C1. Absence of prolonged hunger

Hunger is a subjective sensation of appetite and is difficult to assess in non-human animals, but well-designed indicators can suggest probable hunger or satiety (Dawkins Reference Dawkins2006; Clegg et al. Reference Clegg, Borger-Turner and Eskelinen2015). Following the Welfare Quality® model, the indicators proposed to assess this criterion and evaluate both the degree of detail included in dietary records as well as the body condition of the animal, as further discussed below.

The diet of captive chimpanzees has been associated with multiple conditions, including obesity, cardiac, dental, and gastrointestinal disease (Obanda et al. Reference Obanda, Omondi and Chiyo2014), as well as contributing to abnormal behavioural repertoires, including regurgitation and reingestion (Cabana et al. Reference Cabana, Jasmi and Maguire2018). The provision of a suitable diet, as assessed by dietary records, is thus of the utmost importance to captive chimpanzee management and encompasses both appropriate dietary composition and presentation. The nutrient composition of foods consumed by wild chimpanzees is high in fibre and low in water-soluble carbohydrates and lipids (Conklin-Brittain et al. Reference Conklin-Brittain, Wrangham and Hunt1998). Studies have demonstrated that as the proportion of dietary water-soluble carbohydrates decreases, the neutral detergent fibre (NDF) fraction increases, significantly influencing chimpanzee behaviour; inactivity and abnormal behaviour patterns decreased while locomotion, foraging, exploratory and affiliative social behaviours increased (Cabana et al. Reference Cabana, Jasmi and Maguire2018). In addition to diet composition, food presentation is vital to captive chimpanzee management. Wild chimpanzees spend a large proportion of their time (22.5–30.7%) engaged in feeding and foraging behaviour (Hockings et al. Reference Hockings, Anderson and Matsuzawa2009) and providing abundant foraging and feeding opportunities likely represents a source of social and cognitive stimulation, reducing inactivity and the performance of some abnormal behaviours (Celli et al. Reference Celli, Tomonaga, Udono, Teramoto and Nagano2003). Research by Bloomsmith and Lambeth (Reference Bloomsmith and Lambeth1995) has also shown that unpredictable feeding schedules reduced inactivity levels before feeding while feeding whole items may increase feeding time and replicate wild manipulatory behaviours in chimpanzees. A good welfare score would thus be assigned if a chimpanzee’s dietary records reflected the current state of knowledge in this area. While this indicator does not directly assess the animal’s physical condition or feeding behaviour, it evaluates the comprehensiveness and quality of the dietary records maintained by the facility. Importantly, this indicator also serves to assess the appropriateness of the diet offered to the chimpanzees and looks for evidence that the diet provided is nutritionally appropriate and aligned with the natural foraging and feeding behaviours of chimpanzees.

Body Condition Scoring (BCS) is a well-established and widely used welfare tool that can be used to assess animal health. Obesity is a common problem when managing captive chimpanzees (Reamer et al. Reference Reamer, Webb, Jones, Thiele, Haller, Schapiro, Lambeth and Hanley2020), and a study by Obanda et al. (Reference Obanda, Omondi and Chiyo2014) highlighted similarities between humans and chimpanzees concerning obesity and the subsequent risk of developing inflammatory disease. Frequent assessment of body condition should therefore be an integral part of health management in captive chimpanzees, and caretakers should be able to quickly identify and assess overweight or obese individuals. Reamer et al. (Reference Reamer, Webb, Jones, Thiele, Haller, Schapiro, Lambeth and Hanley2020) developed a ten-point scale for non-invasive observational ratings and this has been validated against actual weights and skin fold measurements from sedated chimpanzees.

C2. Absence of prolonged thirst

The continuous availability of water, the number of water-points, and the cleanliness of water-points were considered suitable indicators to assess this criterion, but they reflect the challenge in finding a feasible animal-based indicator (Botreau et al. Reference Botreau, Veissier and Perny2009). Indicators to assess dehydration, such as plasma osmolarity, involve taking a blood sample and thus do not satisfy the requirements for a welfare assessment to be non-invasive, feasible to apply, and undemanding on resources (Jones et al. Reference Jones, Sherwen, Robbins, McLelland and Whittaker2022). Chimpanzees should have easy access to multiple functioning water-points, and the number provided should consider the group size, social dynamics, and habitat design. Water-points should be cleaned regularly and free of crusts of dirt (for example, faeces) and decayed food residue with a good welfare score assigned where water is readily available in clean, functioning water-points to all troop members.

