DAVID J. MELLOR AND NGAIO J. BEAUSOLEIL

Animal Welfare Science and Bioethics Centre, School of Veterinary Science; Massey University, New Zealand

Early in the twentieth century many nutritional, environmental, disease, and other animal-related problems were poorly understood (Webster, 1994; Fraser, 2008; Mellor et al., 2009). This seriously hampered the capacity to manage animals effectively, so that farmers in particular struggled to keep their livestock alive and healthily productive. It was recognized that improving scientific knowledge of the underlying biological processes could lead to solutions. Accordingly, there followed at least seven decades of problem-based research into body mechanisms and their manipulation, especially in farm livestock. This commonly involved a now well-recognized iterative approach to these investigations: identify the problem; clarify the underlying mechanism; and manipulate the mechanism to solve the problem (Mellor et al., 2009). The outcome was the development of numerous strategies for more effectively meeting animals’ basic survival needs related mainly to the adequacy of their nutrition, shelter/shade and protection against disease (Mellor et al., 2009; Fraser et al., 2013). Application of this scientific knowledge mitigated many problems that otherwise would have had major negative functional and production impacts on animals (Mellor et al., 2009). Such problem-based research, extending well beyond farm animals, continues to this day, but during the past 20–30 years this has increasingly included consideration of animal welfare impacts. This was partly because formulation of the Five Freedoms paradigm, which incorporated consideration of nutrition, environment, health, behavioral expression, and some mental experiences (Anonymous, 1993; Webster, 1994, 2005), provided an accessible way of understanding animal welfare (Mellor 2016a,b). Additionally, many of the earlier health- and production-based remedies were then understood to have had previously unrecognized animal welfare benefits (Broom, 1986; Webster, 1994, 2005; Fraser, 2008; Mellor et al., 2009; Fraser et al., 2013).

5.2.2 A problem-based perspective on the Five Freedoms paradigm

After its formulation in 1993/1994 the Five Freedoms paradigm had a profound impact on animal welfare thinking and management (Anonymous, 1993; Webster, 2005; Mellor, 2016a,b). Each freedom represented a specific welfare aim and was aligned with a stated Provision that indicated the practical means of achieving the aim (Webster, 1994, 2005; Mellor, 2016a,b). Together the freedoms and provisions focused attention on avoiding or minimizing negative affective experiences (thirst, hunger, discomfort, pain, fear, and distress) and functional states (malnutrition, injury, disease, and behavioral restriction) that were understood to reflect the presence of potentially serious welfare problems (Anonymous, 1993; Webster, 1994, 2005). Thus, a major aim of the paradigm was to motivate the discovery and application of practical remedies for these welfare problems. The general focus was therefore on problems and the aim was to be free of them, which represented an overall orientation toward animal welfare that was conceptually consistent and usefully informative (Mellor, 2016a). Note, however, that four of the freedoms were described in much more specific terms than this: namely, freedom from: (i) thirst, hunger, and malnutrition; (ii) discomfort and exposure; (iii) pain, injury, and disease; and (iv) fear and distress (Anonymous, 1993; Webster, 1994, 2005). The fifth freedom was expressed more generally and focused on restricted opportunities to engage in normal behaviors (Anonymous, 1993; Webster, 1994).

5.3 Limitations of the Five Freedoms Paradigm

It is important to reiterate that the Five Freedoms paradigm beneficially influenced animal welfare thinking and management for most of the first 20 years after it was formulated (Mellor, 2016a,b). It was understood initially that use of the word ‘freedom’ as part of each welfare aim was not intended to mean that animals should to be kept ‘completely free’ from the stipulated negative affective experiences or states; rather, it was understood that animals should be kept ‘as free as possible’ from them (Anonymous, 1993; Webster, 1994). However, the meaning and usage of the term ‘freedom’ changed over time, being strongly influenced, for example, by individuals or groups who had only a cursory knowledge of animal welfare science or who favored animal rights thinking (Mellor, 2016a). Thus, in some sectors the aim of being ‘free’ of negative affects and states was taken to imply that completely eliminating them was both desirable and possible (Mellor, 2016a). This misconception was reinforced because the particular affects and states referred to in the freedoms were described simply as negative or unpleasant. They were therefore considered to be ‘bad’, and, importantly, no corrective clarification was offered regarding their critical biological roles or significance (Mellor, 2016a).

