Motivation to dust-bathe of laying hens housed in cages and in aviaries

New housing systems for commercial egg production, furnished cages and non-cage systems, should improve the welfare of laying hens. In particular, thanks to the presence of a litter area, these new housing systems are thought to satisfy the dust-bathing motivation of hens more than in conventional cages, in which no litter area is present. However, although apparently obvious, there is no concrete evidence that non-cage systems, particularly aviaries, satisfy hens' motivation to dust-bathe and thus improve hens' welfare in terms of dust-bathing behaviour. The aim of this study was to compare hens' dust-bathing motivation when housed for a long time under similar conditions to commercial conditions in laying aviaries (with litter) and in conventional cages (without litter). Three treatments were compared: hens reared in floor pens then housed in conventional cages, hens reared in furnished floor pens then housed in a laying aviary, and hens reared in rearing aviaries then housed in a laying aviary. All three treatments provided access to litter during the rearing period. After transfer to the laying systems, access to litter was maintained for the aviary hens but stopped for the cage hens. Twelve groups of four hens per treatment were tested 36 to 43 weeks after transfer. The hens were placed in sawdust-filled testing arenas, and latency to dust-bathe, duration and number of dust baths, and number of hens dust-bathing were recorded. Latency to dust-bathe was shorter, dust baths were longer and more numerous and more hens dust-bathed among cage hens than among aviary hens. Our results indicate that hens' motivation to dust-bathe was more satisfied in laying aviaries than in conventional cages. Thus, laying aviaries improve hens' welfare in term of dust-bathing behaviour compared with conventional cages.     

Welfare, health, and hygiene of laying hens housed in furnished cages and in alternative housing systems

The aim of this review was to compare welfare, health, and hygienic status of laying hens housed in furnished cages and in alternative systems. In alternative systems (floor housing and aviaries), birds have more freedom of movement and a more complex environment than in furnished cages. However, housing birds in much larger flocks in alternative systems leads to an increased risk of feather-pecking. Furthermore, air quality can be poorer in alternative systems than in furnished cages. This can affect health and hygienic status. There are only limited data on a direct comparison between furnished cages and alternative systems. Therefore, there is a need for an on-farm comparison of welfare, health, and hygienic status in these systems.     

Furnished cages for laying hens: study of the effects of group size and litter provision on laying location, zootechnical performance and egg quality

The furnished cage is a new housing system for layers. A current trend in furnished cage design is to increase group size and replace the litter box with a mat provided with litter. An experiment was set up to determine the effects of group size and litter provision on laying performance and egg quality of beak-trimmed ISA Brown hens housed in large furnished cages with more than 12 hens. Six treatments, each of 18 furnished cages (768 cm(2)/hen including nest and litter area) were compared in a 3 × 2 experimental trial: three group sizes (S20 (20 hens per cage), S40 (40) and S60 (60)), with or without feed as litter distributed on the mat of the litter area. The provision of facilities per hen was equal in all treatments. Mortality, laying rate, mean egg weight, feed intake and feed conversion ratio were unaffected by group size over the 53-week laying period, and performance exceeded the ISA production standards. The overall percentage of eggs laid in the nest exceeded 95% except that it was slightly lower in group S20 (92.0% ± 6.4% v. S40: 96.0% ± 3.3% and S60: 96.2% ± 2.7%) leading to a higher proportion of dirty eggs (S20: 1.6% ± 2.2%, S40: 1.4% ± 1.5%, S60: 1.0% ± 1.0%). At 66 to 70 weeks, eggs laid outside the nest had a slightly higher count of mesophilic bacteria on the eggshell (5.0 log CFU/egg ± 0.4) than those laid in the nest (4.8 log CFU/egg ± 0.5) but no difference in contamination was observed between group sizes. Litter provision had no effect on mortality, egg weight or egg quality traits except for a higher proportion of broken eggs in cages with litter (5.3% ± 6.2% v. 4.6% ± 5.7%). Providing hens with feed for litter was associated with a higher laying rate (97.3% ± 3.2% v. 94.8% ± 4.4% at 23 weeks) and an apparent improvement in feed efficiency at the beginning of the laying period (feed conversion ratio based on feed consumption at the trough: 2.18 ± 0.06 with litter v. 2.28 ± 0.09 without litter at 25 weeks). The results of this study showed that a high level of productivity and good egg quality could be obtained in large furnished cages. Further research is needed to assess the impact on hens' welfare and performance of using more economically competitive substrates than feed for litter.     

