Re-Assessing the Importance of Evidence-Based Inputs for Positive Zoo and Aquarium Animal Welfare Outputs

Abstract

The welfare of animals in zoos has come under increasing scrutiny as public awareness grows around the biological needs of captive species. It is also becoming clear that promoting positive welfare experiences upholds population management and conservation aims. This paper re-evaluates current welfare frameworks in zoological institutions, advocating for evidence-based practices, multi-dimensional welfare metrics, and greater emphasis on species-specific needs, as well as the importance of input-based approaches to assess the welfare of zoo animals. By evaluating the limitations of current welfare practices (e.g., a lack of species-specific assessment protocols or sound husbandry evidence to base measures on) and presenting potential areas for improvement, this paper identifies ways that sound baselines for meaningful zoo animal welfare outputs can be created. Although current welfare policies from large zoo membership organisations stipulate assessment of welfare outputs as key to improving animal welfare standards, such outputs can only be positive if inputs are species-specific and relevant to the animals being housed. Practices such as the use of environmental enrichment (for example) need to be further refined to ensure they provide meaningful outputs (for the individuals) from the inputs that create them. Understanding the animal’s needs to ensure that the goal of enrichment is clear benefits both the animal who is provided with the enrichment and the human caregivers as husbandry and management becomes easier. A focus on welfare outputs is commendable and (especially when considering emotional outputs) is indeed a gold standard to aim for, yet we must not lose sight of striving for improvements to housing, husbandry, and species-specific care. Without such fundamental support from correct inputs, outputs are unlikely to be truly (or meaningfully) positive. Therefore, consistent re-examination of inputs is required to make sure they uphold an individual’s attainment of good welfare.

1. Introduction

The gulf between our knowledge on the ecology and biology of the species we house in zoos and aquariums (hereafter “zoos”) and what we do not understand about their requirements under human care was first highlighted by Melfi [1], who called for the development of evidence-based animal husbandry. We define evidence-based animal care as the scientific basis for zoo animal housing and husbandry, where empirical information on species’ evolutionary history, ecological specifics, and behaviour patterns is embedded into better practice approaches for in-zoo management. Prior to Melfi [1], there had already been calls for further scrutiny of how animal housing was designed to ensure that it was relevant and appropriate for the species being housed [2,3]. Whilst there has been substantial evolution in the care provided to some species, including the writing, publishing, and sharing of better practice guidelines to uphold care at the most appropriate level that is currently understood, for many species, housing, husbandry, and management decisions are lacking in research focus [1,4,5] and, therefore, may be extrapolated from other species or are based on “trial and error” approaches. A lack of reliable evidence (or indeed any evidence at all) can result in zoo animals experiencing poor welfare [6]. The current trend to focus on emotional and psychological outputs in zoo welfare assessment protocols is admirable but shortsighted when correct care is not being provided in the first instance.

This paper, which is the amalgamation of content from two presentations at the 4th International Meeting on Zoo Research, Conservation & Biodiversity at Serengeti-Park Hodenhagen (April 2024) aims to consider where the evidence for good care and good welfare inputs is needed and how this evidence can be collected and then implemented into zoo housing, husbandry, and welfare assessment tools. We discuss how evidence for inputs should be collected in a valid and robust manner to ensure such evidence is ecologically relevant to each species that we are keeping. We consider the importance of deciphering what zoo and aquarium animals need from their care so that welfare-positive interventions, such as environmental enrichment, can be applied in a meaningful way. We provide future concepts for consideration to build our evidence base and knowledge of what welfare inputs are and how to ensure they are applied in the zoo, across taxonomic groups. Ultimately, we hope to encourage the continued re-examination and review of species-specific input-based welfare factors so that welfare outputs accurately relate to the care being provided. We are not advocating for the use of inputs alone, but rather for combining both (validated, species-specific) outputs and (biologically relevant) inputs to achieve the highest standards of care for zoo and aquarium-housed species.

