Townsend (1997) observed increased exploration in enriched cages. Orok-Edem and Key’s (1994) study was 5 days (the duration of Townsend’s (1997) study was not specified), and it is possible that the addition of enrichment objects could have excited the rats temporarily causing an increase in ‘exploration’ (D’Aquila et al., 2000; Galani et al., 2001), but that subsequent habituation and a consequent drop in exploration would not have been observed in these relatively short studies. Interestingly, while reporting increases in exploratory behaviour with enrichment (Kitchen and Martin, 1996; Townsend, 1997), a reduction in other measures of general activity was also observed. We found that rats in unenriched cages spent significantly longer moving or climbing on the cage lid during the dark phase of the light/dark cycle than did rats in enriched cages. Again, this could simply be due to enriched rats spending time in other activities such as interaction with enrichment objects (e.g. ladder, rope and shelter) that also allowed climbing behaviour. Others have reported similar decreases in activity for enriched cages (Brenda et al., 2005; Townsend, 1997; Van der Harst et al., 2003). The only reported exceptions are Foulkes (2004) who found no differences in rats housed for six weeks in conventional or enriched cages (tubes and chewing objects) and Orok-Edem and Key (1994) who found increased jumping in older rats (2 years). Possibly, additional structures induce a temporary increase in activity in older rats only. 4.3. Sleep behaviour Rats in the unenriched cages slept less frequently and for shorter durations in the light phase than enriched rats, despite the fact that they were generally more active and agonistic and hence might be expected to have been more tired. Furthermore, if lying unalert with both eyes shut (our behavioural definition of sleep) partly reflected motivation to avoid light, as well as real sleep, one might also have expected more behaviour of this sort in unenriched rats who had access to fewer shelters. Conversely, the lack of shelters may actually have inhibited sleep behaviour by limiting the ability of unenriched rats to adequately avoid bright light during the light phase. Previous studies indicated that unavoidable light is likely

178

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

to be stressful for nocturnal animals (rats) resulting in a marked decrement of both types of sleep (REM and SWS) (Fishman and Roffwarg, 1972). Similarly to our results, Orok-Edem and Key (1994) reported an increase in frequency of behaviourally recorded sleep in enrichedhoused rats as against unenriched housed rats in the light phase of the light/dark cycle, however the increase was not statistically significant. The authors suggested that this trend could reflect the rats’ high activity level in the dark phase, however in our study, enriched rats were less active in the dark phase of the light/dark cycle. 4.4. Agonistic behaviour Housing rats in enriched cages greatly decreased all components of their agonistic activities, and these differences remained unchanged even after data were adjusted for the time spent interacting with the enrichment objects. Our findings thus demonstrate that this reduction in agonistic activity under enriched housed conditions was not merely a consequence of rats redirecting their behaviours away from agonistic behaviours and towards interaction with enrichment objects in their environment. One possibility is that in unenriched cages, due to the lack of any physical structures, rats themselves may be a more prominent focus of their cage-mates’ attention, and therefore attract relatively more non-aggressive and aggressive interaction. Non-aggressive social investigation has previously been reported to promote aggressive responses (e.g. Robitaille and Bovet, 1976). The lack of cage structures may also simply increase the chance that two rats come into direct contact with each other. Close proximity of animals was reported to increase aggression in mice (Van Loo et al., 2001), and the haphazard collision of two running rats might also induce an aggressive response (e.g. Robitaille and Bovet, 1976). The provision of additional structures such as a shelter, bedding and nesting materials and crawling balls into enriched cages may also have provided rats with an increased opportunity to escape aggressive attacks by cage mates, whereas in unenriched cages, retreat may have been less easy to achieve, due to the lack of suitable shelters and refuges, resulting in a frustrated escape response (see Hurst et al., 1999) and corresponding increase in escape-related behaviours such as ‘exploration’. The ability to escape an attacker in enriched cages may have resulted in fewer decisive aggressive encounters whereas in unenriched cages, where escape may have been more difficult, loser rats may have ended up adopting the submissive posture more frequently, resulting in the dominant likely embarking in a bout of aggressive grooming (see Hurst et al., 1999). The reduction of agonistic behaviours in enriched conditions may also have been due to the nature of the husbandry procedures used in this experiment, cage cleaning in particular. Cages of all rats were cleaned out completely once a week. While this process was carried out in exactly the same way for both housing types, it involved returning some of the enrichment items such as the shelter, uncleaned. The re-introduction of these items to the enriched cages may have imposed some level o familiarity to the environment (by providing familiar olfactory and visual cues), which may in turn have reduced agonistic interaction between rats in these cages. It has been shown in laboratory mice that incomplete cage cleaning regimes that allow retention of some uncleaned objects (e.g. nesting but not bedding materials) decreased intermale aggression (Van Loo et al., 2000). Orok-Edem and Key (1994) demonstrated that provision of some chewable items, such as gnawing blocks and wooden tongue depressors to the cages of laboratory rats, reduced the frequency of fighting, although the study only lasted 5 days and any novelty effects may not have worn off in this time. It is possible that allowing the animals to exert simple control over the environ-