Effects of housing conditions during the rearing and laying period on adrenal reactivity, immune response and heterophil to lymphocyte (H/L) ratios in laying hens

This study was conducted to evaluate the effect of early rearing conditions on physiological, haematological and immunological responses relevant to adaptation and long-term stress in white Leghorn hens with intact beaks housed in furnished cages (FC) or conventional cages (CC) during the laying period. Pullets were cage reared (CR) or litter floor reared (FR). From 16 to 76 weeks of age, hens were housed in FC (eight hens per cage) or in CC (three hens per cage). As measures of long-term stress at the end of the laying period, adrenal reactivity was quantified by assessing corticosterone responses to adrenocorticotropin challenge, and immune response was assessed by measuring antibody responses after immunization with sheep red blood cells (SRBC) and keyhole limpet haemocyanin (KLH). Heterophil to lymphocyte (H/L) ratio was employed as an indicator of stress. Rearing conditions significantly affected anti-SRBC titres (P < 0.0001) and tended to affect H/L ratios (P = 0.07), with the highest values found in FR hens. Layer housing affected H/L ratio (P < 0.01); the highest ratio was found in FR birds housed in FC during the laying period. This study shows that early rearing environment affects immunological indicators that are widely used to assess stress in laying hens. However, while results on H/L ratio indicated that FR birds experienced more stress particularly when they were housed in FC during the laying period, the immune responses to SRBC in FR hens was improved, indicating the opposite. This contradiction suggests that the effects on immune response may have been associated with pathogenic load due to environmental complexity in FR and FC hens rather than stress due to rearing system or housing system per se.     

Does Rearing Laying Hens in Aviaries Adversely Affect Long-Term Welfare following Transfer to Furnished Cages?

This study tests the hypothesis that hens that are reared in aviaries but produce in furnished cages experience poorer welfare in production than hens reared in caged systems. This hypothesis is based on the suggestion that the spatial restriction associated with the transfer from aviaries to cages results in frustration or stress for the aviary reared birds. To assess the difference in welfare between aviary and cage reared hens in production, non-beak trimmed white leghorn birds from both rearing backgrounds were filmed at a commercial farm that used furnished cage housing. The videos were taken at 19 and 21 weeks of age, following the birds'' transition to the production environment at 16 weeks. Videos were analysed in terms of the performance of aversion-related behaviour in undisturbed birds, comfort behaviour in undisturbed birds, and alert behaviour directed to a novel object in the home cage. A decrease in the performance of the former behaviour and increase in the performance of the latter two behaviours indicates improved welfare. The results showed that aviary reared birds performed more alert behaviour near to the object than did cage reared birds at 19 but not at 21 weeks of age (P = 0.03). Blood glucose concentrations did not differ between the treatments (P>0.10). There was a significant difference in mortality between treatments (P = 0.000), with more death in aviary reared birds (5.52%) compared to cage birds (2.48%). The higher mortality of aviary-reared birds indicates a negative effect of aviary rearing on bird welfare, whereas the higher duration of alert behavior suggests a positive effect of aviary rearing.     

