Conventional laboratory housing increases morbidity and mortality in research rodents: results of a meta-analysis.

BACKGROUND: Over 120 million mice and rats are used annually in research, conventionally housed in shoebox-sized cages that restrict natural behaviours (e.g. nesting and burrowing). This can reduce physical fitness, impair thermoregulation and reduce welfare (e.g. inducing abnormal stereotypic behaviours). In humans, chronic stress has biological costs, increasing disease risks and potentially shortening life. Using a pre-registered protocol ( https://atrium.lib.uoguelph.ca/xmlui/handle/10214/17955 ), this meta-analysis therefore tested the hypothesis that, compared to rodents in 'enriched' housing that better meets their needs, conventional housing increases stress-related morbidity and all-cause mortality. RESULTS: Comprehensive searches (via Ovid, CABI, Web of Science, Proquest and SCOPUS on May 24 2020) yielded 10,094 publications. Screening for inclusion criteria (published in English, using mice or rats and providing 'enrichments' in long-term housing) yielded 214 studies (within 165 articles, using 6495 animals: 59.1% mice; 68.2% male; 31.8% isolation-housed), and data on all-cause mortality plus five experimentally induced stress-sensitive diseases: anxiety, cancer, cardiovascular disease, depression and stroke. The Systematic Review Center for Laboratory animal Experimentation (SYRCLE) tool assessed individual studies' risks of bias. Random-effects meta-analyses supported the hypothesis: conventional housing significantly exacerbated disease severity with medium to large effect sizes: cancer (SMD = 0.71, 95% CI = 0.54-0.88); cardiovascular disease (SMD = 0.72, 95% CI = 0.35-1.09); stroke (SMD = 0.87, 95% CI = 0.59-1.15); signs of anxiety (SMD = 0.91, 95% CI = 0.56-1.25); signs of depression (SMD = 1.24, 95% CI = 0.98-1.49). It also increased mortality rates (hazard ratio = 1.48, 95% CI = 1.25-1.74; relative median survival = 0.91, 95% CI = 0.89-0.94). Meta-regressions indicated that such housing effects were ubiquitous across species and sexes, but could not identify the most impactful improvements to conventional housing. Data variability (assessed via coefficient of variation) was also not increased by 'enriched' housing. CONCLUSIONS: Conventional housing appears sufficiently distressing to compromise rodent health, raising ethical concerns. Results also add to previous work to show that research rodents are typically CRAMPED (cold, rotund, abnormal, male-biased, poorly surviving, enclosed and distressed), raising questions about the validity and generalisability of the data they generate. This research was funded by NSERC, Canada.

Don't use it? Don't lose it! Why active use is not required for stimuli, resources or "enrichments" to have welfare value.

Current frameworks for designing and evaluating good enclosures and "enrichments" typically focus on animals' active interactions with these features. This has undoubtedly improved the welfare of zoo-housed animals over the last 30 years or more. However, literature reviews from this same period identify persistent gaps in how such frameworks are applied: experiences and behaviors that do not rely on active interaction with stimuli or resources are largely ignored, when evaluating the welfare value of enclosures and enrichments within them. Here, we review research evidence demonstrating that active interaction is not always a reliable measure of welfare value, showing that items that elicit little or no interaction can nevertheless still reduce stress and improve well-being. This evidence largely comes from research on humans, lab animals and farm animals, but also from some zoo studies too. We then investigate why. We review psychology and ethology literatures to show that such welfare benefits can arise from five, non-mutually exclusive, processes or mechanisms that are well-understood in humans and domestic animals: (1) some motivations are sated quickly by interaction with resources, yet still have large welfare benefits; (2) active interaction may just be a way to achieve a goal or solve a problem, without being beneficial for welfare in itself; (3) having opportunities for choice and control may be inherently beneficial, even when not acted on; (4) some enclosure features meet social needs for structure, landmarks, and blocked sightlines; and (5) some stimuli may be preferred because they signaled good environments to an animal's ancestors. We use this information to identify  improved ways of enhancing and assessing zoo animal welfare. Incorporating these concepts should expand the scope of behaviors and subjective experiences that are targeted, to now include those that involve little active interaction and yet still are important for good welfare.

Practice does not make perfect: Juvenile object play does not improve the predation skills of adult mink in "simulated prey" tests.