5.4 Biological Significance of Negative Affective Experiences

Research during the past 15–20 years has expanded the list of negative affects that most mammals are now considered likely to experience (Denton et al., 2009; Mellor et al., 2009; Mellor, 2012a; Mellor and Beausoleil, 2015; McMillan, 2016). Two main types have been distinguished: those that are survival-related and others that are situation-related (Mellor and Beausoleil, 2015; Mellor, 2016a, 2017). They have broadly different sensory origins, biological roles, and impacts on animal welfare.

5.4.1 Survival-related or survival-critical negative affects

Originally referred to as homeostasis-related motivational affects (Mellor, 2015a), then survival-related negative affects (Mellor and Beausoleil, 2015), and finally survival-critical negative affects (Mellor, 2017), they are now considered to include the qualitatively different experiences of breathlessness, thirst, hunger, pain, nausea, dizziness, debility, weakness, and sickness (Fraser and Duncan, 1998; Gregory, 2004; Panksepp, 2005; Denton et al., 2009; Mellor et al., 2009; Mellor, 2012a; Beausoleil and Mellor, 2015; Mellor and Beausoleil, 2015, 2017). Generated mainly by sensory inputs that register imbalances or disruptions in the internal physical/functional state of the animal, these negative affects are essential components of genetically embedded mechanisms which elicit behaviors that are designed to secure the survival of the animals (Fraser and Duncan, 1998; Panksepp, 2005; Denton et al., 2009). The unmistakable negativity of these affects creates a sense of urgency to engage in behaviors that are specific to each affect (Fraser and Duncan 1998; Denton et al., 2009); for example, breathlessness elicits increased respiratory activity, thirst provokes water seeking and drinking, hunger food acquisition, and pain escape or avoidance responses to injury (Gregory, 2004; Denton et al., 2009; Verbeek et al., 2012; Beausoleil and Mellor, 2015; Mellor and Beausoleil, 2015). Moreover, the greater the intensity of the negative affect, the greater is the sense of urgency to engage in the aligned behavior, and vice versa, so that once the behavior achieves the required corrective physical/functional outcome the intensity of the negative affect declines (Fraser and Duncan 1998; Denton et al., 2009).

It is apparent that animals are genetically pre-programmed to experience these affects in order to motivate them to engage in affect-specific behaviors that are essential for their survival. Accordingly, a key welfare aim in managed animals is to keep the intensity of these negative affects at tolerably low levels that nevertheless still motivate the essential life-sustaining behaviors (Mellor, 2015a; Mellor and Beausoleil, 2015). As will be outlined in Section 5.7.3, this has other benefits, especially with regard to impacts on animals’ motivation to engage in rewarding behaviors.

5.4.2 Situation-related negative affects

These negative affects arise from brain processing of sensory inputs that originate from outside the body. Such outward-focused sensory inputs contribute to the animal’s cognitive perceptions of its external circumstances, i.e., its situation (Denton et al., 2009 ; Mellor, 2016a). Specifically, these affects are considered to include frustration, anger, helplessness, loneliness, boredom, depression, anxiety, fear, panic and nervous vigilance (Wemelsfelder, 1997, 2005; King and Rowan, 2005; Panksepp, 2005; Mason and Rushen, 2006; Boissy et al., 2007; Beausoleil et al., 2008; Boissy and Lee, 2014; Mellor, 2015c, 2016a). Animals may experience various combinations of these affects when in impoverished and/or threatening situations.

Examples of impoverished circumstances include indoor or outdoor enclosures distinguished by their invariant or barren features and very limited space, situations which severely restrict opportunities to engage in environment-focused exploration and/or interactive social behaviors; also nutrient-dense feeds are provided as small meals that are consumed rapidly (Mason and Rushen, 2006; Boissy et al., 2007; Mellor, 2015a,b,c; Mellor and Beausoleil, 2015). The development of such affects in severely restricted circumstances has been attributed to the thwarting of genetically pre-programmed or learned motivations to engage in behaviors that animals find rewarding, and/or failure to gain the anticipated rewards (Kirkden and Pajor, 2006; Mason and Rushen, 2006; Spinka and Wemelsfelder, 2011). Rewarding behaviors are considered in more detail in Section 5.5.3.