Essential Elements of Zoo and Aquarium Animal Welfare

Welfare is the state of an individual as it attempts to cope with its environment [7]. Therefore, in zoos, welfare should consider the animal’s ability to adapt and respond positively to its environment, with welfare inputs serving as the foundation for positive welfare outputs [8]. Because welfare is multifaceted [9], comprising physical, behavioural, and psychological components (Figure 1), each requires systematic attention to protect an animal’s overall wellbeing and longer-term good quality of life. Zoo welfare frameworks increasingly advocate for evaluations that are both specific and repeatable to facilitate ongoing welfare improvements across various species. Figure 1 shows how an individual’s welfare experience comprises (i) physical attributes of the environment and the animal themselves; (ii) the behaviour patterns that an individual has evolved to gain adaptive benefits from and their choice of what to perform and when; (iii) and the psychological outputs that come from environmental interactions, meeting their preferences and receiving what is needed or wanted at a particular time. Species-specific husbandry and housing inputs should be aligned with these components for all animals housed in the zoo to ensure that the animals have the best opportunity of experiencing positive welfare.

Species-specific inputs, encompassing environmental enrichment, behavioural opportunities, and care frameworks that support personnel in delivering and reviewing such correct inputs, play a critical role in fostering positive mental and emotional states in animals. Proper inputs support the Five Domains Model [10], linking physical provisions and environmental conditions to mental wellbeing and emotional outputs. As validated, robust and reputable ways of addressing the fifth (emotional) domain for multiple non-mammalian species (and indeed many mammals) do not exist, efforts should be centred on what evidence we can gather to improve such inputs to scaffold an individual’s chances of attaining positive outputs. Correct inputs are critical to supporting optimal animal welfare, as they address both the physical and psychological needs of animals, enabling them to exhibit ecologically relevant behaviour patterns. When we care for our animals properly, using evidence for this care that we have confidence in, only then can they develop the capacity to achieve positive mental outputs and express such emotional outputs in a way that is relevant and meaningful to that species.

We know that adaptation to an artificial environment can occur in a relatively short period of time [11], across only a few generations [12], and that the behaviour patterns and morphology of captive animals will then differ from wild counterparts [13,14]. This information should galvanise us to ensure that our species management regimes in the zoo do not cause the loss of adaptive traits that are important for a life in the wild. Given the strong conservation aim of zoos, it should, therefore, be integral to the goal of providing appropriate husbandry inputs to ensure that wild traits are not lost from individuals in ex situ populations. Such evidence-based approaches would further support the involvement of zoo-housed populations in One Plan conservation approaches [15,16].

In the wild, animals encounter various challenges that foster resilience [17], such as competition, predation, and social interactions. Zoo environments must emulate aspects of these challenges to support animals’ natural development and behavioural diversity. We recommend incorporating some aspects of unpredictability in daily husbandry practices, thereby mirroring the natural, variable conditions that help animals build resilience. In order to provide the best possible welfare for zoo animals, it is necessary to have science-based welfare assessment methods and to know what animals need. Welfare assessment methods are needed to identify challenges and monitor improvement strategies. Knowing what animals need is necessary to design optimal facilities for animals and adjust husbandry practices to the specific needs of each animal. In the following sections, we will discuss both aspects (welfare assessments and animals’ needs), emphasising the importance of welfare inputs.

2. The Relevance of Input-Based Factors to Zoo Welfare Assessment

Effective welfare management depends on reliable metrics and validated measures that enable objective evaluation for individuals. Animal welfare metrics can be divided into two categories: output-based measures, also known as animal-based indicators, and input-based factors, which are considered resource-based indicators [18]. Examples of the former are abnormal behaviours; changes in the frequency, duration, or intensity of normal behaviours; disease prevalence and incidence; life span; changes to body condition and physical appearance; fluctuation in physiological parameters; mood and body language inferences from qualitative behavioural assessment (QBA) [19]. Examples of input-based factors are housing a species in an appropriately sized and designed exhibit; provision of ecologically relevant environmental enrichment; providing species-appropriate climatic conditions (such as temperature humidity and lighting) to enable species-typical behavioural performance.