In carnivores, juvenile object play is hypothesized to improve the development of adult predation ability. We tested this hypothesis in a model carnivore, the American mink (Neovison vison). Play was induced via the provision of diverse "enrichment" objects to 32 litters from 4 to 15 weeks of age on a rotating schedule; control mink (32 litters) were instead provided with one static object. Behavioral observations confirmed that enriched juveniles engaged in more object play than control juveniles (p < .01). Tenacity, strength, and chasing motivations were assessed in 52 mink at 27-29 weeks of age using five "simulated prey" tests. These tests showed internal consistency, with measures from the same test being weakly to very strongly correlated (r or ρ: |.33-.95|). Across tests assessing similar skills, some measures of tenacity and strength were also moderately correlated. Treatment had little effect, however, other than on maximum force during vertical pulling, in which control mink unexpectedly scored higher (1.01 ± 0.06 kg vs. 0.84 ± 0.06 kg, p = .0458). Although we did not find evidence to support the hypothesis that object play leads to improved predatory skills, the tests developed have the potential to be used in humane, prey-free future investigations of predation abilities.

Nature calls: intelligence and natural foraging style predict poor welfare in captive parrots.

Understanding why some species thrive in captivity, while others struggle to adjust, can suggest new ways to improve animal care. Approximately half of all Psittaciformes, a highly threatened order, live in zoos, breeding centres and private homes. Here, some species are prone to behavioural and reproductive problems that raise conservation and ethical concerns. To identify risk factors, we analysed data on hatching rates in breeding centres (115 species, 10 255 pairs) and stereotypic behaviour (SB) in private homes (50 species, 1378 individuals), using phylogenetic comparative methods (PCMs). Small captive population sizes predicted low hatch rates, potentially due to genetic bottlenecks, inbreeding and low availability of compatible mates. Species naturally reliant on diets requiring substantial handling were most prone to feather-damaging behaviours (e.g. self-plucking), indicating inadequacies in the composition or presentation of feed (often highly processed). Parrot species with relatively large brains were most prone to oral and whole-body SB: the first empirical evidence that intelligence can confer poor captive welfare. Together, results suggest that more naturalistic diets would improve welfare, and that intelligent psittacines need increased cognitive stimulation. These findings should help improve captive parrot care and inspire further PCM research to understand species differences in responses to captivity.

Phylogenetic comparative methods: Harnessing the power of species diversity to investigate welfare issues in captive wild animals.

This paper reviews a way of investigating health and welfare problems in captive wild animals (e.g., those in zoos, aviaries, aquaria, or aquaculture systems) that has great potential, but to date has been little used: systematically comparing species with few or no health and welfare issues to those more prone to problems. Doing so empirically pinpoints species-typical welfare risk and protective factors (such as aspects of their natural behavioral biology): information which can then be used to help prevent or remedy problems by suggesting new ways to improve housing and husbandry, and by identifying species intrinsically best suited to captivity. We provide a detailed, step-by-step "how to" guide for researchers interested in using these techniques, including guidance on how to statistically control for the inherent similarities shared by related species: an important concern because simple, cross-species comparisons that do not do this may well fail to meet statistical assumptions of non-independence. The few relevant studies that have investigated captive wild animals' welfare problems using this method are described. Overall, such approaches reap value from the great number and diversity of species held in captivity (e.g., the many thousands of species held in zoos); can yield new insights from existing data and published results; render previously intractable welfare questions (such as "do birds need to fly?" or "do Carnivora need to hunt?") amenable to study; and generate evidence-based principles for integrating animal welfare into collection planning.

Mixed-strain housing for female C57BL/6, DBA/2, and BALB/c mice: validating a split-plot design that promotes refinement and reduction.

BACKGROUND: Inefficient experimental designs are common in animal-based biomedical research, wasting resources and potentially leading to unreplicable results. Here we illustrate the intrinsic statistical power of split-plot designs, wherein three or more sub-units (e.g. individual subjects) differing in a variable of interest (e.g. genotype) share an experimental unit (e.g. a cage or litter) to which a treatment is applied (e.g. a drug, diet, or cage manipulation). We also empirically validate one example of such a design, mixing different mouse strains -- C57BL/6, DBA/2, and BALB/c -- within cages varying in degree of enrichment. As well as boosting statistical power, no other manipulations are needed for individual identification if co-housed strains are differentially pigmented, so also sparing mice from stressful marking procedures. METHODS: The validation involved housing 240 females from weaning to 5 months of age in single- or mixed- strain trios, in cages allocated to enriched or standard treatments. Mice were screened for a range of 26 commonly-measured behavioural, physiological and haematological variables. RESULTS: Living in mixed-strain trios did not compromise mouse welfare (assessed via corticosterone metabolite output, stereotypic behaviour, signs of aggression, and other variables). It also did not alter the direction or magnitude of any strain- or enrichment-typical difference across the 26 measured variables, or increase variance in the data: indeed variance was significantly decreased by mixed- strain housing. Furthermore, using Monte Carlo simulations to quantify the statistical power benefits of this approach over a conventional design demonstrated that for our effect sizes, the split- plot design would require significantly fewer mice (under half in most cases) to achieve a power of 80%. CONCLUSIONS: Mixed-strain housing allows several strains to be tested at once, and potentially refines traditional marking practices for research mice. Furthermore, it dramatically illustrates the enhanced statistical power of split-plot designs, allowing many fewer animals to be used. More powerful designs can also increase the chances of replicable findings, and increase the ability of small-scale studies to yield significant results. Using mixed-strain housing for female C57BL/6, DBA/2 and BALB/c mice is therefore an effective, efficient way to promote both refinement and the reduction of animal-use in research.