ment (e.g. by providing objects that can promote the expression of species-specific behaviours such as gnawing), motivation for agonistic interactions may decrease. Reduced aggressive encounters and subsequent improvement in welfare, between members of group-housed animals in enriched conditions, have also been reported in pigs (O’Connell and Beattie, 1999), laying hens (Gvaryahu et al., 1994), captive primates (Kitchen and Martin, 1996) and mice (Armstrong et al., 1998; Van Loo et al., 2002, 2003). In contrast, other experiments have reported an increase in agonistic behaviours, particularly in enrichedhoused mice (Haemisch and Gartner, 1997; Haemisch et al., 1994; Van Loo et al., 2002). The increase of agonistic behaviours in enriched cages was interpreted as to be due to either close proximity and exposure of parts of the body that are vulnerable for biting, such as tails (Van Loo et al., 2002), or stimulation of territorial behaviour (Haemisch and Gartner, 1997; Haemisch et al., 1994). This increased aggression in enriched environments appears to be a species difference in response to the housing environment, with male laboratory mice being less socially tolerant than other species (e.g. Gray and Hurst, 1995). 4.5. Abnormal behaviour No significant differences in the incidence of apparently abnormal behaviours such as tail chasing or bar chewing were observed between rats housed in the different housing conditions. Previous studies have found that animals in enriched housing exhibit less stereotypic and abnormal behaviour (e.g. Würbel et al., 1998). In our study, the incidence of these behaviours, even in the unenriched housed rats, was observed very infrequently, and hence was difficult to decrease further. This may have been due to the strain used (Wistar) or perhaps different sampling techniques may have been necessary to more sensitively record these rare behaviours. 5. Conclusions The main findings of our study were that rats in the enriched environment exhibited higher levels of sleep behaviour and lower levels of agonistic behaviour than rats in unenriched environments, though no differences in abnormal behaviour were observed. The findings of this experiment importantly demonstrated that the behavioural changes in the enriched environment were not directly attributable to changes in time allocation resulting from rats interacting with the enrichment items in their environment and, for example, thereby reducing the time available in which to exhibit aggression, but seemingly because of the presence of the enrichments themselves exerting a more general effect on behavioural expression. Acknowledgements The authors would like to acknowledge the Home Office Animal Procedures Committee (HOAPC) for funding the study, and The Egyptian Government (The Egyptian Educational and Cultural Bureau) for funding U.A. Abou-Ismail. References Abou-Ismail, U.A., Burman, O.H.B., Nicol, C.J., Mendl, M., 2007. Can sleep behaviour be used as an indicator of stress in group-housed rats (Rattus norvegicus)? Animal Welfare 16, 185–188. Anzaldo, A.J., Harrison, P.C., Maghirang, R.-G., Gonyou, H.W., 1994. Increasing welfare of laboratory rats with the help of spatially enhanced cage. Animal Welfare Information Center Newsletter 5, 1–5. Anzaldo, A.J., Harrison, P.C., Riskowski, G.L., Sebek, L.A., Maghirang, R., Stricklin, W.R., Gonyou, H.W., 1995. Behavioral evaluation of spatially enhanced caging for laboratory rats at high density. Cotemporary Topics of Laboratory Animal Science 34, 56–60.