Input-based factors sometimes fail to provide accurate information on welfare of animals because the effect (on welfare) of a given environmental feature will vary across individuals, and welfare is an attribute of the individual animal (and their experiences). Environmental features also interact with one another and their effect on an animal may be difficult to predict. Resource-based factors are more easily standardised and audited, although they do not guarantee good welfare without corresponding positive animal outcomes to justify the relevance of the input. As a result of these limitations, it is often advised that welfare assessment tools rely as much as possible on validated and appropriate output-based approaches [20]. Traditionally, zoos have relied on physical factors for welfare inferences, such as changes in body mass, body condition, or health status [21], but these factors alone are insufficient for a comprehensive welfare assessment. An important aspect of welfare assessment is determining the emotional or mental state of animals, for which output-based indicators, such as QBA, behavioural observations for time activity or positive behavioural diversity, hormone analyses, and cognitive bias testing [22], can provide deeper insights as long as they are validated and replicable within individuals and across zoos. For example, preference testing [23,24] can be used for an individual as long as the preferences being offered represent an actual positive choice for the individual (e.g., they are ecologically relevant to the species) and not “the best of a bad situation”. Whilst the importance of considering the mental component of welfare is described in popular frameworks, such as the Five Domains Model [10], zoos need robust and replicable ways of inferring species-specific psychological outputs if such data are to be used to inform care.

However, output-based measures that are meant to provide direct information on animals’ mental state have several limitations. For example, for some taxa, such as fish or reptiles, understanding specific emotional states may be complex, as behavioural indicators of emotional states are less overt. In such cases, the focus may shift to ensuring that environments meet core needs and allow opportunities for positive affect. Even for mammals and birds, using behavioural or physiological indicators to assess emotional states can be challenging.

Behavioural indicators stand out as the predominant tools in the evaluation of animal welfare [5], mainly because of the non-invasive nature of behavioural observations [25] and their cost-effective application [26]. The most frequently used welfare indicators based on behaviour are those that assess abnormal behaviours and changes in the expression (i.e., frequency, duration, or intensity) of normal behaviours [27]. Behavioural indicators, however, have several limitations, including observer bias and fatigue, as well as lack of training and experience, which can lead to inconsistent data. Temporal variability is another challenge, as animal behaviour can vary significantly throughout the day and across different seasons. The presence of an observer (especially if the observer is known to the animal) can also alter behaviour through direct engagement or anticipation of activities. Training is also required to ensure that scientific methods are consistently and rigorously [28] applied, and interobserver reliability [29] may need to be calculated if multiple people are collecting data. Animal care staff have busy schedules [30], and being “time poor” [31] can be a barrier to engagement with behavioural data collection.

Longitudinal studies and continuous monitoring, as well as advances in technology, such as remote behavioural monitoring and/or automated tracking systems, can help mitigate some of these limitations [32]. For example, remote spatial monitoring can be used to measure enclosure usage, social relationships, and functional locomotion in zoo-housed elephants (Proboscidea) [33,34]. However, there are also technological limitations, including technical issues, calibration errors, and high initial setup costs. Moreover, not all species or behaviours are easily tracked with current technology. Finally, the meaning of any behavioural change in terms of animal welfare is not always straightforward. For example, abnormal repetitive behaviours are widely considered indicators of poor welfare [35], but abnormal repetitive behaviours that have been shown for a long period of time may become “fixed” and, therefore, there is the possibility that an animal currently maintained in a suitable environment may exhibit stereotypic behaviour as a result of a previously unsuitable environment [36]. Similarly, behaviours that are not inherently abnormal but may be expressed in ways that are not species- or context-appropriate can result from past housing and husbandry failings [37]. This can explain the performance of displacement activities [38], for example. Therefore, welfare inferences from such unusual behaviour needs to consider an individual’s previous experiences of inputs. Therefore, it is important to recognise that behavioural indicators may not always accurately reflect affective state, as animals can develop coping mechanisms that mask underlying distress [39,40]. This potential mismatch highlights the importance of ensuring robust input-based welfare provisions rely on comprehensive ethograms of species-typical behaviour patterns, individual animal personal scores, and mixed-method approaches (such as comparisons of space use, physical condition, and health status).

The term “physiological indicators” includes a diverse group of biochemical, neuroendocrine, haematological, cardiovascular, respiratory, immunological, and/or cellular biomarkers [41]. The physiological indicators related to the stress response are particularly interesting for zoo-housed animals and are the most commonly used [42]. Although potentially interesting, their interpretation can be challenging and requires in-depth knowledge of the characteristics of each indicator and biological matrix, individual, species, and context [43].