Play in juvenile mink: litter effects, stability over time, and motivational heterogeneity.

Mink are potentially ideal for investigating the functions of play: deleterious effects of early social isolation suggest a crucial developmental role for play; and huge numbers of highly playful juvenile subjects can be studied on farms. We collected descriptive data on 186 pairs from 93 litters, half provided with play-eliciting environmental enrichment objects in their home cages, to test three hypotheses: (1) play frequency is subject to litter effects; (2) relative playfulness is stable over time; (3) play sub-types share a single, common motivational basis. We found weak litter effects that were driven by stronger litter effects on general activity, and weakly stable individual differences in both total and rough-and-tumble play. Experimentally increasing object play did not inhibit rough-and-tumble play, showing these sub-types are not motivational substitutes. Frequencies of these sub-types were also uncorrelated, and changed differently with time of day and age, further supporting this conclusion.

Effects of environmental enrichment and stereotypic behavior on maternal behavior and infant viability in a model carnivore, the American mink (Neovison vison).

In several species, stress compromises maternal behaviors that are important for infant viability (e.g. licking and grooming). Understanding how stress in captivity affects maternal behavior could therefore be beneficial, especially for carnivores in zoos and breeding centers where infant mortality is often high. We used a model carnivore--American mink--to test two hypotheses, namely that maternal investment and/or behavior is i. improved by environmental enrichment; and ii. compromised by stereotypic behavior. We observed 22 females raised in an indoor facility, 9 enriched, 13 non-enriched. At birth, and at post-natal day 20 when altricial infants were still fully dependent on their mothers, the following offspring variables were recorded: litter size, infant mortality, litter sex ratio (post-natal day 1), and weight. Maternal behavior was assessed by recording nest shape (post-natal day 1), and the frequency of licking and grooming (post-natal days 1-7). Non-enriched females stereotyped more, had female-skewed litters at birth, and tended to make poorer, flatter nests. Maternal licking and grooming showed large, stable individual differences, but appeared unaffected by enrichment. High levels of maternal stereotypic behavior predicted slower offspring growth, replicating previous findings for farmed mink. Nevertheless, enrichment did not significantly increase infant growth rates nor decrease infant mortality. Due to small sample sizes, our study now needs replicating, particularly to explore the potential benefits of enrichment on nest building, sex ratio effects, and the implications of maternal licking and grooming for offspring stress reactivity. Findings could then apply to endangered mustelids like the European mink.

Do housing-induced changes in brain activity cause stereotypic behaviours in laboratory mice?

Abnormal repetitive stereotypic behaviours (SBs) (e.g. pacing, body-rocking) are common in animals with poor welfare (e.g. socially isolated/in barren housing). But how (or even whether) poor housing alters animals' brains to induce SBs remains uncertain. To date, there is little evidence for environmental effects on the brain that also correlate with individual SB performance. Using female mice from two strains (SB-prone DBA/2s; SB-resistant C57/BL/6s), displaying two forms of SB (route-tracing; bar-mouthing), we investigated how housing (conventional laboratory conditions vs. well-resourced 'enriched' cages) affects long-term neuronal activity as assessed via cytochrome oxidase histochemistry in 13 regions of interest (across cortex, striatum, basal ganglia and thalamus). Conventional housing reduced activity in the cortex and striatum. However, DBA mice had no cortical or striatal differences from C57 mice (just greater basal ganglia output activity, independent of housing). Neural correlates for individual levels of bar-mouthing (positive correlations in the substantia nigra and thalamus) were also independent of housing; while route-tracing levels had no clear neural correlates at all. Thus conventional laboratory housing can suppress cortico-striatal activity, but such changes are unrelated to SB (since not mirrored by congruent individual and strain differences). Furthermore, the neural correlates of SB at individual and strain levels seem to reflect underlying predispositions, not housing-mediated changes. To aid further work, hypothesis-generating model fit analyses highlighted this unexplained housing effect, and also suggested several regions of interest across cortex, striatum, thalamus and substantia nigra for future investigation (ideally with improved power to reduce risks of Type II error).

