The 24 / 7 approach to promoting optimal welfare for captive wild animals 1

We have an ethical responsibility to provide captive animals with environments that allow them to experience good welfare. Husbandry activities are often scheduled for the convenience of care staff working within the constraints of the facility, rather than considering the biological and psychological requirements of the animals themselves. The animal welfare 24/7 across the lifespan concept provides a holistic framework to map features of the animal's life cycle, taking into account their natural history, in relation to variations in the captive environment, across day and night, weekdays, weekends, and seasons. In order for animals to have the opportunity to thrive, we argue the need to consider their lifetime experience, integrated into the environments we provide, and with their perspective in mind. Here, we propose a welfare assessment tool based upon 14 criteria, to allow care staff to determine if their animals' welfare needs are met. We conclude that animal habitat management will be enhanced with the use of integrated technologies that provide the animals with more opportunities to engineer their own environments, providing them with complexity, choice and control.

The natural history of an animal, its biology, ecology and diet, sensory systems, 185 natural habitat, social structure, major life history events, activity patterns, and 186 human-animal interactions are among the many topics taken into account when 187 developing species-specific animal welfare programs (see our website 188 247animalwelfare.eu for a worked example with common marmosets). Looking at the 189 life cycle of a species, we find different life stages commonly divided into birth, baby, 190 juvenile, adolescence, reproductive age, senescence and death (see Table 1). When we 191 consider different life stages we can identify key features and considerations likely to 192 be of importance to the welfare of the species. On our website, we provide a table  193 highlighting examples of these features across a wide range of species. To manage a 194 provide animals with an environment that is designed on an understanding of 260 individual differences. An environment that provides sufficient complexity, choice 261 and control will allow animals to thrive within their own capacity, and to develop 262 abilities to cope with the challenges they may face in a captive environment. Early life 263 is a critical stage in this regard. 264 Early experiences affect the brain, given its plasticity when developing (Knudsen, 265 2004 is also some empirical research that illustrates that exposure to some mild or moderate 273 early life stress may provide resilience to subsequent stressors encountered in 274 adulthood (reviewed in Parker and Maestripieri, 2011). It is a fine balance to provide 275 the best early life care to enable the animal to cope as an adult in their future 276 environment. 277