Partly due to the previously mentioned challenges and limitations of output-based indicators, a crucial aspect of welfare assessment is determining whether it is necessary to evaluate each animal’s emotional state or whether it is sufficient to create conditions that naturally support positive emotional states. This focus on “inputs”—creating environments that foster positive states—is posited as potentially more effective than direct emotional state assessments, as the efforts of zoo staff can be better placed in supporting the implementation of evidence-based animal care rather than trying to work out the mood of each individual animal. This emphasis on input-based approaches does not diminish the importance of behavioural observations, as observing behaviours indicative of comfort, engagement, and positive interactions with the environment can serve as proxies for welfare outputs. In this context, positive welfare is not solely the absence of stress but rather the presence of behaviours that reflect positive engagement with environmental features [44]. This can include social/mutual grooming in primates [45], nest-building activities in birds [46], and exploration and wider environmental engagement [47]. A welfare output may not always be a direct measure of an animal’s emotional state but can instead focus on “natural living” by retaining and encouraging natural abilities. This approach aligns with the World Zoo and Aquarium Animal Welfare Strategy [48], which emphasises supporting animals’ engagement with their environment and their specific ecological roles within the zoo.

3. Considering What Animals Need from Their Care

When caring for animals in zoos, it is essential to ask “what do they truly need?” Beyond basic survival, their wellbeing depends on a complex mix of physical, psychological, and social factors. Good welfare goes beyond just meeting their basic needs; it is grounded in ensuring that animals have experiences to thrive. We acknowledge that making any inferences of welfare for non-human animals is challenging, especially when considering the many non-mammalian species housed in zoos that are likely to be less understood. We believe that robust measurements of outcomes that are the result of how an individual engages with their care are required (calculation of time-activity budgets, space use according to different husbandry or enrichment regimes, patterns of positive change with husbandry or housing is amended, the likelihood of a return to baseline conditions when resources are removed or depleted) to fully appreciate the relevance of inputs as a support for positive welfare. A dyadic approach (presence or absence) is not going to be a valid or robust inference of the suitability of a specific input as a scaffold for positive welfare. Multiple measures will be required, and these will be nuanced by the individual animals themselves, their previous experiences, and their personality. Further work on developing methods for scoring responses to inputs that can be validated by those familiar with the species in the zoo and then shared with animal care teams would help provide data-based evidence for how inputs that are currently being provided can meet an animal’s core needs in the zoo. If we can create key performance indicators for each species in the zoo, for example, how a species behaves when a certain degree of care is provided or how a group dynamic can function when a certain social environment is provided, we may be able to share more ecologically valid protocols for housing and husbandry that fulfil individual animal (and population) needs.

3.1. Understanding How to Provide Correct Inputs

Correct housing and husbandry inputs directly influence animal welfare by addressing both physical needs and psychological demands of animals housed in zoos and aquariums. Positive animal outputs, including reduced stress, higher levels of engagement, and evidence of pleasure states—are closely tied to the quality of inputs provided. For example, proper care reduces chronic stress indicators [49,50], which can be measured via cumulative matrices, such as circulating cortisol concentrations in blood/metabolite concentrations in hair or faeces [51], alongside assessing positive relationships with keepers, (characterised by familiarity and consistent care) [52]. Research shows that suboptimal environments alter the performance of species-typical behaviours [53], causing species to perform unusual, aberrant, or abnormal actions as a potential means of coping [54,55], and thus empowering zoos and aquariums to ensure that housing and husbandry are ecologically relevant for each and every species being maintained. The question that follows is, therefore, “of what are the correct inputs for each species?”

Animals can thrive when allowed to express behaviours that are innate and biologically significant [56]. The expression of some natural behaviours has welfare benefits and can be considered a requisite for good welfare. Indeed, as far back as 1965, the Brambell Report mentioned that a degree of confinement that frustrates most of an animal’s natural behaviour should be seen as inadequate from an animal welfare standpoint [57]. More recently, the relationship between natural behaviour and welfare has been discussed in the context of positive welfare, and it has been argued that positive emotions are closely linked to the performance of some rewarding behaviours, such as play, exploration, and maternal care [58]. Therefore, animal care staff should consider monitoring the performance of behaviours that promote a good physical and mental state.