C3. Comfort around resting

Chimpanzees build nests for sleeping by combining tree branches and vegetation (Goodall Reference Goodall1962; Koops et al. Reference Koops, McGrew, Matsuzawa and Knapp2012), and these nests serve various functions, including shelter from climatic conditions, defence against predators, and comfort for sleep or rest (Pruetz et al. Reference Pruetz, Fulton, Marchant, McGrew, Schiel and Waller2008; Stewart & Pruetz Reference Stewart and Pruetz2013). Captive chimpanzees may be provided with materials for nest-building, including hay, woodchips, blankets, and paper (Brent Reference Brent1992), or with intact browse materials for constructing a more typical nest (Videan Reference Videan2006). It is worth mentioning at this point that there is a need to distinguish between a behaviour that is ‘natural’ and one that is ‘motivated’ in an individual as there is limited evidence that ‘natural behaviours’ reliably either lead to or signify better well-being, unless an individual values that behaviour (Yeates Reference Yeates2018). For example, it is not expected that animals enjoy flight, sickness or antagonistic behaviours commonly encountered in ‘natural’ environments (Hermansen et al. Reference Hermansen, Strudsholm and Horsted2004; Miao et al. Reference Miao, Glatz and Ru2004), while animals may find enjoyment in unnatural affiliative interactions with human caregivers (Yeates & Main Reference Yeates and Main2008). Thus, it is impossible to determine a priori what ‘natural behaviours’ improve well-being (Yeates Reference Yeates2010).

Nesting behaviour forms a fundamental and valued part of a chimpanzee’s behavioural repertoire (Fruth & Hohmann Reference Fruth and Hohmann1996), and thus the provision of nesting materials has the potential to impact the welfare of captive chimpanzees. Multiple studies have demonstrated this, including Brent (Reference Brent1992), who found that supplying woodchip bedding to juvenile chimpanzees significantly reduced their abnormal behaviours. More recently, Anderson et al. (Reference Anderson, Amarasekaran and Riba2021) demonstrated that an individual’s welfare was highest when presented with destructible, synthetic nesting materials. It is also worth noting that nest-building ability has been shown to depend on early rearing conditions and that bed-building techniques differ between wild- and captive-born chimpanzees (Bernstein Reference Bernstein1962; Videan Reference Videan2006).

Nests can be built on the ground or at a height, and so the provision of multilevel resting sites is the second indicator chosen to assess comfort around resting in this species. Night nests are typically constructed in a tree (Reynolds & Reynolds Reference Reynolds and Reynolds1965), while day nests are often at ground level (Goodall Reference Goodall1962; Brownlow et al. Reference Brownlow, Plumptre, Reynolds and Ward2001). Samson and Hunt (Reference Samson and Hunt2014) investigated the physical comfort levels of tree and ground nests and reported that ground nests could be advantageous owing to reduced energy expenditure and a homeostatic microclimate. Chimpanzees construct a new nest each night, and a study by Lock and Anderson (Reference Lock and Anderson2013) concluded that night-time spatial arrangements and individual preferences for specific sleeping areas were broadly comparable to nesting patterns reported in free-living chimpanzees. This study suggested that in the interests of captive ape welfare, habitats should include multilevel nesting areas and a choice of multiple sleeping sites.

C4. Thermal comfort

Chimpanzees inhabit various environments, from the semi-arid savanna to the warm and humid tropical forests (Pruetz & Bertolani Reference Pruetz and Bertolani2007), implying behavioural flexibility to adjust to various habitats. Two indicators were developed to assess thermal comfort, adapted from the Welfare Quality® model: availability of shelter and shade and appropriate environmental monitoring records.