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180 Armstrong, K.R., Clark, T.R., Peterson, M.R., 1998. Use of corn-husk nesting material to reduce aggression in caged mice. Contemporary Topics in Laboratory Animal Science 37, 64–66. Belz, E.E., Kennell, J.S., Czambel, K., Rubin, R.T., Rhodes, M.E., 2003. Environmental enrichment lowers stress responsive hormones in singly housed male and female rats. Pharmacology, Biochemistry and Behaviour 76, 481–486. Bennett, E.L., Rosenzweig, M.R., Diamond, M.C., 1969. Rat brain: effects of environmental enrichment on wet and dry weights. Science 163, 825–826. Black, J.E., Sirevaag, A.M., Wallace, C.S., Savin, M.H., Greenough, W.T., 1989. Effects of complex experience on somatic growth and organ development in rats. Developmental Psychobiology 22, 727–752. Brenda, M., Elliott, B.M., Grunberg, N.E., 2005. Effects of social and physical enrichment on open field activity differ in male and female Sprague–Dawley rats. Behavioural Brain Research 165, 187–196. Callard, M.D., Bursten, S.N., Price, E.O., 2000. Repetitive backflipping behaviour in captive roof rats (Rattus rattus) and the effects of cage enrichment. Animal Welfare 9, 139–152. Chamove, A.S., 1989. Cage design reduces emotionality in mice. Laboratory Animals 23, 215–219. Chmiel, J.R.D.J., Noonan, M., 1996. Preference of laboratory rats for potentially enriching stimulus objects. Laboratory Animals 30, 97–101. D’Aquila, P.S., Peana, A.T., Carboni, V., Serra, G., 2000. Exploratory behaviour and grooming after repeated restraint and chronic mild stress: effect of desipramine. European Journal of Pharmacology 399, 43–47. Davenport, J.W., Gonzalez, L.M., Carey, J.C., Bishop, S.B., Hagquist, W.W., 1976. Environmental stimulation reduces learning deficits in experimental cretinism. Science 191, 578–579. Denny, M.S., 1975. The rat’s long-term preference for complexity in its environment. Animal Learning and Behaviour 3, 245–249. Fiala, B., Snow, F.M., Greenough, W.T., 1977. “Impoverished” rats weigh more than “Enriched” rats because they eat more. Developmental Psychobiology 10, 537–541. Fishman, R., Roffwarg, H.P., 1972. REM sleep inhibition by light in the albino rat. Experimental Neurology 36, 166–178. Foulkes, A., 2004. Do laboratory rats benefit from more cage spaces? Animal Welfare Institute Quarterly (AWI Quarterly) 53, 18–19. Fraser, A.F., Broom, D.R., 1997. Farm Animal Behaviour and Welfare, 3rd ed. Book craft, Midsomer Norton, UK. Galani, R., Duconseille, E., Bildstein, O., Cassel, J.-C., 2001. Effects of room and cage familiarity on locomotor activity measures in rats. Physiology and Behavior 74, 1–4. Gamallo, A., Villanua, A., Trancho, G., Fraile, A., 1986. Stress adaptation and adrenal activity in isolated and crowded rats. Physiology and Behavior 36 (2), 217–221. Gray, S., Hurst, J.L., 1995. The effects of cage cleaning on aggression within groups of male laboratory mice. Animal Behaviour 49, 821–826. Greenough, W.T., West, R.W., DeVoogd, T.J., 1978. Subsynaptic plate perforations: changes with age and experience in the rat. Science 202, 1096–1098. Gvaryahu, G., Ararat, E., Asaf, F.E., Lev, S.M., Weller, J.I., Robinzon, B., Snapir, N., 1994. An enrichment object that reduces aggressiveness and mortality in caged laying hens. Physiology and Behavior 55, 313–316. Haemisch, A., Gartner, K., 1997. Effects of cage enrichment on territorial aggression and stress physiology in male laboratory mice. Acta Physiologica Scandinavica 161, 73–76. Haemisch, A., Voss, T., Gartner, K., 1994. Effects of environmental enrichment on aggressive behaviour, dominance hierarchies and endocrine states in male DBA/2J mice. Physiology and Behavior 56, 1041–1048. Henderson, N., 1970. Brain weight increases resulting from environmental enrichment: a directional dominance in mice. Science 169, 776–778. Hockly, E., Cordery, P.M., Woodman, B., Mahal, A., van Dellen, A., Blakemore, C., Lewis, C.M., Hannan, A.J., Bates, G.P., 2002. Environmental enrichment slows disease progression in R6/2 Huntington’s disease mice. Annals of Neurology 51, 235–242. Home Office, 2002. Statistics of Scientific Procedures on Living Animals. Home office figures, London, Great Britain. Huck, U.W., Price, E.O., 1975. Differential effects of environmental enrichment on the open-field behaviour of wild and domestic Norway rats. Journal of Comparative and Physiological Psychology 89, 892–898. Hurst, J.L., Barnard, C.J., Hare, R., Wheeldon, E.B., West, C.D., 1996. Housing and welfare in laboratory rats: time-budgeting and pathophysiology in single-sex groups. Animal Behaviour 52, 335–360. Hurst, J.L., Barnard, C.J., Tolladay, U., Nevison, C.M., West, C.D., 1999. Housing and welfare in laboratory rats: effects of cage stocking density and behavioural predictors of welfare. Animal Behaviour 58, 563–586. Johnson, S.R., Patterson-Kane, E.G., Niel, L., 2004. Foraging enrichment for laboratory rats. Animal Welfare 13, 305–312. Kitchen, A.M., Martin, A.A., 1996. The effect of cage size and complexity on the behaviour of captive common marmosets. Callithrix jacchus jacchus. Laboratory Animal 30, 317–326. Kiyono, S., Seo, M.L., Shibagaki, M., 1981. Effects of rearing environments upon sleep–waking parameters in rats. Physiology and Behavior 26, 391–394. Lawlor, M.M., 2002. Comfortable quarter for rats in research institutions. In: Reinhardt, V., Reinhardt, A. (Eds.), Comfortable Quarters for Laboratory Animals, 9th ed. Animal Welfare Institute, Washington, DC, USA. Line, S.W., Morgan, K.N., 1991. The effects of two novel objects on the behavior of singly caged adult rhesus macaques. Laboratory Animal Science 41, 365–369.