However, deciding which behaviours are important for each species is difficult as there are differences both across species and across environments. Ideally, we would need an experimental approach to find out whether, for each target species kept in each possible environment, the performance of a given behaviour has welfare benefits or, alternatively, whether the frustration of such behaviour has a harmful effect. This approach is not realistic given the sheer diversity of zoo-kept species and environments. Therefore, we suggest a more practical approach based on the assumption that, unless otherwise proven, behaviours that have been found to be important for several species, including animals living in farm and laboratory environments, are likely to also be important to the welfare of zoo-kept animals of some species in some situations that are similar. Following this assumption, exploration, foraging behaviour, and social interaction with conspecifics, among others, could be labelled as being relevant for welfare [59]. This approach does not mean that detailed knowledge of the natural behaviour of the target species is not important, as each species has its own natural foraging habits and social settings, for example, and understanding these (how they provide adaptive benefits, for example) is necessary to provide opportunities for the expression of important behaviours.

Providing environments where animals can make choices and control aspects of their daily routines enhances their sense of agency [44,60,61]. Effective enrichment strategies, such as cognitive challenges and foraging-related tasks, are inherently rewarding [17], as demonstrated by research on diverse taxa from primates [62] to teleost fish [63]. Problem-solving activities, for instance, yield positive emotional responses across species [17] and an enrichment plan that works for one species could easily be adapted and modified to work with another [64], although care is needed with generalisations and individual measurements of engagement with enrichment should be performed [65]. However, understanding the intended outcomes of enrichment is essential before its design. “What is the desired goal or function of the enrichment?” should be the question that is posed before the enrichment is created, implemented, evaluated, and re-deployed. Such “goal-orientated enrichment” focusses on the outputs from the enrichment, enhancing positive behavioural diversity or opportunities for specific behaviour patterns seen in the wild (but not always in zoos). This approach has been picked up and developed by zoos recently [66] to provide further opportunities for enriched experiences [67]. Figure 2 outlines the process of providing an enriched opportunity that is grounded in ecologically relevant inputs to produce enriched behavioural outputs.

Figure 2 illustrates that a structured, evidence-based approach to enrichment design begins with identifying the behaviour that the animal should be displaying, using in-zoo observations to highlight potential behavioural deficits. Understanding the role and function of this behaviour in the wild provides essential context, allowing enrichment goals to be informed by the species’ natural history. The next step involves assessing what aspects of the current environment may be lacking, thus limiting the expression of this behaviour. Enrichment strategies can then be designed and implemented to enhance the environment in a way that actively promotes the desired behaviour, with objectives again grounded in wild evidence. Finally, the effectiveness of the enrichment should be assessed by measuring changes in behaviour that are indicative of positive welfare outcomes, thereby providing a robust, welfare-focused evaluation of the intervention. Ultimately, this process leads to the provision of correct inputs that can enhance welfare, to be evaluated against key welfare indicators. Or, in the case of species where little knowledge of welfare outputs is known, this process can be used to generate key welfare indicators.

The quality of the evidence linking a given input with a desired welfare output must be considered. One case study on one population in one facility is a useful starting point but is not, in itself, evidence for how care should be delivered. Data collection across multiple facilities that provides information on how animals respond to their care or interact with the environment created for them are needed in all cases for all zoo-housed species. A case study on a population at one zoo provides a good benchmark for why further investigation may be needed. Descriptive data on how a species is kept can be gained from a husbandry survey, and then individual animal responses to their husbandry can be gleaned from observational data collection at each zoo that houses said species. Replication across seasons, populations, and environments is necessary to accurately judge engagement with key welfare inputs in a valid and meaningful way. Documents such as the “Polar Bear Research Prospectus”, which outlines key research themes and provides a framework for strategic study, from the zoo community in conjunction with the IUCN Bear Specialist Group [68] is an excellent example of what should be created for all captive species to drive a flow of information from the wild into zoos and to highlight the value of zoo-housed individuals to integrated conservation and management actions. Multi-zoo studies, such as that by Wark and Cronin [69], which provided information on how animals respond to captive care to provide a benchmark of what is commonplace for that species, are fundamentally important to how we understand species’ responses to being in ex situ populations. Such research activities need to be rolled out across more species as a way of critiquing inputs, and their impact on outputs (such as behaviour), to understand which aspects of husbandry and management are most appropriate for that species’ behavioural needs.