Chimpanzee activity budget, terrestriality, and sun exposure have been found to be related to environmental temperatures (Kosheleff & Anderson Reference Kosheleff and Anderson2009). Studies investigating how chimpanzees vary their arboreal versus terrestrial budgets due to environmental temperatures have yielded conflicting results, with Kosheleff and Anderson (Reference Kosheleff and Anderson2009) concluding that chimpanzee terrestriality correlated negatively with temperature, while Takemoto (Reference Takemoto2004) found increased terrestrial activities with higher ambient temperatures. Nevertheless, it is apparent that chimpanzees exhibit behavioural flexibility and adapt to diverse conditions by employing various behavioural strategies for thermoregulation, allowing them to respond to thermal challenges associated with housing them away from their natural environment. Examples of such strategies come from studies such as that by Pruetz and Bertolani (Reference Pruetz and Bertolani2007), who demonstrated that savannah chimpanzees use caves to avoid heat in an exposed environment, while Duncan and Pillay (Reference Duncan and Pillay2013) concluded that chimpanzees in captivity adopted a sun-avoidance strategy, with ultraviolet radiation and humidity levels predicting shade utilisation patterns. Other studies assessing space use and environmental complexity have also demonstrated the tendency of captive apes to spend time close to buildings (Ogden et al. Reference Ogden, Finlay and Maple1990), which may relate to the thermoregulatory benefits of microclimates around these structures (Stoinski et al. Reference Stoinski, Hoff and Maple2001). Thus, the best welfare score for thermal comfort would be assigned to enclosures with multiple shaded and sheltered areas available (for example, climbing structures and platforms), as these areas create a choice of differing microclimates for thermoregulation (Swaisgood & Shepherdson Reference Swaisgood and Shepherdson2005).

In colder countries, captive chimpanzees may have to contend with low environmental temperatures and the provision of heated indoor enclosures and ‘hot spots’ are usually necessary. The highest welfare score for the second Welfare Quality® indicator relating to thermal comfort would evaluate environmental monitoring records and protocols to assess and prevent enclosure temperatures outside of chimpanzees’ Preferred Optimal Temperature Zone, with a remediation process outlined should temperatures fall outside of this range.

C5. Appropriate environment

This was the sole criterion altered from the Welfare Quality® protocol and was adapted to account for chimpanzees’ complex and advanced social structure and intellect. Like Clegg et al. (Reference Clegg, Borger-Turner and Eskelinen2015) in the adaption of Welfare Quality® for captive bottlenose dolphins, this criterion was changed from ‘Ease of movement’ to ‘Appropriate environment’.

There has been an increasing focus on enhancing functionally appropriate captive environments (FACEs) for chimpanzees, such that natural ranging behaviours and other valued species-typical behaviours are promoted (Webb et al. Reference Webb, Hau and Schapiro2018a). Environments inhabited by wild animals often serve as the ‘gold standard’ when judging captive environments, and an abiding critique of captivity is the degree to which it differs to a wild environment, particularly concerning the space available for animals to range and travel. Animal density or space per animal is thus an essential focus of these FACEs, and studies on space per chimpanzee have demonstrated that higher animal densities are related to increased social tension and reduced allogrooming (Aureli & Waal Reference Aureli and Waal2000). Consequently, chimpanzees have been shown to use conflict-avoidance and tension-reduction strategies to cope with reduced space (Aureli & De Waal Reference Aureli and FBM1997). Chimpanzees have also been shown to engage in more abnormal and self-directed behaviours when confined at an increased density (Duncan et al. Reference Duncan, Jones, Lierop and Pillay2013). Thus, the best welfare score for spatial density would be assigned where a behaviourally relevant minimum size threshold was considered. This is supported by literature that suggests that welfare is enhanced when space is increased beyond a suboptimal minimum, with increasing space beyond this leading to smaller and smaller changes in behaviour and increases in welfare (Ross et al. Reference Ross, Wagner, Schapiro, Hau and Lukas2011b; Reamer et al. Reference Reamer, Talbot, Hopper, Mareno, Hall, Brosnan, Lambeth and Schapiro2015).

The topography of zoo enclosures has been associated with animal well-being, with novelty and enclosure complexity regarded as crucial to improving welfare (Jensvold et al. Reference Jensvold, Sanz, Fouts and Fouts2001; Hosey Reference Hosey2005). Such complexity is offered through spatial topography, such as vertical space (Caws et al. Reference Caws, Wehnelt and Aureli2008), available spaces (Bettinger et al. Reference Bettinger, Wallis and Carter1994; Ross et al. Reference Ross, Calcutt, Schapiro and Hau2011a; Herrelko et al. Reference Herrelko, Buchanan-Smith and Vick2015), escape routes (Ross & Lukas Reference Ross and Lukas2006), and opportunities to exhibit control (Videan et al. Reference Videan, Fritz, Schwandt, Smith and Howell2005b; Bloomsmith et al. Reference Bloomsmith, Hasenau and Bohm2017; Webb et al. Reference Webb, Hau and Schapiro2018b). Access to outdoor space is a further indicator purported to influence captive chimpanzee behaviour, as it results in significant volumetric expansion, increased naturalism, and increased enclosure complexity (Ross et al. Reference Ross, Wagner, Schapiro, Hau and Lukas2011b). Thus, the appropriateness of the environment could be assessed by these three indicators, whereby more topographically complex and dynamic enclosures with access to outdoor space achieve better welfare scores.