Effects of separation of resources on behaviour of high-, medium- and low-ranked hens in furnished cages

In our previous studies, we demonstrated that dominant hens had priority in using the dust bath, resulted in increased competition for the resource. It seemed that the problem was that the resource was placed on one side of the cage (‘localised’). Therefore, we designed a medium-sized furnished cage with a dust bath and nest box on both sides of the cage (‘separated’, MFS). To evaluate the effects of separation of these resources, we compared the behaviour of high-, medium- and low-ranked hens in MFS cage with that in small (SF) and medium furnished (MFL) cages with a localised resource. In total, 150 White Leghorn layers were used. At the age of 17 weeks, the hens were randomly divided into three groups and moved to small furnished cages (SF, 90 cm wide; five birds per cage) and two types of medium furnished cages (180 cm wide; 10 birds per cage) with a nest box and dust bath on both sides (MFS) and a nest box and dust bath on one side of the cage (MFL). The total dust bath and nest box areas per hen were same for the three cages. The dominance hierarchy was determined by observing the aggressive interactions and by this high-, medium- and low-ranked hens in each cage were identified. The behaviour, use of facilities and physical condition of these hens were measured. Data were analysed by using repeated measure ANOVA. A significant interaction between social order and cage design was found in the proportions of time spent in the dust bath and on performing dust-bathing (both P < 0.001), and these proportions tended to be higher in higher-ranked hens in SF and MFL. Conversely, the MFS low-ranked hens tended to use the dust bath more than the SF and MFL low-ranked hens. Thus, hens from each rank used the dust bath equally in MFS, though the MFS high-ranked hens tended to use the resource less than the SF and MFL high-ranked hens. While the frequency of pre-laying sitting was lower among low-ranked hens (P < 0.05), the proportion of time in the nest box was higher among low- than high-ranked hens (P < 0.01). The low-ranked hens spent more time performing escaping, moving and standing in the nest box. In conclusion, it is suggested that separation of the dust bath to two locations would be an effective arrangement to promote more equal usage of the dust bath by hens from each rank in the furnished cages. It was also confirmed in the present study that nest boxes were not only used for laying eggs but also as a refuge by lower ranked hens.     

A machine vision system to detect and count laying hens in battery cages

Manually counting hens in battery cages on large commercial poultry farms is a challenging task: time-consuming and often inaccurate. Therefore, the aim of this study was to develop a machine vision system that automatically counts the number of hens in battery cages. Automatically counting hens can help a regulatory agency or inspecting officer to estimate the number of living birds in a cage and, thus animal density, to ensure that they conform to government regulations or quality certification requirements. The test hen house was 87 m long, containing 37 battery cages stacked in 6-story high rows on both sides of the structure. Each cage housed 18 to 30 hens, for a total of approximately 11 000 laying hens. A feeder moves along the cages. A camera was installed on an arm connected to the feeder, which was specifically developed for this purpose. A wide-angle lens was used in order to frame an entire cage in the field of view. Detection and tracking algorithms were designed to detect hens in cages; the recorded videos were first processed using a convolutional neural network (CNN) object detection algorithm called Faster R-CNN, with an input of multi-angular view shifted images. After the initial detection, the hens’ relative location along the feeder was tracked and saved using a tracking algorithm. Information was added with every additional frame, as the camera arm moved along the cages. The algorithm count was compared with that made by a human observer (the ‘gold standard’). A validation dataset of about 2000 images achieved 89.6% accuracy at cage level, with a mean absolute error of 2.5 hens per cage. These results indicate that the model developed in this study is practicable for obtaining fairly good estimates of the number of laying hens in battery cages.     