3.2. Ensuring That Care Is Species-Specific

Species-specific care ensures that the inputs provided align with the physiological and psychological needs of the animal. Tailored inputs prevent the one-size-fits-all approach that can overlook individual and species-level requirements. Although there is good evidence regarding key welfare inputs for some species [70,71], this is not the case for the majority of zoo-kept animals. Therefore, key input-based factors should be generated by (1) using ecological knowledge to improve enclosure design, (2) carrying out systematic literature reviews centred around directed questions, and (3) using indicators for ecologically similar species for which evidence is available. Dismissing the importance of natural behaviour is worrying. Zoo animals must not become domesticated versions of what they have evolved to be. If we do not consider what a species has evolved to do, and we do not or cannot provide for this, our conservation, education, and research missions are disingenuous, and the role of the animal in the zoo is diminished.

What Do Caregivers Need to Do? (So Animals Can Do What They Need to Do)

Advancing welfare standards across all taxa is an essential future focus for all stakeholders engaged in evidence-based care and the development of welfare-focussed husbandry. Zoo welfare standards often disproportionately emphasise mammals and certain avian species, while non-mammalian taxa receive less attention. We point to a need for more inclusive welfare strategies that address the unique behavioural and physiological needs of all species within a zoo. Developing behavioural inventories that identify positive welfare indicators for less-studied taxa, such as amphibians or invertebrates, is essential to create comprehensive welfare guidelines. Considering species-specific requirements in zoos is also essential, for example, generic “bird welfare assessment tools” are not going to capture the information required to advance care for individual species of birds housed in zoos. We believe that zoos should work around the following key areas of development to provide relevant inputs (that will lead to the performance of welfare-positive behaviours) for the animals in their care.

(1)

Developing science-based environmental enrichment programs that are tailored to each species and individual needs will ultimately promote more opportunities for more individuals of more species to perform welfare-positive behaviour patterns. Environmental enrichment plays a critical role in animal welfare by enabling physical fitness, mental stimulation, and behavioural expression. Incorporating environmental complexity not only aids in maintaining physical health but also supports the psychological welfare of the animals. These enrichments should be tailored to each species’ needs and should also consider variability in temperament within the species. For instance, complex climbing structures or sensory stimuli that provoke the use of different sensory modalities may be provided for arboreal primates. Encouraging animals to express their natural behaviours within a zoo setting is a cornerstone of welfare strategy. This includes providing opportunities for behaviours related to feeding, social structure, and territory, which align with a species’ ecological roles. An overall goal must be to support animals in ways that allow them to engage authentically with their surroundings. In this way, the environment itself can become enriching.

(2)

Implementing evidence-based care practices. Building a sound baseline for welfare requires adopting evidence-based practices across all species. Research on species-specific needs and behaviours should guide welfare inputs, enabling a more reliable foundation for assessing welfare outputs. This approach ensures that animals are more likely to experience positive welfare states and, in turn, exhibit behaviours that reflect wellbeing.

(3)

Bridging the gap between proactive and reactive welfare. The zoo industry traditionally relies on reactive welfare measures, addressing issues as they arise rather than pre-emptively providing for welfare needs. We advocate for a shift towards proactive welfare management by establishing welfare inputs that inherently promote positive welfare states. Proactive approaches not only enhance welfare outcomes but also improve consistency in welfare standards across institutions.

4. Future Directions for Improving Evidence Used for Decisions on Zoo Welfare

When reassessing welfare priorities, it is clear that establishing robust welfare inputs tailored to species-specific needs provides a sound basis for meaningful welfare outputs. The process of identifying welfare inputs that have positive effects on welfare outputs must be based on scientific evidence, and we suggest that there are two main methodologies to achieve this. First, longitudinal studies using an epidemiological approach can identify whether the presence or absence of a particular input has an effect on a given welfare output. This sort of study requires a relatively large number of animals that are ideally kept in different environments or exposed to different inputs; therefore, the collaboration of several zoological institutions is required. A good example of this approach is found in the research by Meehan et al. [72], which evaluated the welfare of African (Loxodonta africana) and Asian (Elephas maximus) elephants in 68 accredited North American zoos. These authors wished to identify aspects of a zoo elephant’s physical environment, social life, and management that are most important to good welfare, thus providing practical information on how to prevent and reduce a variety of species-specific welfare problems. Secondly, a comparative approach can identify behavioural priorities that are important for several species across different taxa [73,74,75,76]. This helps identify environmental features that allow for the expression of such species-typical behaviours that enable adaptive benefits to accrue. Welfare inputs must be tailored to the biology of each species, but such a comparative approach at least provides a starting point to formulate hypotheses for testing on which inputs are important for a given species, for which there currently exists a paucity of information. We suggest a stepwise process for generating evidence for inputs, taking relevant methods and approaches from case studies (e.g., a single population in one zoo) and building on the robust qualities of good research design in this case study to expand where these are used (e.g., across facilities and populations). Data review by those with expertise in the care and management of the focal species can help with the re-formulation of data collection methods (e.g., to address any limitations or challenges in research design or validity) and can provide knowledge on how to best apply the findings from larger-scale research to a species’ in-zoo husbandry for welfare benefits to accrue. We emphasise the importance of expanding welfare assessment frameworks to cover all zoo taxa, thereby ensuring comprehensive and inclusive welfare management.