C6. Absence of injuries

Socially housed chimpanzees are at risk of increased frequency of injury owing to intraspecific aggression, as exists in the wild. Therefore, captive management must balance the social needs of the animals with the desire to curtail wounding owing to natural intraspecific aggression (Williams et al. Reference Williams, Nash, Scarry, Videan and Fritz2010). As agonistic behaviour is a natural part of the chimpanzee’s behavioural repertoire, its elimination from the captive setting is an unrealistic and perhaps undesirable objective. It is also worth noting that social conflict provides cognitive stimulation, variability, and problem-solving opportunities (de Waal et al. Reference de Waal, Nishida, McGrew, Marler, Pickford and de Waal1992). However, captive chimpanzees engage in more aggression than their wild counterparts, and a number of studies have identified factors that influence agonistic encounters and wounding in captivity, including: (1) introductions of new individuals (Alford et al. Reference Alford, Bloomsmith, Keeling and Beck1995); (2) the age and sex of the individual (Baker Reference Baker2000); (3) group composition (Ross et al. Reference Ross, Bloomsmith, Bettinger and Wagner2009); (4) social stability (Bloomsmith et al. Reference Bloomsmith, Lambeth and Alford1992); (5) the presence of in-oestrus females (Bloomsmith et al. Reference Bloomsmith, Lambeth and Alford1992); and (6) the level of human activity (Lambeth et al. Reference Lambeth, Bloomsmith and Alford1997). Management policies can be formulated to reduce the risk of serious wounding by identifying the main factors influencing wounding aggression in captive chimpanzees. The general recommendation that emerges from the literature cited above is that captive management should focus on maintaining long-term, stable social groups and minimise changes in group composition as Williams et al. (Reference Williams, Nash, Scarry, Videan and Fritz2010) demonstrated that wounding aggression decreased in groups with long-term social stability. Two indicators were chosen to assess the ‘Absence of injuries’ criterion, reflecting the main risk of injury to captive individuals, which is intraspecific wounding. The highest welfare score for the resource-based indicators would be assigned to a stable social grouping, while the highest score for the animal-based indicator would be assigned to a chimpanzee with no evidence of wounding.

C7. Absence of disease

Chimpanzees, like humans, can be affected by a wide range of diseases, and the Welfare Quality® indicators chosen to assess this criterion focus on the most prevalent chronic and acute diseases of captive chimpanzees.

The most frequent cause of death in captive chimpanzees has been repeatedly confirmed as cardiovascular disease (Kumar et al. Reference Kumar, Laurence, Owston, Sharp, Williams, Lanford, Hubbard and Dick2017; Raindi et al. Reference Raindi, Rees, Hirschfeld, Wright, Dobbs, Moitti, White, Stahl, Martin, Redrobe, Hughes, Liptovszky, Baiker and Grant2022; Ross et al. Reference Ross, Joshi, Terio and Gamble2022), with a higher percentage of chimpanzees affected by this condition when compared to humans (Seiler et al. Reference Seiler, Dick, Guardado-Mendoza, Vandeberg, Williams, Mubiru and Hubbard2009). Interstitial fibrosing cardiomyopathy is the most common pathology identified, and sudden collapse and death are thought to result from defective conduction or sudden cardiac arrhythmia (Baldessari et al. Reference Baldessari, Snyder, Ahrens and Murnane2013). Whereas humans typically suffer ischaemic myocardial injury due to occlusive vascular disease built up over a lifetime, chimpanzees develop myocardial fibrosis without the vascular component (Varki et al. Reference Varki, Anderson, Herndon, Pham, Gregg, Cheriyan, Murphy, Strobert, Fritz, Else and Varki2009). Various initiatives, such as the Ape Heart Project based at Twycross Zoo and the International Primate Heart Project, have been established to further our understanding of heart disease in great apes, with causal factors remaining poorly understood despite extensive research (Strong et al. Reference Strong, Moittie, Sheppard, Liptovszky, White, Redrobe, Cobb and Baiker2020).