The role of animal 278
Animals do not choose the role they get assigned in a zoo, people do this for them. 279 Some of the common roles animals are assigned to fulfil are: exhibit animal, petting 280 zoo/touch pool animal, ambassador, interaction (e.g. photo opportunity) and show 281 animal, research animal, breeder, and reintroduction candidate, but this list is by no 282 means exhaustive. Roles may require animals to move between zoos regularly, as 283 they are important to the genetic diversity of the captive populations. Roles may 284 require a predominately hands-off approach when animals are raised in environments 285 that prepare them for successful release, such as the "mother condor puppet" feeding 286 method for California condor chicks (Kasielke, 2007, p.151). Alternatively they may 287 receive anti-predator training, as is used with the greater rhea, to improve introduction 288 success (de Azevedo et al., 2012). Other roles take animals into human arms, when 289 children and adults in petting zoos cradle rabbits, or dolphins are kissed in interactive 290 programs. It is important to emphasize that animals can have multiple roles 291 (simultaneously and/or consecutively), such as the bottlenose dolphin being part of a 292 breeding program, a participant in research projects and an interaction/show animal. 293 The impacts of cumulative stress on animals that are assigned multiple roles need to 294 be considered. 295 To illustrate the impact of the role assigned, we shall use an example of domestic rats. 296 Rats are often bred in zoos as food for other animals, and mostly housed in simple and 297 small cages. The same rats can also be display animals in zoos however, housed in 298 larger and complex exhibits with reversed light-cycle, showcasing the highly 299 adaptable, curious, active and social nature of the animal. These same rats can 300 become food for other animals in the zoo when they are old and not suitable to be on 301 display anymore. The role the animal is assigned by humans will affect his/her quality  The purpose for outreach and interactive programs through touch pools and other 307 education programs with hands-on experiences may be laudable. It is important to 308 educate children that, for example, snakes are not slimy, but should be admired and 309 protected. There is also evidence that keeping animals in the classroom improves the 310 children's learning outcomes, for example using more science facts and vocabulary in 311 student writing (Trainin et al., 2005). It also positively affects the cognitive and/or 312  presence of a retreat area, which the public cannot enter, gives the animals some 353 control over their interactions with visitors and may have ameliorated some negative 354 effects. Majchrzak et al. (2015) found that performing rides did not increase cortisol 355 levels in camels, claiming the rides were more akin to environmental enrichment. 356 However no behavioural data were collected to help interpret the findings. It is widely 357 accepted that the use of cortisol as a sole measure does not provide enough 358 information to understand animal welfare and caution against using a one-method 359 approach (e.g. Novak et al., 2013). There are many reasons that glucocorticoid 360 measurements may increase including but not limited to seasonal variation in sex 361 hormones, activity levels, and/or the stress response. A multi-method integrated 362 approach is fundamental for a holistic understanding of animal welfare (Mason and 363 Mendl, 1993). 364 Public presentations and shows are popular with the public and still commonplace. 365 Whilst the chimpanzee tea party is an event of the past in modern zoos, displays with 366 birds of prey, parrots and small mammals, as well as various marine mammal shows 367 still take place. Although such displays can be educational, highlighting and showing 368 skills and adaptations, they also raise specific welfare concerns (Brando, 2016b). 369 Shows can often attract large and noisy crowds, and the method of showing animals 370 has the potential for negative impact. 371 For example, with birds of prey, frequently used old-fashioned and classical falconry 372 methods require that the diet is restricted to ensure the birds are hungry and "work" 373 during the shows and return to the handler (Ford, 1992). Bird of prey training 374 focusing on positive reinforcement is on the rise and weight management to the 375 detriment of the bird is recommended against (IAATE, 2008). Birds of prey may be 376 hooded (eyes covered to limit sensory input) to prevent them becoming distracted or 377 frightened (Ford, 1992), and parrots might be feather clipped to prevent them from 378 flying away. Housing may be designed to make interactions easy and quick with high 379 visitor visibility, and birds of prey are often tethered to perches limiting movement, or 380 held in small crates and holding areas between presentations and shows. Whilst many 381 birds appear to adapt as long periods of being stationary and without food intake may 382 be part of their natural history, food restriction, tethering and hooding, may have 383 negative welfare consequences. Scientific research on these topics is currently 384 lacking. Their participation in shows may disrupt their desired activity patterns, but 385 often provides the only opportunity for the animals to fly freely. Restricted housing 386 for easy access and handling is not limited to birds of prey, but is also common in 387 other species used in education programs, including small mammals such as rats and 388 armadillos. We will illustrate this with an example of animal feeding repertoires. Many species 500 housed in zoos would naturally be feeding, or engaging in their natural feeding 501 repertoire, at times that zoo care professionals are not usually at work. In the wild, 502 some animals start feeding very early, other species feed when the sun goes down or 503 at night, while other species might forage throughout a 24-hour timeframe. The 504 feeding repertoire may include many facets like feelings of hunger, anticipating food, 505 gathering, manipulation, hiding and recovering food, to digestive behaviours and 506 processes. Animals could be eating throughout the day and every day, or gorge feed 507 with meals spread weeks apart, and this in turn can be seasonally dependent. Feeding 508 can be a solitary or social event, and can have social functions to strengthen bonds or 509 maintain hierarchies. These are only a few of the many considerations regarding an 510 animal's feeding repertoire. 511 Due to caregiver routines and working hours, species-specific and appropriate feeding 512 presentation might be different to what is preferred or necessary for the species in  Bloomsmith and Lambeth (1995) studied the response of chimpanzees to predictable 532 and unpredictable feeding schedules and found more abnormal behaviour and 533 inactivity in the pre-feeding period in the predictable schedule due to anticipation. 534 Feeding enrichment designed to increase temporal variability of feeding times 535 (change from feeding at set times), and to increase the number of feeding times/day is 536 proposed by Swaisgood and Shepherdson (2005). provide the necessary information for suitable conditions for indoor-outdoor 554 environments, rather than blindly following traditions or species guidelines (i.e. input 555 parameters described in guidelines have not all been empirically tested using animal-556 based parameters). 557

Variations between day and night 558
Although the use of night cameras is increasing, it is remarkable how little practical Depending on geographical location, many seasonal variations exist, such as daylight 625 hours and climatic variations, with the changing seasons also affecting enclosure 626 quality, quantity and outdoor access. For example, in countries in northern Europe 627 (e.g. Finland) days can be very long, even having 24-hour daylight in summer, while 628 winter days are shortened to a few daylight hours. If zoos operate with the closed 629 access policy (i.e. animals do not have free access between indoor and outdoor areas 630 whenever they want to), the length of opening hours will have an effect on the 631 animals opportunity to choose where they want to be. There is evidence that a 632 combination of indoor and outdoor housing does improve welfare for a variety of 633 primate species, and for some outdoor access is of particular importance (e.g. common marmosets; Pines et al., 2007). Outdoor facilities should provide shelter, 635 heating and protection to allow access in inclement weather. Particular care must be 636 taken for smaller monkey species like the Callitrichidae, or carried infants, 637 particularly those unable to return inside of their own volition when they get cold. 638 Geographical area and associated weather also influences the amount of time that 639 animals can spend outdoors, and may impact on adaptations for seasonal breeding. Tailored programs to suit an individual facility will assist in providing the most 653 optimal environment to promote animal welfare. 654

Zoo visitors 655
Visitor numbers the type of audience, behaviour and noise levels varies across the 656 seasons, with spring and autumn seeing more people at the weekends, while school