Testing resource value in group-housed animals: an investigation of cage height preference in laying hens

To avoid unpredictable social effects, animals' behavioural priorities are almost always tested using individuals housed singly, yet many species kept commercially are social animals housed in groups. Our aim was to develop a method of investigating environmental preference in group-housed laying hens, Gallus gallus domesticus, that maximised the external validity of our findings. In a simple test of preference, eight groups of ten hens were given free choice between furnished cages with minimum heights of 38 cm (low) and 45 cm (high). A preference for one cage height over the other would be evident as a shift from a binomial distribution of flock sizes in the two cages. No height preference was found as hens distributed evenly between the two cages more frequently than was expected. This suggests at high stocking densities maximising average inter-individual distance could be a priority over increased cage height. In a second experiment, to investigate the value that hens placed on a change in cage height; a 'cost' in the form of a narrow gap was imposed on movement from a low or high start cage to a high or low target cage, respectively. Cage height did not influence the latency of the first three hens to enter the target cage. However, latencies for subsequent hens were shorter and more hens worked to access a high target cage than a low target cage. We suggest that titrating animals' willingness to tolerate higher stocking densities against access to a resource could be an effective way to compare responses of group-housed animals to resources that are expected to satisfy the same motivational state.     

Do Laying Hens with Keel Bone Fractures Experience Pain?

The European ban on battery cages has forced a change towards the use of non-cage or furnished cage systems, but unexpectedly this has been associated with an increased prevalence of keel bone fractures in laying hens. Bone fractures are acutely painful in mammals, but the effect of fractures on bird welfare is unclear. We recently reported that keel bone fractures have an effect on bird mobility. One possible explanation for this is that flying becomes mechanically impaired. However it is also possible that if birds have a capacity to feel pain, then ongoing pain resulting from the fracture could contribute to decreased mobility. The aim was to provide proof of concept that administration of appropriate analgesic drugs improves mobility in birds with keel fracture; thereby contributing to the debate about the capacity of birds to experience pain and whether fractures are associated with pain in laying hens. In hens with keel fractures, butorphanol decreased the latency to land from perches compared with latencies recorded for these hens following saline (mean (SEM) landing time (seconds) birds with keel fractures treated with butorphanol and saline from the 50, 100 and 150 cm perch heights respectively 1.7 (0.3), 2.2 (0.3), p = 0.05, 50 cm; 12.5 (6.6), 16.9 (6.7), p = 0.03, 100 cm; 20.6 (7.4), 26.3 (7.6), p = 0.02 150 cm). Mobility indices were largely unchanged in birds without keel fractures following butorphanol. Critically, butorphanol can be considered analgesic in our study because it improved the ability of birds to perform a complex behaviour that requires both motivation and higher cognitive processing. This is the first study to provide a solid evidential base that birds with keel fractures experience pain, a finding that has significant implications for the welfare of laying hens that are housed in non-cage or furnished caged systems.     

The European Union ban on conventional cages for laying hens: history and prospects

Since the publication of Animal Machines (Harrison, 1964), there has been widespread public pressure in Europe-supported by European institutions-to “ban the battery cage. ”The European Union (EU) and national governments (particularly in Northern Europe) funded research on noncage systems for egg production and enriched cages. In 1986, the EU passed a Directive specifying a minimum size for cages, but public opinion-again particularly in the North-continued to require more. A market sector emerged that would pay more for noncage eggs. Denmark, Sweden, and Switzerland passed more stringent legislation than the rest of Europe. A 1999 Directive with details based on advice from the EU's Scientific Veterinary Committee will phase out conventional laying cages but allow enriched cages. Implementation depends on various factors, including negotiations in the World Trade Organization. In the next 10 years, however, major changes to the housing of most laying hens in Europe almost certainly will occur. Similar changes in other countries will follow. As in Europe, change probably will be piecemeal, affected both by public pressure and by all sectors of society: producers, retailers, consumers, legislators, and the media.     