Our understanding of what other diseases affect captive chimpanzees derives from various studies retrospectively examining mortality in captive chimpanzee groups over long periods. Ross et al. (Reference Ross, Joshi, Terio and Gamble2022) analysed the cause of death of 224 chimpanzees at 42 accredited zoos over 25 years and concluded that younger individuals are particularly at risk of trauma and infectious disease, while degenerative conditions continue to be the leading cause of death for adult chimpanzees. Ross et al. (Reference Ross, Joshi, Terio and Gamble2022) examined fifteen years of records relating to the deaths of 268 chimpanzees at a primate facility in New Mexico and determined that gastrointestinal disease was the primary cause of death in 41% of the cases there. Hubbard et al. (Reference Hubbard, Lee and Eichberg1991) and Kumar et al. (Reference Kumar, Laurence, Owston, Sharp, Williams, Lanford, Hubbard and Dick2017) reviewed post mortem examination records for a colony of chimpanzees at The Southwest Biomedical Foundation in Texas and identified cardiovascular, gastrointestinal, and respiratory disease as the leading causes of death in this colony. An additional study conducted by Nunamaker et al. (Reference Nunamaker, Lee and Lammey2012) sought to characterise the prevalence of chronic disease in a colony of sixteen geriatric female chimpanzees and documented a 23% incidence of renal disease. Thus, five animal-based indicators were chosen to assess an ‘Absence of disease’, reflecting the most commonly recognised health concerns of captive chimpanzees: cardiovascular disease, respiratory disease, gastrointestinal disease, musculoskeletal and renal disease. A single resource-based indicator reflects the formulation of a comprehensive preventative healthcare programme, with routine parasitology, regular review of medical histories, daily observations, and opportunistic physical and diagnostic examinations as core components of this.

C8. Absence of pain induced by management procedures

Welfare Quality® includes a criterion for assessing pain associated with management procedures, such as blood draws, injections, and capture for immobilisation. In addition, captive chimpanzees are commonly moved within and among enclosures for feeding, training, veterinary care, or in emergencies where rapid access or separation is needed; thus, the ability to achieve this with voluntary animal participation is critical to animal health, safety, and welfare. The indicator proposed to assess this criterion is whether a behavioural management programme (BMP), such as positive reinforcement training (PRT), is in operation.

Eliciting co-operation from captive animals during necessary routine procedures through PRT is desirable, rather than forcing compliance using restraints or aversive measures (Reinhardt Reference Reinhardt2003). Multiple studies have now been published which have variably focused on training captive chimpanzees to co-operate during routine medical procedures (Videan et al. Reference Videan, Fritz, Murphy, Borman, Smith and Howell2005a), moving cages (Bloomsmith et al. Reference Bloomsmith, Stone and Laule1998), blood and urine collection (Schapiro et al. Reference Schapiro, Bloomsmith and Laule2003; Reamer et al. Reference Reamer, Haller, Thiele, Freeman, Lambeth and Schapiro2014; Bloomsmith et al. Reference Bloomsmith, Neu, Franklin, Griffis and McMillan2015), as well as behavioural management that can reduce social tension during feeding (Bloomsmith et al. Reference Bloomsmith, Laule, Alford and Thurston1994). In addition to the benefits offered to the animal in terms of stress reduction, the involuntary collection of blood from distressed animals can result in blood samples that are not physiologically representative of the animal’s normal hormone levels during stress-free periods (Hassimoto et al. Reference Hassimoto, Harada and Harada2004). A study by Lambeth et al. (Reference Lambeth, Hau, Perlman, Martino and Schapiro2006) also demonstrated that chimpanzees trained to present for an anaesthetic injection had lower white blood cell counts and blood glucose levels, indicating a reduced physiological stress response. Thus, PRT programmes allow animals to participate actively in interactions that improve veterinary care (Fernandez Reference Fernandez2022).