Animal welfare assessments: A practical evaluative framework 682
Within the constraints of attempting to take the animal's perspective (but supporting 683 critical anthropomorphism, Burghardt, 1985), understanding individual differences 684 and seasonal changes, we developed an assessment criteria framework as a tool to 685 assist care staff consider the experiences of animals, through a thorough 686 understanding of natural history and biological adaptations. In this section we 687 compare our framework with others available, and introduce a practical online tool. 688

Animal Welfare Frameworks 689
The first welfare framework was developed for farm animals (Brambell, 1965). Farm 690 animals make up the majority of kept animals (e.g. Fraser and MacRae, 2011) and the 691 early attempt to promote welfare was encapsulated in the Five Freedoms (Brambell, 692 1965). More recently these have been extended to a Five Domains model (Mellor,693 2016) and a 12 welfare criteria (Welfare Quality, 2009). We have combined these two 694 approaches, and adapted and extended the 12 welfare criteria to assess any captive 695 animal in our 24/7 approach (see below). We focus here on zoo-housed animals, but 696 the approach is applicable to any captive animal. 697 Other zoo welfare frameworks have also been developed recently. We recognise that 698 efforts to promote optimal welfare in zoos must be comprehensive, coordinated and 699 there must be commitment at all levels for it to be successful at institution level. any animal, and their use will be helpful in the development of an action plan for 726 improvement, and to set priorities. We fully support both of their evidence-based 727 approaches. What the 24/7 approach adds is the need to fully research the natural 728 history and adaptations as a starting point prior to any assessment, and to consider 729 welfare across the life-course and the impact of variations described in this paper. 730 For 24/7, we propose that welfare assessments should be planned at specific times, at 731 important changes and/or transitions (e.g. particular requirements such as shelter, 732 heating or cooling might be necessary with a change of season). A change in the care 733 staff is another example. When someone who has cared for certain individuals for 734 many years is retiring, early planning and continued animal welfare assessment can 735 help identify, prevent and address possible negative impacts. Assessment is also 736 needed when an animal is moved in or out of the group (through death or transfer). 737 Revisiting the workshop questions at pre-determined times would increase the 738 likelihood that an animal's needs and preferences are attended too as much as 739 possible.  To develop these high standards and professional animal care programs, it is critical 749 to understand an animal's ecology, behaviour, biology, sensory systems, social life, 750 and nutritional needs. We base our approach on the 12-point welfare assessment 751 criteria framework. In 2004, a large multi-country and multi-institution project in 752 Europe was initiated, named Welfare Quality®, science and society improving animal 753 welfare in the food quality chain (Welfare Quality, 2009). The goal of the project was 754 to develop European standards for on-farm welfare assessment, product information 755 systems that create transparency about the welfare of farm animals during production, 756 as well as practical strategies for improving animal welfare. The project took five 757 years to complete and identified four key principles: good feeding, good housing, 758 good health and appropriate behaviour. The principles incorporate, expand and update 759 the Five Freedoms. Within these 4 key principles, 12 animal welfare assessment 760 criteria were defined. 761 We have adapted these 12 welfare assessment criteria to be more relevant to zoo 762 animal welfare. We propose two additional criteria (Table 2: criteria no. 2 on feeding 763 and no. 7 on perceived control) to determine whether welfare needs are met. 764 The orange-winged Amazon parrot was chosen as an example species, to illustrate the 765 adapted animal welfare assessment criteria ( Table 2). The reason for choosing this 766 species is that there has been a considerable amount of research done on different 767 aspects of the behaviour, biology and care in captivity, such as environmental 768 enrichment, feeding, and health, which makes completing the criteria easier and 769 decisions for care evidence-based (Melfi, 2009). This parrot species is a resident 770 breeding bird in South America, Tobago and Trinidad, reaching up to sixty years of 771 age. In the wild, the diet consists primarily of all types of fruits, nuts, seeds, blossoms, 772 leaf buds and berries, feeding solely during the day, early in the morning or late in the husbandry manual for the orange-winged Amazon parrot could not be located at the 777 time of writing and should be professionally developed to ensure a standardized 778 approach to promoting good welfare for orange-winged Amazon parrots (or any other 779 animal species kept in captivity for that matter). 780 As previously noted, we have a dedicated website for the 24/7 welfare across lifespan 787 approach: www.247animalwelfare.eu. Amongst other materials, this website provides 788 a link to an interactive and evaluative workshop using common marmosets as an 789 example. The "Wild versus Captive" workshop provides an evidence base to highlight 790 where there may be a mismatch between the wild and captivity, and so potentially 791 highlight a welfare problem. Rather than caregivers providing the bulk of the care, the environment can be 803 designed to provide for activities and opportunities to the animals to choose. The shift 804 to habitat management should be further developed in zoos, reflecting a 24/7 across 805 lifespan approach. Coe (1989Coe ( , 1996Coe ( , 2003Coe ( , 2011, an experienced landscape and zoo 806 architect, has long designed and written about the importance of environments that 807 strongly contribute to the mental and physical wellbeing of animals, including urban 808 wildlife in zoos. Coe (2009) describes the design of environments that specifically 809 suit particular life stages, such as an exhibit for bachelor groups of gorillas. Coe 810