The humoral immune response and the productivity of laying hens kept on the ground or in cages

The effects of two different keeping systems on the humoral immune response and productivity were compared for 80 laying hens, divided into four groups. Two groups each of 20 hens were kept on the ground and two were kept in cages. All the birds were immunised subcutaneously with human serum immunoglobulin G (IgG) at a dose of 100(microg per injection. The immunisations were performed twice at 4-week intervals. The lipopeptide Pam(3)Cys-Ser-(Lys)(4) was used as an adjuvant at a dose of 0.25mg per injection in one group from each housing system. In the second group from each housing system, the hens were immunised without any adjuvant (antigen control groups). The mean egg yield was significantly higher in both the antigen control group and the adjuvant group, when laying hens were kept in cages. Total egg weight remained constant in both of the housing systems. Keeping hens in cages resulted in higher mean specific antibody titres and mean immunoglobulin Y concentrations in the egg yolk.     

Animal Welfare Issues in the Poultry Industry: Is There a Lesson to Be Learned?

Many of the conditions in which poultry live and the procedures to which they are subjected compromise their welfare. This article describes these welfare problems in the hope that they may serve as warnings to the rest of animal agriculture, which then might take steps to avoid the same pitfalls. The article discusses poultry welfare problems under the headings of battery cages for laying hens, forced molting, disposal of spent laying hens, fast growth problems in meat poultry, catching and transportation, food restriction of broiler breeders, hyperaggressiveness in broiler breeder males, elective surgeries, and water-bath stunning.     

Assessment of the effect of housing on feather damage in laying hens using IR thermography

Plumage damage represents one of the animal-based measures of laying hens welfare. Damage occurs predominantly due to age, environment and damaging pecking. IR thermography, due to its non-invasiveness, objectivity and repeatability is a promising alternative to feather damage scoring systems such as the system included in the Welfare Quality^R assessment protocol for poultry. The aim of this study was to apply IR thermography for the assessment of feather damage in laying hens kept in two housing systems and to compare the results with feather scoring. At the start of the experiment, 16-week-old laying hens (n=30) were divided into two treatments such as deep litter pen and enriched cage. During 4 months, feather damage was assessed regularly in 2-week intervals. One more single assessment was done nine and a half months after the start of the experiment. The feather damage on four body regions was assessed by scoring and IR thermography: head and neck, back and rump, belly, and underneck and breast. Two variables obtained by IR thermography were used: the difference between the body surface temperature and ambient temperature (ΔT_B) and the proportion of featherless areas, which were defined as areas with a temperature >33.5°C. Data were analyzed using a GLM model. The effects of housing, time, region and their interactions on feather damage, measured by the feather scoring and by both IR thermography measures, were all significant (P<0.001). The ΔT_B in all assessed regions correlated positively with the feather score. Feather scoring revealed higher damage in enriched cages compared with deep litter pens starting from week 6 of the experiment on the belly and back and rump regions, whereas ΔT_B from week 6 in the belly and from week 8 on the back and rump region. The proportion of featherless areas in the belly region differed significantly between the housings from week 8 of the experiment and on the back and rump region from week 12. The IR thermography assessment of the feather damage revealed differences between hens kept in different housing systems in agreement with the feather scoring. In conclusion, it was demonstrated that the IR thermography is a useful tool for the assessment of poultry feather cover quality that is not biased by the subjective component and provides higher precision than feather damage scoring.     

Behaviour of laboratory mice in different housing conditions when allowed to self-administer an anxiolytic

Standard cages prevent mice from performing several natural behaviours for which they are motivated. As a consequence, abnormal behaviours sometimes develop and mice often spend long periods inactive. To improve welfare, cages are sometimes furnished with items such as nesting material, shelters and running wheels. We have previously reported that when allowed to self-administer an anxiolytic, mice in furnished cages consume less anxiolytic than mice in standard cages. This paper presents the results of behaviour studies of the mice in the same experiment. Female C57BL/6J mice (3 per cage) were housed in Standard (n = 10), Unpredictable (n = 10) or Furnished (n = 6) cages. Unpredictable cages were identical to Standard cages, but were exposed to unpredictable events two to three times a week. Furnished cages were double the size of Standard cages and contained nesting material, nest box, tubes, chew blocks and a running wheel. During three consecutive periods, mice had access to only water (control), water or an anxiolytic solution on a daily alternating schedule (forced consumption), and finally, both water and anxiolytic (self-administration). Behaviour was analysed from video recordings taken during the dark phase. The housing type affected behaviour both under the control and the self-administration conditions. Overall, mice in Furnished cages spent less time resting and performing bar-related behaviours and more time on exploratory/locomotory behaviours. Mice in Furnished cages also performed less bar-circling stereotypies than mice in Standard cages. The Unpredictable treatment did not significantly affect behaviour compared to mice in the Standard conditions. There was an overall effect of anxiolytic availability on rest-related behaviours and on exploration-locomotion behaviours, in that mice rested more and spent less time on exploration and locomotion when they were able to self-administer the anxiolytic.     