PRT has additionally been shown to benefit animals by influencing social interactions with conspecifics, serving as a form of enrichment, and reducing undesirable behaviours. Bloomsmith et al. (Reference Bloomsmith, Laule, Alford and Thurston1994) effectively reduced aggression during feeding times using PRT in group-housed chimpanzees. Pomerantz and Terkel (Reference Pomerantz and Terkel2009) used PRT to increase chimpanzee affiliative behaviours outside of training sessions, while they also demonstrated reduced abnormal and stress-related behaviours during these PRT sessions. Thus, the highest welfare scores for this criterion would be assigned to facilities with an established and active behavioural management programme in operation.

C9. Expression of social behaviours

Chimpanzees live in large, highly complex, multi-male, multi-female social communities with fission-fusion dynamics (Goodall Reference Goodall1986). Thus, they require social interactions with conspecifics to promote the expression of natural social behaviour. Group dynamics are somewhat artificial in captivity due to behavioural management programmes, operational decisions regarding housing, and the limited scope for escape or avoidance. Therefore, the indicators chosen to assess this criterion included group size and systematically measuring for agonistic and affiliative behaviours, like those exhibited by wild chimpanzees.

Humane standards for the long-term maintenance of captive chimpanzees mandate that they are socially housed, with a multitude of studies demonstrating that aggression and abnormal behaviours are higher in socially impoverished conditions (Spijkerman et al. Reference Spijkerman, Dienske, van Hooff and Jens1994; Williams et al. Reference Williams, Nash, Scarry, Videan and Fritz2010; Khan Reference Khan2013). In terms of group size, the National Institutes of Health (NIH) and the Association of Zoos and Aquariums (AZA) recommend that captive chimpanzees are housed in multi-male, multi-female, age-diverse groups of no less than seven individuals (NIH 2013). Neal Webb et al. (Reference Neal Webb, Hau and Schapiro2019) investigated this recommendation by examining the behavioural effects of group size, group age, and percentage of males on the group and found evidence to support the existing group size and composition recommendations. They also demonstrated in this study that chimpanzees living in more age-diverse social groups, with at least 50% males, showed the highest levels of affiliation.

It has been suggested that behavioural assessment, specifically social interactions, may inform welfare assessment more than physiological indicators (Hill & Broom Reference Hill and Broom2009). The comparison of captive chimpanzee behaviour to that of wild conspecifics is often used as a proxy for animal welfare assessment, although Howell and Cheyne (Reference Howell and Cheyne2019) emphasise that all behaviours are stimulus-driven and that good welfare does not necessarily mean that behavioural budgets must reflect those of wild populations. A better approach may be determining if the animal is displaying affiliative behaviours such as allogrooming and social play, indicative of a positive emotional state (Boissy et al. Reference Boissy, Manteuffel, Jensen, Moe, Spruijt, Keeling, Winckler, Forkman, Dimitrov, Langbein, Bakken, Veissier and Aubert2007). Boissy et al. (Reference Boissy, Manteuffel, Jensen, Moe, Spruijt, Keeling, Winckler, Forkman, Dimitrov, Langbein, Bakken, Veissier and Aubert2007) suggest that such behaviours may help infer good welfare as they often occur only when other physical needs are met. Furthermore, affiliative interactions have been shown to help alleviate stress in primates (Aureli & Yates Reference Aureli and Yates2010), while Yamanashi et al. (Reference Yamanashi, Nogami, Teramoto, Morimura and Hirata2018) demonstrated how adult-adult social play was used to reduce social tension in multi-male groups.

Conversely, there is little evidence to suggest the ‘normal’ frequency of agonistic behaviour among chimpanzees, but multiple studies demonstrate declining welfare states with the expression of agonistic behaviour in high frequencies (Alford et al. Reference Alford, Bloomsmith, Keeling and Beck1995; Videan & Fritz Reference Videan and Fritz2007; Ross et al. Reference Ross, Bloomsmith, Bettinger and Wagner2009; Fultz et al. Reference Fultz, Jackson-Jewett, Higginbotham and Lane2015; Hosey et al. Reference Hosey, Melfi, Formella, Ward, Tokarski, Brunger, Brice and Hill2016). Intra-specific aggression also often leads to injuries and social stress and has therefore been included to assess this criterion in all the adapted zoo animal Welfare Quality® protocols outlined in Table 2. While agonistic behaviour is an expected and natural part of the chimpanzee’s behavioural repertoire, a higher welfare score would be assigned where agonistic behaviour is observed at lower frequencies.