Situations that are cognitively perceived as threatening include possible or actual attack by predators (including humans), victimization by conspecifics in confined spaces, separation from others that provide protection, overstimulation, or challenging novelty, and hazardous environmental events such as fire, flood, or extreme winds (Mellor, 2015b,c). The negative affects experienced in these situations likely include anxiety, fear, panic, and/or nervous vigilance (Panksepp 2005; Boissy et al., 2007; Beausoleil et al., 2008; Mellor, 2015c).

As decisions about the way animals are managed are responsible for many of these situations, introducing more congenial circumstances is an obvious path to devising remedies. The overall objective is to keep social animals together with congenial others in spacious, appropriately variable, and safe environments, as opposed to keeping them closely confined and isolated in barren and/or threatening environments. This is a well-established foundational principle underlying many environmental enrichment strategies (Young, 2003; Mason and Rushen, 2006; Edgar et al., 2013; Fraser et al., 2013; Mellor and Beausoleil, 2015; Mellor et al., 2015).

5.5 Surviving or Thriving as Animal Management Objectives

Ethically responsible husbandry practices aim at achieving more than merely keeping animals alive; their aim is, or should be, for animals to thrive (Kagan et al., 2015).

5.5.1 Survival – the minimalist approach

Survival-focused husbandry is directed at meeting animals’ basic needs for water/food, shade/shelter, and disease/injury control, which are undoubtedly important, but such a restricted focus cannot deliver overall welfare experiences that are positive (Webster, 1994, 2011; Fraser, 2008; Yeates and Main, 2008; Mellor et al., 2009; Edgar et al., 2013; Mellor, 2016a). This is because the negativity of the aligned survival-related affects is critical for motivating essential life-sustaining behaviors and therefore can at best be neutralized temporarily, not eliminated (Denton et al., 2009; Mellor, 2012a, 2015a; Mellor and Beausoleil, 2015). In other words, even in animals whose nutrition, environment, and health are well managed, the valence of the associated affects ranges from negative to neutral, so that with regard to these affects the overall experience remains negative (Mellor, 2017). Of course, when such management is poor the animals’ overall affective experience can be more intensely negative.

5.5.2 Improving upon mere survival

In this regard, the situation-related negative affects stand in marked contrast to the survival-critical affects. This is because introducing environmental improvements aimed at providing opportunities for animals to engage in rewarding behaviors can replace some negative affects with positive ones, leading to a better overall experiential balance for the situation-related affects (Mellor and Beausoleil, 2015; Mellor, 2015a,b, 2016a,b, 2017). Note that this has been understood for some years based largely on investigations into resource inputs, behavioral outputs, and assessments of animals’ preferences, aversions, and priorities (Young, 2003; Dawkins, 2006; Mason and Rushen, 2006; Kirkden and Pajor, 2006; Yeates and Main, 2008; Fraser and Nicol, 2011; Edgar et al., 2013; Fraser et al., 2013); note also, it is only recently in this context that increasing reference has been made to the specific positive affects the animals may experience (Mellor and Beausoleil, 2015; Mellor et al., 2015; Mellor, 2015a,b, 2016a,b, 2017). A combination of both orientations may guide initiatives that contribute to animals thriving.

5.5.3 Rewarding behaviors that may contribute to animals thriving

Rewarding behaviors are those that reflect ‘positive affective engagement’, key constituents of which are the anticipation, current experience, and/or memory of all positive affects linked to each behavior (Mellor, 2015a). These behaviors are more commonly observed when social animals are kept with congenial others in spacious, stimulus-rich, and safe environments (Mason and Rushen, 2006; Fraser, 2008; Yeates and Main, 2008; Mellor and Beausoleil, 2015; Mellor, 2016b). They include environment-focused activities of exploration and food acquisition (foraging or hunting), and the animal-to-animal interactive activities of bonding and bond affirmation, maternal, paternal, or group care of young, play behavior, and sexual activity (Mellor, 2016a,b). In general terms, the associated positive affects are considered likely to include various forms of comfort, pleasure, interest, confidence, and a sense of being in control (Yeates and Main, 2008; Fraser, 2008; Mellor et al., 2015; Mellor, 2015c). More specifically, they may include feelings of being energized, engaged, affectionately sociable, rewarded maternally, paternally or as a group, nurtured, secure or protected, excitedly joyful, and/or sexually gratified (McMillan, 2005, 2007; Panksepp, 2005; Mellor and Beausoleil, 2015; Mellor, 2015a,b,c).