Improving housing conditions for laboratory mice: a review of "environmental enrichment"

Laboratory animal facilities have been designed to provide a standard environment where animals can be kept in good physical health at the same time as economic and ergonomic considerations are met. Recognizing the potential welfare problem associated with behavioural restriction in such housing systems, a number of attempts have been made to improve this environment, generally described under the term "environmental enrichment". Modifications of cages for mice usually consist of providing material for nest building and structures which can serve as hiding places and/or for climbing. We have reviewed 40 studies carried out between 1987 and 2000, in which preferences as well as the effect of housing modifications have been studied. Mice will work for access to nesting material and make use of this material to make nests in which they rest. They prefer a more complex cage to the standard cage and will also work for access to cages with shelter and raised platforms. On the basis of present knowledge, it is recommended that mice should have access to nesting material. Strategies for future research are outlined in the article.     

The effect of preweaning and postweaning housing on the behaviour of the laboratory mouse (Mus musculus)

Standard housing for laboratory mice severely restricts natural behaviour and the control that the animal has over its environment. Providing the cage with objects is a method that has been used to both increase environmental complexity, promote the performance of natural behaviour and provide greater controllability for the animal. This method of furnishing cages has mostly been studied in adult animals, and little is known about the influence that the preweaning environment has on the behaviour of mice as adults. This study aimed to investigate the effects on mice behaviour of preweaning and postweaning housing environment. In this experiment, 64 pairs of animals of the strain C57BL/6J were used. Half of the animals were born and reared until weaning in standard cages and the other half in cages twice the size of the standard and furnished with nesting material, a cardboard tube, a PVC nest box and a wooden chewblock. After weaning, half the animals in each group were changed to the other type of cage, whereas the other half remained in the same environment; in both cases they were kept in single-sex pairs of littermates. Behaviour during the dark, active period was studied through video recordings. We found no main effects of preweaning environment on behaviour; however, mice moved from furnished to standard cages at weaning showed a decrease in inactive behaviour at four weeks of age. Mice housed after weaning in standard cages spent less time inactive, and more time engaging in activities like feeding and drinking, self-grooming and allogrooming. A sex difference was also found, in that females showed a greater performance of exploratory behaviour as well as a greater prevalence of stereotypies. The use of different objects and locations within the furnished cage was also analysed at both ages. Results show that at eight weeks of age mice spent more time at the top of the cage, and that the use of the nest box (although not for resting) increased between four and eight weeks. Mice were found to use the nest box as a nesting site/sleeping place only at age four weeks, whereas they always used the nesting material for sleeping.     

Review: Automated techniques for monitoring the behaviour and welfare of broilers and laying hens: towards the goal of precision livestock farming