C10. Expression of other behaviours

For captive chimpanzees, ‘abnormal behaviours’ are considered species-atypical behaviours that are observed either exclusively in captivity or at higher rates in captivity than in the wild (Hopper et al. Reference Hopper, Freeman and Ross2016a). Both types are used as benchmarks for evaluating captive chimpanzee welfare, and several stereotypic behaviours have been empirically linked to elevated stress and suboptimal social and physical environments in primates, thus validating them as reliable indicators of poor welfare (Garner Reference Garner2005). A study by Birkett and Newton-Fisher (Reference Birkett and Newton-Fisher2011) examined the behaviour of forty socially housed captive chimpanzees from six collections in the USA and the UK and concluded that abnormal behaviour is endemic in captive chimpanzee populations, despite enrichment efforts. Spijkerman et al. (Reference Spijkerman, Dienske, van Hooff and Jens1994) followed the behavioural development of ninety chimpanzees and demonstrated that body rocking resulted from the inability to cope with frustrating social circumstances, as well as early maternal separation. Baker and Aureli (Reference Baker and Aureli1997) demonstrated that yawning, a self-directed behaviour related to anxiety, increased during periods of social tension in captive chimpanzees, while self-scratching has been observed to increase in chimpanzees in stressful situations, such as immediately after a conflict (Leavens et al. Reference Leavens, Hopkins and Thomas2004; Koski & Sterck Reference Koski and Sterck2007), and to decrease after reconciliation (Fraser et al. Reference Fraser, Stahl and Aureli2008). However, it is worth noting that a clear link between abnormal behaviour and compromised welfare cannot be found in all studies. Recent research on coprophagy (deliberately eating faeces) by both Jacobson et al. (Reference Jacobson, Ross and Bloomsmith2016b) and Hopper et al. (Reference Hopper, Freeman and Ross2016a), suggested an ambiguous relationship between this ‘abnormal behaviour’ and welfare, as it is associated more frequently with desirable conditions, such as mother-rearing. The first Welfare Quality® indicator chosen to assess the ‘expression of other behaviour’ criterion assesses for the presence or absence of ‘abnormal’ behaviours, with the highest welfare scores assigned to animals not displaying species-atypical behaviour at the time of assessment.

As discussed in the Introduction, welfare science has shifted away from the avoidance of negative affective states, towards the promotion of positive ones (Yeates & Main Reference Yeates and Main2008; Mellor Reference Mellor2012), with Boissy et al. (Reference Boissy, Manteuffel, Jensen, Moe, Spruijt, Keeling, Winckler, Forkman, Dimitrov, Langbein, Bakken, Veissier and Aubert2007) even suggesting that the presence of positive affective states may be more relevant to welfare assessment. Therefore, the second indicator chosen to assess this criterion looks for the presence of ‘play and exploratory behaviours’, thus acknowledging the need for new welfare assessment tools to accommodate this shift and include positive welfare indicators where possible. Play behaviour is a valid indicator of positive welfare as it only occurs when all other needs are met (Held & Špinka Reference Held and Špinka2011), while curiosity-driven exploration is a goal-directed behaviour, proposed to be indicative of a positive affective state (Mellor Reference Mellor2015a,Reference Mellorb).

The operation of an environmental enrichment strategy was chosen as the third indicator to assess this criterion. Chimpanzees are highly social and cognitively complex animals, often living in environments that differ substantially from their wild counterparts, and thus appropriate enrichment is required to ensure their psychological well-being (de Groot & Cheyne Reference de Groot and Cheyne2016). The provision of enrichment seeks to encourage behaviours indicative of positive welfare, such as play, affiliation, and exploration (Boissy et al. Reference Boissy, Manteuffel, Jensen, Moe, Spruijt, Keeling, Winckler, Forkman, Dimitrov, Langbein, Bakken, Veissier and Aubert2007) while either reducing or decreasing the occurrence of negative behaviours (Tarou & Bashaw Reference Tarou and Bashaw2007; Wallace et al. Reference Wallace, Altschul, Korfer, Benti, Kaeser, Lambeth, Waller and Slocombe2017). Such enrichment can be sensory (Grunauer & Walguarnery Reference Grunauer and Walguarnery2018; Obsen et al. Reference Obsen, Madsen, Sloth, Gintberg, Jensen, Pertoldi, Pagh and Alstrup2021), tactile (Cerrone Reference Cerrone2019), cognitive (Yamanashi et al. Reference Yamanashi, Matsunaga, Shimada, Kado and Tanaka2016), physical (Hopper et al. Reference Hopper, Shender and Ross2016b), or food-related (Padrell et al. Reference Padrell, Amici, Cordoba, Giberga, Broekman, Almagro and Llorente2021).