179

Manser, C.E., Broom, D.M., Overend, P., Moris, T.H., 1998. Operant studies to determine the strength of preference in laboratory rats for nest-boxes and nesting materials. Laboratory Animals 32, 36–41. Martin, P., Bateson, P. (Eds.), 1993. Measuring Behaviour. Cambridge University Press, Cambridge, UK. McGregor, P.K., Ayling, S.J., 1999. Varied cages result in more aggression in male CFLP mice. Applied Animal Behaviour Science 26, 277–281. Mench, J., Mason, G., 1997. Using behaviour to assess animal welfare. In: Appleby, M., Hughes, B.O. (Eds.), Animal Welfare. CAB International, Wallingford, pp. 127–141. Mirmiran, M.H., Van Den Dungen, H., Uylings, H.B.M., 1982. Sleep patterns during rearing under different environmental conditions in juvenile rats. Brain Research 233, 287–298. Moncek, F., Duncko, R., Johansson, B.B., Jezova, D., 2004. Effects of environmental enrichment on stress related systems in rats. Journal of Neuroendocrinology 16, 423–431. Nevison, C.M., Hurst, J.L., Barnard, C.J., 1999. Strain-specific effects of cage enrichment in male laboratory mice (Mus Musculus). Animal Welfare 8, 361–379. Newberry, R., 1995. Environmental enrichment: Increasing the biological relevance of captive environments. Applied Animal Behaviour Science 44, 229–243. O’Connell, N.E., Beattie, V.E., 1999. Influence of environmental enrichment on aggressive behaviour and dominance relationship in growing pigs. Animal Welfare 8, 269–279. Orok-Edem, E., Key, D., 1994. Response of rats (Rattus norvegicus) to enrichment objects. Animal Technology 45, 25–30. Passineau, M.J., Green, E.J., Dietrich, W.D., 2001. Therapeutic effects of environmental enrichment on cognitive function and tissue integrity following severe traumatic brain injury in rats. Experimental Neurology 168, 373–384. Patterson-Kane, E.G., Hunt, M., Harper, D.N., 1999. Behavioral indexes of poor welfare in laboratory rats. Journal of Applied Animal Welfare Science 2, 97–110. Powell, S.B., Newman, H.A., McDonald, T.A., Bugenhagen, P., Lewis, M.H., 2000. Development of spontaneous stereotyped behavior in Deer mice: effects of early and late exposure to a more complex environment. Developmental Psychobiology 37, 100–108. Robitaille, J.A., Bovet, J., 1976. Field observations on the social behaviour of the Norway rat, Rattus norvegicus (Berkenhout). Biology of Behavior 1, 289–308. Rosenzweig, M.R., Bennett, E.L., Diamond, M.C., Wu, S.Y., Slagle, R.W., Saffran, E., 1969. Influences of environmental complexity and visual stimulation on development of occipital cortex in rat. Brain Research 14, 427–445. Roy, V., Belzung, C., Delarue, C., Chapillon, P., 2001. Environmental enrichment in BALP/c mice: effects in classical tests of anxiety and exposure to a predatory odor. Physiology and Behavior 74, 313–320. Sherwin, C.M., 2004. The influence of standard laboratory cages on rodents and the validity of research data. Animal Welfare 13, S9–15. Spangenberg, E.M.F., Augustsson, H., Dahlborn, K., Essen-Gustavsson, B., Cvek, K., 2005. Housing-related