Understanding this should increase the drive to provide animals with environments that offer such opportunities, especially as this is likely to achieve longer term and more varied beneficial welfare outcomes which would contribute to animals thriving. Of course, as with negative affects, the duration, intensity, and frequency of animals’ experiences of positive affects are likely to vary (Mellor, 2015c).

5.6 Five Domains Model for Animal Welfare Assessment

5.6.1 General background trends in animal welfare understanding

It is apparent that during most of the past 30 years animal welfare science has largely focused on negative states, their recognition and management (Broom, 1986; Webster, 1994, 2011; Gregory, 2004; Mellor et al., 2009; Appleby et al., 2011; Grandin, 2015; Mellor, 2016a). Initially, poor welfare states were identified using predominantly physical, physiological, and clinical indices, characterized as the ‘biological functioning’ orientation, which reflected the then commonly used problem-solving, mechanism-focused approach (see Section 5.2). Subsequently, increasing reference was made to inferred negative subjective experiences or affects that animals may have, an approach characterized as the ‘affective state’ orientation (Fraser et al., 1997; Green and Mellor, 2011; Mellor, 2016a, 2017).

Until about 10 years ago these two orientations appeared to be competing frameworks that gave rise to apparently conflicting conclusions about animal welfare (Fraser et al., 1997), for example, low-stress animals exhibiting high productivity but having poor affective experiences. Now, however, they are widely accepted to be dynamically interactive elements within the body operating as a whole integrated entity (Green and Mellor, 2011; Hemsworth et al., 2015; Mellor, 2016a, 2017). Thus, it is currently understood that the welfare state of an animal refers largely to the affects it may experience, and that these reflect the operation of biological mechanisms linked to physical/functional states within the body and to other mechanisms underlying the brain processing that gives rise to animals’ cognitive perception of their external circumstances (Mellor, 2016a). Finally, during the past 10–15 years attention has increasingly been given to positive welfare states, their recognition and promotion (McMillan, 2005, 2007; Panksepp, 2005; Fraser, 2008; Yeates and Main, 2008; Mellor 2015a,b,c; Mellor and Beausoleil, 2015; Clegg et al., 2017).

5.6.2 Formulation and evolution of the model

When formulated in 1994 (Mellor and Reid, 1994), the Five Domains model was the most comprehensive and systematic means then available to prospectively and retrospectively identify and grade the negative welfare impacts caused by research, teaching, and testing (RTT) procedures conducted on sentient animals (Orlans, 2000; Elzanowski, 2006; Williams et al., 2006). In 1997, use of the model became a mandatory part of the New Zealand institutional animal ethics committee approval system for proposed RTT manipulations (Williams et al., 2006), and its mandatory use by all animal-based science investigators in New Zealand continues to this day.

The original explicit focus of the model was animal welfare compromise. The five domains are: (i) nutrition; (ii) environment; (iii) health; (iv) behavior; and (v) mental state. The first three domains mainly focus on internal imbalances or disturbances, whereas the focus of the fourth domain is on external restrictive confinement or restraint or otherwise unusual space availability, and/or negative impacts of the presence or absence of other animals (Mellor and Reid, 1994; Mellor, 2004a). Once objective evidence is collated by reference to the first four domains, the subjective, emotional, or affective experiences inferred to be associated with those disturbances or restrictions are assigned to the fifth mental domain, and it is these experiences that form the endpoint of the welfare assessment.

After its formulation in 1994 the model was regularly refined and extended to take account of fresh developments in animal welfare science thinking at each stage. Nevertheless, the original model anticipated some later developments in thinking. For example, its structure already reflected the dynamically integrated biological functioning/affective state interactivity that underlies an animal’s overall welfare state expressed in terms of what the animal itself experiences (Green and Mellor, 2011; Hemsworth et al., 2015; Mellor, 2016a, 2017). At that time the focus was entirely on welfare compromise so that the named experiences were negative. The original list was limited to thirst, hunger, pain, anxiety, fear, and distress (Mellor and Reid, 1994).

Subsequently, this list was progressively expanded in order to specify what additional affects may be included under the generic term ‘distress’. This specificity was designed to direct attention to previously unrecognized negative impacts (Mellor et al., 2009; Mellor, 2012b; Mellor and Webster, 2014). As outlined above, the current list totals 19, which have been characterized as survival-critical negative affects and situation-related negative affects (see Sections 5.4.1 and 5.4.2). Clear descriptions of the key features of each of these negative affects have now also been published (Ledger and Mellor, 2018).