C11. Good human-animal relationship

The human-animal relationship (HAR) was outlined by Hosey (Reference Hosey2008) as the history of interactions between the human and the animal, which allows each to make predictions about the behaviour of the other. Negative interactions are generally aversive, unpredictable, and rough (Waiblinger et al. Reference Waiblinger, Boivin, Pedersen, Tosi, Janczak, Visser and Jones2006), and Hosey and Melfi (Reference Hosey and Melfi2015) highlighted the negative impact of such interactions on zoo animal welfare. They demonstrated the negative impact of keepers with deficient handling skills and the stressing consequences of poor human-animal relationships when multiple keepers care for the same animal. Conversely, recent work has shown that positively reinforcing and affiliative relationships can enhance welfare, established through positive reinforcement training, considerate handling, familiarity, and increased interactions (Hosey Reference Hosey2008; Brando Reference Brando2012). Positive interactions involve reassuring verbal and considered physical effort (Hemsworth Reference Hemsworth2003), and a significant body of evidence shows that positive human-animal relationships broadly improve captive chimpanzee welfare (Jensvold Reference Jensvold2008; Chelluri et al. Reference Chelluri, Ross and Wagner2013). Baker (Reference Baker2004) demonstrated how increased human interaction led to increased affiliative encounters and reduced abnormal behaviours in a group of zoo-housed chimpanzees, while multiple authors have proposed that these positive interactions can further serve as social enrichment with positive welfare benefits (Jensvold et al. Reference Jensvold, Buckner and Stadtner2010; Claxton Reference Claxton2011; Chelluri et al. Reference Chelluri, Ross and Wagner2013).

Chimpanzees actively participate in daily husbandry schedules, such as shifting enclosures to facilitate cleaning, and these frequent animal-keeper interactions make HAR assessment particularly pertinent as a welfare metric, with two indicators chosen to assess this criterion for Welfare Quality®. The indicator proposed to assess this criterion is the response of a chimpanzee to a caregiver when not under stimulus control, with the highest welfare scores assigned to a positive response and the lowest welfare score assigned in cases of apparent fear avoidance, stress, or overt human-directed aggression.

C12. Positive emotional state

Increasing focus on the area of positive emotions, in combination with the assertion that captive wild animals should experience good welfare and ‘a good life’ (Mellor Reference Mellor2016), is leading to a greater aspiration among welfare scientists to promote positive experiences and enhance long-term positive emotional states. However, as highlighted by Boissy et al. (Reference Boissy, Manteuffel, Jensen, Moe, Spruijt, Keeling, Winckler, Forkman, Dimitrov, Langbein, Bakken, Veissier and Aubert2007), such positive emotional states are usually more subtle, often less expressive, and more difficult to reliably distinguish when compared to negative ones. Qualitative Behavioural Assessment (QBA) is a scientific method that was initially developed by Wemelsfelder et al. (Reference Wemelsfelder, Hunter, Mendl and Lawrence2000, Reference Wemelsfelder, Hunter, Mendl and Lawrence2001) and has been validated to contribute to the identification of the main dimensions of animal emotional states (Mendl et al. Reference Mendl, Burman and Paul2010; Rutherford et al. Reference Rutherford, Donald, Lawrence and Wemelsfelder2012; Temple et al. Reference Temple, Manteca, Dalmau and Velarde2013; Carreras et al. Reference Carreras, Mainau, Arroyo, Moles, González, Bassols, Dalmau, Faucitano, Manteca and Velarde2016). This tool captures the expressive qualities of animal behaviour by observing an animal’s body language and then scores several behavioural descriptors such as ‘tense’, ‘calm’, ‘playful’, or ‘fearful’. The emotional connotations of these behavioural terms directly relate to animal welfare in that they describe the animals’ experience of their own state (Wemelsfelder et al. Reference Wemelsfelder, Hunter, Mendl and Lawrence2001; Wemelsfelder & Farish Reference Wemelsfelder and Farish2004). Ultimately, this method enables a highly experienced caretaker to capture subtle shifts in behaviour, posture, attitude, expression, or movement and to translate them into quantitative measures that can be analysed statistically.