Correction: 'The welfare problems of wide-ranging Carnivora reflect naturally itinerant lifestyles' (2023), by Bandeli et al.

[This corrects the article DOI: 10.1098/rsos.230437.][This corrects the article DOI: 10.1098/rsos.230437.].

The welfare problems of wide-ranging Carnivora reflect naturally itinerant lifestyles.

Carnivora with naturally small home ranges readily adjust to the evolutionarily new environment of captivity, but wider-ranging species seem prone to stress. Understanding why would advance both collection planning and enclosure design. We therefore investigated which aspects of wide-ranging lifestyles are key. We identified eight correlates of home range size (reflecting energetic needs, movement, intra-specific interactions, and itinerant lifestyles). We systematically assessed whether these correlates predict welfare better than range size per se, using data on captive juvenile mortality (from 13 518 individuals across 42 species) and stereotypic route-tracing (456 individuals, 27 species). Naturally itinerant lifestyles (quantified via ratios of daily to annual travel distances) were found to confer risk, predicting greater captive juvenile losses and stereotypic time-budgets. This finding advances our understanding of the evolutionary basis for welfare problems in captive Carnivora, helping explain why naturally sedentary species (e.g. American mink) may breed even in intensive farm conditions, while others (e.g. polar bears, giant pandas) can struggle even in modern zoos and conservation breeding centres. Naturally itinerant lifestyles involve decision-making, and strategic shifts between locations, suggesting that supplying more novelty, cognitive challenge and/or opportunities for control will be effective ways to meet these animals' welfare needs in captivity.

Assessment of Mouse Judgment Bias through an Olfactory Digging Task.

Judgment biases (JB) are differences in the way that individuals in positive and negative affective/emotional states interpret ambiguous information. This phenomenon has long been observed in humans, with individuals in positive states responding to ambiguity 'optimistically' and those in negative states instead showing 'pessimism'. Researchers aiming to assess animal affect have taken advantage of these differential responses, developing tasks to assess judgment bias as an indicator of affective state. These tasks are becoming increasingly popular across diverse species and fields of research. However, for laboratory mice, the most widely used vertebrates in research and a species heavily relied upon to model affective disorders, only one JB task has been successfully validated as sensitive to changes in affective state. Here, we provide a detailed description of this novel murine JB task, and evidence of its sensitivity to mouse affect. Though refinements are still necessary, assessment of mouse JB opens the door for answering both practical questions regarding mouse welfare, and fundamental questions about the impact of affective state in translational research.

Waking inactivity as a welfare indicator in laboratory mice: investigating postures, facial expressions and depression-like states.

Animal welfare assessment relies on valid and practical indicators of affect. In mice, the most widely used research vertebrates, lying still with eyes open, inactive-but-awake (IBA) in the home cage, has potential to be one such indicator. IBA is elevated in barren, conventional housing compared with well-resourced, enriched housing, and predicts immobility in Forced Swim Tests, a common measure of 'helplessness' in depression research. In Experiment 1, using females from three strains (C57BL/6, Balb/c and DBA/2), we first replicated past findings, confirming higher levels of IBA in conventional cages and a positive relationship between IBA and helplessness. We then extended this research to three other signs of depression: changes in weight and sleep, and reduced hippocampal volume. Here, IBA positively covaried with body mass index, with sleep in DBA/2s and conventionally housed BALB/cs, and negatively covaried with hippocampal volume in conventionally housed C57BL/6s. In Experiment 2, we sought to refine the phenotype of IBA to improve its accuracy as a welfare indicator. Here, scoring IBA performed in hunched postures appeared to improve its accuracy as an indicator in Balb/c mice. Additional research is now needed to further refine the phenotype of IBA and to confirm whether it reflects states consistent with depression, or instead other underlying poor welfare conditions.

Social Buffering as a Tool for Improving Rodent Welfare.