Table 1. Examples of concepts that need addressing in zoos to create species-appropriate inputs as a foundation for positive welfare outputs.

Concept	Reasoning
Species-specific inputs	Welfare assessment templates should be species-specific and institutionally transferable, emphasising the needs and natural behaviours of each animal.
Challenge and resilience	Incorporating environmental challenges helps build resilience and promotes behavioural diversity, aligning in-zoo scenarios more closely with the animal’s natural lifestyle.
Ensuring valid welfare inferences	While emotional state assessment remains challenging in many taxa, prioritising environments that facilitate positive outcomes may offer more practical welfare inferences.
Inclusive welfare frameworks	Expanding welfare assessments to include underrepresented taxa is essential to ensure all zoo animals benefit from review of their care to ensure its relevance.
Combining input- and output-based measures	Both input- and output-based approaches have their strengths and weaknesses, and a combination of both is likely to be most effective. Providing this starts from a solid foundation of species-relevant inputs being provided.
Using knowledge obtained from non-zoo species, including farm and laboratory animals	Welfare assessment templates must be species-specific. However, work on farm and laboratory animals can provide help formulate hypotheses to be tested in the relevant species that are then applied to zoo/aquarium concepts.
Using data from field studies	A sound knowledge of the habitat and natural history of a species is very valuable to understanding their biological needs [78,79]. Zoo/aquarium husbandry and housing can only move forwards when these are aligned fully with evolution and ecology.

provides some suggestions for how this can be realised. Work that has been undertaken in different industries, outside of the zoo, can also be useful in this regard [77] by identifying methods and approaches to data collection and validation that have worked in one scenario and, therefore, could be trialled elsewhere.

Although some taxa predominate in zoo research and welfare activities [4,5], we should broaden our focus to ensure all species can be provided with evidence-based care. We can learn from these “research-popular” species by borrowing methods and approaches to gathering such evidence. Even for popular/familiar mammalian species, there is disparity in what species are commonly studied; this also applies to domesticated species [80]. Asking directed research questions can help to focus scientific efforts, taking methods from well-researched taxa and applying this to other species and situations. For example, identifying areas of husbandry and housing that are lacking and promoting the need for investigation into this can generate research that enables data-based answers. Such examples of directed questions can be species-specific, for example, the study by Rose et al. [81] or husbandry-specific, e.g., that by Ward et al. [82]. We encourage collaborative efforts to address some of the challenges in Table 1, while continuing to evaluate and review the relevance and validity of currently available evidence to ensure it remains fit for purpose. Such collaborative approaches (such as sharing research protocols, welfare assessment databases, or cross-institutional studies), together with a focus on species-specific needs and how these are factored in validated and credible welfare assessment frameworks, would help to reconcile some of the issues noted with the currently popular ways of assessing welfare, such as the limitations surrounding the Five Domains Model [83]. Such limitations centre on a largely qualitative approach alongside subjectivity, inconsistent application, and lack of practical measurement. Attempting to infer the fifth (mental) domain is admirable and the ultimate gold standard. But without a basis in appropriate care, animals in zoos are very unlikely to achieve the positive emotional outputs associated with this. Animal care staff and the personnel in charge of ex situ populations must continue to investigate and evidence what animals need to thrive, and once these are realised, implemented, and evaluated, further consideration can be given to how animals may be feeling and thinking. Comfort can come from feeling safe, secure, and having autonomy over the immediate environment, hence the need for husbandry to align with ecology and evolutionary history. From a practical perspective, encouraging continued research into species’ needs, as well as collaboration with colleagues across zoological organisations to review, evaluate, and re-review husbandry practices, goal-orientated enrichment applications, and the management of a species’ social environment over time and life stages, for example, will generate applied evidence that ensures the relevance of welfare-focused inputs.