activity in rats: effects on body weight, urinary corticosterone levels, muscle properties and performance. Laboratory Animal 39, 45–57. Spring, S.E., Clifford, J.O., Tomoko, D.L., 1997. Effect of environmental enrichment devices on behaviors of single- and group-housed squirrel monkeys (Saimiri sciureus). Contemporary Topics in Laboratory Animal Science 36, 72– 75. Tagney, J., 1973. Sleep patterns related to rearing rats in enriched and impoverished environments. Brain Research 53, 353–361. Townsend, P., 1997. Use of in-cage shelters by laboratory rats. Animal Welfare 6, 95–103. Turner, C.A., Lewis, M.H., King, M.A., 2003. Environmental enrichment: effects on stereotyped behaviour and dendritic morphology. Developmental Psychobiology 43, 20–27. Van der Harst, J.E., Fermont, P.C.J., Bilstra, A.E., Spruijt, B.M., 2003. Access to enriched housing is rewarding to rats as reflected by their anticipatory behaviour. Animal Behaviour 66, 493–504. Van Gool, W.A., Mirmiran, M., 1986. Effects of aging and housing in an enriched environment on sleep–wake patterns in rats. Sleep 9, 335–347. Van Loo, P.L.P., Kruitwangen, C.L.J.J., Koolhaas, J.M., van de Weerd, H.A., van Zutphen, L.F.M., Baumans, V., 2002. Influence of cage enrichment on aggressive behaviour and physiological parameters in male mice. Applied Animal Behaviour Science 76, 65–81. van Loo, P.L.P., Kruitwangen, C.L.J.J., Koolhaas, J.M., van de Weerd, H.A., van Zutphen, L.F.M., Baumans, V., 2000. Modulation of aggression in male mice: influence of cage cleaning regime and scent marks. Animal Welfare 9, 281–295. van Loo, P.L.P., Mol, J.A., Koolhaas, J.M., van Zutphen, B.F.M., Baumans, V., 2001. Modulation of aggression in male mice: influence of group size. Physiology and Behavior 72, 675–683. Van Loo, P.L.P., van de Meer, E., Kruitwagen, C.L.J.J., Koolhaas, J.M., van Zutphen, L.F.M., Baumans, V., 2003. Strain- specific aggressive behaviour of male mice submitted to different husbandry procedures. Aggressive Behaviour 29, 69–80. Volkmar, F.R., Greenough, W.T., 1972. Rearing complexity affects branching of dendrites in the visual cortex of the rat. Science 176, 1445–1447. Watson, D.S.B., 1993. Evaluation of inanimate objects on commonly monitored variables in preclinical safety studies for mice and rats. Laboratory Animal Science 43, 378–380. Wiepkema, P.R., Koolhaas, J.M., 1993. Stress and animal welfare. Animal Welfare 2, 195–218.

180

U.A. Abou-Ismail et al. / Behavioural Processes 85 (2010) 172–180

Wolfer, D.P., Litvin, O., Morf, S., Nitsch, R.M., Peter-Lipp, H., Würbel, H., 2004. Cage enrichment and mouse behaviour: test responses by laboratory mice are unperturbed by more entertaining housing. Nature 432, 821–822. Würbel, H., Chapman, R., Rutland, C., 1998. Effect of food and environmental enrich-

ment on development of stereotypic wire-gnawing in laboratory mice. Applied Animal Behaviour Science 60, 69–81. Würbel, H., 2001. Ideal homes? Housing effects on rodent brain and behaviour. Trends in Neuroscience 24, 207–211.