Finally, the most extensive revision of the model to date involved including in each of the first four domains internal and external circumstances that give rise to positive affective experiences evaluated in the fifth domain (Mellor and Beausoleil, 2015). Moreover, the distinction between survival-related affects (negative) and situation-related affects (both negative and positive) was first made an explicit part of the model during this revision.

5.6.3 Validating inferences about affect referred to in the model

The discipline of affective neuroscience focuses on the brain processes that underlie aversive and rewarding experiences and their manifestation as specific affects (Panksepp, 2005; Rolls, 2007). Thus, there is now objective neuroscience evidence for cautiously inferring that specific internal states and/or expressed behaviors are suggestive of animals experiencing particular negative or positive affects (see below). Importantly, this evidence successfully challenges accusations of anthropomorphism (Panksepp, 2005). Nevertheless, as the focus of model-based welfare assessment is on specific affects, or groups of affects, and their sources, it is important to consider how well supported the inferences are regarding the presence of each affect (Beausoleil and Mellor, 2017). Confidence in such inferences depends on how well-described the underlying affective neuroscience is, the specificity of any physical/functional indices, and/or the distinctiveness of indicative behaviors, all of which must be evaluated in the context of the animal’s physical, biotic, and social environment (Beausoleil and Mellor, 2017; Mellor, 2017). For further consideration of the assessment of affect (negative and positive) see Chapter 23, this volume.

Survival-critical negative affects

Identification of, and distinguishing between, breathlessness, thirst, hunger, pain, and sickness are well supported by the underlying neuroscience knowledge, physical/functional/clinical indices, and behaviors (Gregory, 1998, 2004; Denton et al., 2009; Mendl et al., 2010; Viňuela-Fernández et al., 2011; Beausoleil and Mellor, 2015; Mellor, 2015a). However, distinguishing between nausea, dizziness, debility, weakness, and sickness is not as easy unless the specific circumstances of the animal and/or specific functional indices provide sufficient justification to identify a particular affect (Gregory, 1998, 2004). If this is not possible, affects might be considered in groups, for example, nausea and dizziness or debility, weakness, and sickness, thereby allowing for less specific, but still informative consideration.

Situation-related negative affects

There are sound neuroscience bases for using indicative behaviors to cautiously distinguish among anxiety, fear, panic, depression, frustration, and anger when evaluated in relation to the animals’ particular circumstances (Panksepp, 1998, 2005, 2006; Gregory, 2004; Panksepp and Zellner, 2004; King and Rowan, 2005; Rolls, 2007; Jones and Boissy, 2011; Panksepp et al., 2011; McMillan, 2016). However, behavioral indices may not enable helplessness, loneliness, and/or boredom to be distinguished as easily (Wemelsfelder, 1997, 2005; Mason and Rushen, 2006; McMillan, 2016), so identifying these affects should be done with greater caution. This caveat is not intended to cast doubt on the existence of these affects; rather, it is to note that identifying or distinguishing between them behaviorally may be difficult.

Situation-related positive affects

Affective neuroscience observations support interpretation of particular behaviors as indicating experiences of ‘positive affective engagement’ (Mellor, 2015a). More specifically, the neuroscience of reward-seeking and the generation of positive affects supports the interpretation that animals will likely have pleasurable experiences when exhibiting the following behaviors (Mellor, 2015b): positively motivated, energized environmental exploration, and food acquisition activities, i.e., which are not motivated by significant negative survival-critical affects (Figs 5.1 and 5.2); bonding and bond affirmation (Fig. 5.3); maternal, paternal, or group care of young (Fig. 5.4); play behavior (Fig. 5.5); and sexual activity (Panksepp, 1993, 2005; Berridge, 1996; Nelson and Panksepp, 1998; Ikemoto and Panksepp, 1999; Numan and Insel, 2003; Burgdorf and Panksepp, 2006; Lim and Young, 2006; Boissy et al., 2007). These largely neuroscience-supported inferences from animals’ behavior generally accord with, and are thereby strengthened by, prior interpretation of predominantly behavior-based investigations of animals’ preferences, aversions, and priorities conducted independently (Wemelsfelder, 1997; Dawkins, 2006; Kirkden and Pajor, 2006; Mason and Rushen, 2006; Fraser and Nicol, 2011).

Stay updated, free articles. Join our Telegram channel

Join