5. Conclusions

This paper shows that focusing on appropriate (evidence-based) inputs and species-specific care leads to improved emotional and mental welfare outputs for zoo animals. By aligning provisions with a species’ behavioural needs and evolutionary ecology, we can foster an environment conducive to their thriving rather than merely surviving. This input-based approach is a cornerstone of effective welfare assessment and husbandry improvement that, by its very nature, will lead to a promotion of positive welfare outputs. As mentioned by others in the context of farm animal welfare assessment [84,85], we suggest that a realistic approach to assess the welfare of zoo animals should be based on the recognition that both input- and output-based measures have their own strengths and weaknesses, and that a combination of both is probably the best strategy. But without knowledge of what inputs are appropriate or relevant to a species, we will not see welfare-positive outputs being produced. In this regard, zoos must focus their attention on developing and appraising housing and husbandry to ensure that the care they provide is actually going to care for that particular species. The evidence-based approach is still the cornerstone of evolving how we keep animals, and this is our surest way of ensuring that the physical, psychological, and behavioural elements of welfare are good (Figure 3).

This evidence-based approach can be obtained from intervention studies where a given change in housing or husbandry is implemented and its effects on the welfare of animals is monitored using a combination of several output-based measures. Even small-scale changes, with one population in one zoo, that show a positive result can then be replicated elsewhere to see whether another population of the same species benefits. Alternatively, epidemiological studies performed across several institutions, which identify those input factors associated with changes in output measures, could be implemented. This second approach is particularly promising and has yielded very valuable insights on what animals need (i.e., on which inputs are relevant for a particular species) [72]. A main limitation, however, is that such epidemiological studies require the collaboration of several zoological institutions and can be logistically and financially challenging to implement.

Ensuring robust welfare baselines is essential to achieving meaningful welfare outputs in zoos. This shift towards proactive welfare management, where inputs are carefully crafted to meet species-specific needs, represents a more holistic approach to animal welfare. By fostering natural behaviours and promoting resilience, zoos can better support positive welfare for all their animals, ensuring that welfare outcomes reflect not just the absence of distress or discomfort but the presence of positive, engaging experiences that build to an overall better quality of life. Zoos operate under increasing ethical scrutiny, particularly regarding the conditions of animals in captivity and the relevance of conservation claims. Transparency in welfare practices and the implementation of rigorous welfare standards are essential for gaining public trust and support. Zoos should engage in open dialogue with the public, explaining the rationale behind welfare practices and addressing common concerns. Public education initiatives need to be linked to information on how and why animals are kept, including how good welfare is promoted. By re-assessing welfare priorities, including ensuring that inputs to care are correct and species-relevant, zoos can demonstrate a commitment to ethical animal care that aligns with both scientific standards and public expectations.

This review has reaffirmed the necessity of integrating both input-based factors and output-based indicators into a holistic framework for promoting good zoo animal welfare. While output-based indicators, particularly those focused on behavioural states, are invaluable tools for welfare assessment, their effectiveness is fundamentally dependent on the presence of appropriate, species-specific inputs. Without valid and evidence-based inputs tailored to each species’ evolutionary and ecological needs, even the most rigorous output assessments may fail to reflect an animal’s true welfare state. We are not suggesting that output-based measures are flawed; rather, we argue that they must be interpreted in the context of the care provided. Welfare outcomes are only meaningful if they emerge from inputs that align with what each species requires to thrive under human care. By prioritising robust, scientifically validated inputs (ranging from housing design and husbandry practices to environmental enrichment and the creation of positive life experiences), zoos can better scaffold animals’ opportunities to achieve positive welfare states. Ultimately, using input factors is not without challenges. For example, the identification of those inputs is relevant for each species, and we have discussed methodologies that could be used to achieve this. Additionally, individual animals will differ in how they respond to a particular environmental or husbandry change (i.e., changes in how an input is accessed or provided) and multiple measures of how animals engage with their inputs is required. We advocate for the continued application of an evidence-based approach to zoo animal management, the use of empirical investigation, and a combination of output measures and input factors to underpin all elements of ex situ care and welfare-friendly husbandry for all individuals of a species across all life stages.