The presence of a conspecific can be calming to some species of animal during stress, a phenomenon known as social buffering. For rodents, social buffering can reduce the perception of and reaction to aversive experiences. With a companion, animals may be less frightened in conditioned fear paradigms, experience faster wound healing, show reduced corticosterone responses to novelty, and become more resilient to everyday stressors like cage-cleaning. Social buffering works in diverse ways across species and life stages. For example, social buffering may rely on specific bonds and interactions between individuals, whereas in other cases, the mere presence of conspecific cues may reduce isolation stress. Social buffering has diverse practical applications for enhancing rodent wellbeing (some of which can be immediately applied, while others need further development via welfare-oriented research). Appropriate social housing will generally increase rodents' abilities to cope with challenges, with affiliative cage mates being the most effective buffers. Thus, when rodents are scheduled to experience distressing research procedures, ensuring that their home lives supply high degrees of affiliative, low stress social contact can be an effective refinement. Furthermore, social buffering research illustrates the stress of acute isolation: stressors experienced outside the cage may thus be less impactful if a companion is present. If a companion cannot be provided for subjects exposed to out-of-cage stressors, odors from unstressed animals can help ameliorate stress, as can proxies such as pieces of synthetic fur. Finally, in cases involving conditioned fear (the learned expectation of harm), newly providing social contact during exposure to negative conditioned stimuli (CS) can modify the CS such that for research rodents repeatedly exposed to aversive stimuli, adding conspecific contact can reduce their conditioned fear. Ultimately, these benefits of social buffering should inspire the use of creative techniques to reduce the impact of stressful procedures on laboratory rodents, so enhancing their welfare.

Recommendations for Abnormal Behaviour Ethograms in Monkey Research.

Laboratory monkey ethograms currently include subcategories of abnormal behaviours that are based on superficial morphological similarity. Yet, such ethograms may be misclassifying behaviour, with potential welfare implications as different abnormal behaviours are likely to have distinct risk factors and treatments. We therefore investigated the convergent validity of four hypothesized subcategories of abnormal behaviours ('motor', e.g., pacing; 'self-stimulation', e.g., self-sucking; 'postural', e.g., hanging; and 'self-abuse', e.g., self-biting). This hypothesis predicts positive relationships between the behaviours within each subcategory. Rhesus macaque (Macaca mulatta) data on 19 abnormal behaviours were obtained from indoor-housed animals (n = 1183). Logistic regression models, controlling for sex, age, and the number of observations, revealed that only 1/6 'motor' behaviours positively predicted pacing, while 2/3 'self-abuse' behaviours positively predicted self-biting (one-tailed p-value < 0.05). Furthermore, 'self-stimulation' behaviours did not predict self-sucking, and none of the 'postural' behaviours predicted hanging. Thus, none of the subcategories fully met convergent validity. Subsequently, we created four new valid subcategories formed of comorbid behaviours. The first consisted of self-biting, self-hitting, self-injurious behaviour, floating limb, leg-lifting, and self-clasping. The second comprised twirling, bouncing, rocking, swinging, and hanging. The third comprised pacing and head-twisting, while the final subcategory consisted of flipping and eye-poking. Self-sucking, hair-plucking, threat-biting, and withdrawn remained as individual behaviours. We encourage laboratories to replicate the validation of these subcategories first, and for scientists working with other species to validate their ethograms before using them in welfare assessments.

Where are you from? Female mice raised in enriched or conventional cages differ socially, and can be discriminated by other mice.

Laboratory rodents raised in environmentally-enriched (EE) cages differ behaviourally and cognitively from conventionally-housed (CH) animals. We hypothesised that mice can detect such differences, testing this using differentially-raised female C57BL/6 s as subjects, and differentially-raised female BALB/cs and DBA/2 s as stimuli, in Social Approach Tests. Because more prone to signs of depression, anxiety, stereotypic behaviour (SB) and aggression, we further hypothesised that CH mice would be less sociable and socially attractive than EE mice. A novel familiarisation paradigm pre-exposed subjects to non-cagemate EE and CH stimulus mice before testing in Social Approach Tests. CH subjects proved less sociable than EE subjects: an effect unrelated to general exploration, anxiety or depression-like traits, and driven specifically by reduced interest in CH stimulus mice. Providing further evidence that CH and EE stimulus mice could be distinguished, subjects proved most attracted to mice from housing unlike their own. CH subjects thus preferred EE over CH stimulus mice, while EE subjects tended to prefer CH over EE: patterns that were not mediated by any measured aspect of stimulus mouse behaviour. Differential bodyweight also seemed unimportant, as was scent: soiled CH and EE bedding/nesting did not elicit the same discrimination. Instead, subjects who avoided CH stimulus mice and were attracted to EE stimulus mice were those who received the most agonism in their home cages. Together this provides the first demonstration that mice can distinguish between individuals raised in enriched or conventional cages, and suggests that receiving agonism from cagemates may motivate mice to seek new, less aggressive companions.