There is increasing public concern about poultry welfare; the quality of animal welfare is closely related to the quality of livestock products and the health of consumers. Good animal welfare promotes the healthy growth of poultry, which can reduce the disease rate and improve the production quality and capacity. As behaviour responses are an important expression of welfare, the study of behaviour is a simple and non-invasive method to assess animal welfare. The use of modern technology offers the possibility to monitor the behaviour of broilers and laying hens in a continuous and automated way. This paper reviews the latest technologies used for monitoring the behaviour of broilers and laying hens under both experimental conditions and commercial applications and discusses the potential of developing a precision livestock farming (PLF) system. The techniques that are presented and discussed include sound analysis, which can be an online tool to automatically monitor poultry behaviour non-invasively at the group level; wireless, wearable sensors with radio-frequency identification devices, which can automatically identify individual chickens, track the location and movement of individuals in real time and quantify some behavioural traits accordingly and image processing technology, which can be considered a direct tool for measuring behaviours, especially activity behaviours and disease early warning. All of these technologies can monitor and analyse poultry behaviour, at the group level or individual level, on commercial farms. However, the popularity and adoption of these technologies has been hampered by the logistics of applying them to thousands and tens of thousands of birds on commercial farms. This review discusses the advantages and disadvantages of these techniques in commercial applications and presents evidence that they provide potential tools to automatically monitor the behaviours of broilers and laying hens on commercial farms. However, there still has a long way to go to develop a PLF system to detect and predict abnormal situations.     

Do hens suffer in battery cages? environmental preferences and welfare

The question ‘Do hens suffer in battery cages?’ is difficult to answer because of the problem of objectively assessing suffering in animals. It is argued that preference tests may be one way of throwing light on this difficult problem. This paper describes some experiments on habitat preference in domestic hens. No preference was observed between a commercial battery cage and a large pen when hens were given continuous access to the two. A simultaneous choice between a battery cage and an outside hen-run showed a clear preference for the run, but choice was strongly influenced by prior experience. The strength of the run preference was investigated by ‘pitting’ the run against food and access to companions.     

Non-exclusive spatial preference in the laying hen

Two experiments were performed to elucidate aspects of the spatial preference of domestic hens kept in battery cages.In the first experiment, 12 hens were given a choice between four cages differing in size and shape. The measure of preference used was the amount of time spent in each cage under free-access conditions. The results show that the hens spent most time in the cage with the largest floor area, but that substantial amounts of time were spent in the remaining three cages. Eight possible explanations for this non-exclusive choice are proposed: variation among hens; change of preference with time; mis-identification; constraints on decision-making; pacing; sampling; monitoring; genuine preference. The first six hypotheses are discussed with respect to this experiment.The second experiment, using 6 hens under similar conditions, was designed to distinguish between the latter two hypotheses. The results suggest there may be a monitoring component to the hens' behaviour in this apparatus, characterised by cage visits of short duration. In addition, the hens may continue to visit small cages in order to perform particular behaviour patterns. Nest-building behaviour is mentioned as one possibility.     

The effects of cage volume and cage shape on the condition and behaviour of captive European starlings (Sturnus vulgaris)

Cage size is widely recognised as an important determinant of captive animal welfare, but in contrast, cage shape has received far less attention. Husbandry recommendations for flying birds state that cages should be long in shape because this allows greater potential for flight. However, so far no studies have investigated the impact of cage shape on the behaviour or welfare of captive flying birds. We measured the effects of cage size and shape on the condition and behaviour of captive wild-caught European starlings (Sturnus vulgaris) using a 2 × 3 factorial design in which birds were individually housed for 1 week in cages of one of two volumes (either medium at 0.3 m³ or large at 1 m³), and one of three shapes (long with an aspect ratio (i.e. length/height) of 3.43), standard with an aspect ratio of 1.72 or tall with an aspect ratio of 0.86). We found effects of cage size and the interaction of cage size and shape on the behaviour and condition of birds. In interpreting the welfare implications of our results we focused on stereotypic behaviour as measured by incidence of somersaulting and a novel statistic that quantifies sequential dependencies in the birds’ locations within the cage. The lowest measures of stereotypic behaviour were recorded in the large cages and the medium long cage. Cage shape was more important in determining the quantity of stereotypic behaviour in the medium cages than the large cages. Our findings support the recommendation that starlings be housed in long-shaped cages.