In order to assess animal welfare we must first understand what animal welfare means. The science of animal welfare has been around for about 30 years, but historically it has been very difficult to define welfare and there is no existing single, unified definition of animal welfare. In the past, animal welfare assessments focused on health and other physical measures, but now there is an increasing emphasis on the affective states (emotions) of an animal. However, there is still no common agreement on how these affective states can be measured and as such, this has led to measurable indices being used that are not always addressing the true nature of welfare.
Perceptions of animal welfare are culturally sensitive, globally variable and subject to change through time. Despite these differences, an animal’s own perspective of its welfare remains the same. With regards to zoos, given these challenges in defining what animal welfare is and how to measure it, it can create difficulties in systematically measuring and improving animal welfare across multiple institutions, habitats and species.
Animal welfare refers to the mental/emotional or affective state of the animal. In simple terms, an affective state can be negative, neutral or positive.
A Negative State: Experiences animals try to avoid, for example fear, hunger or pain, i.e. those experiences an animal is motivated to minimise.
A Positive State: Occurs when physical needs – nutritional, behavioural, health and environmental – are met, i.e. those experiences an animal is motivated to maximise.
A Neutral State: Experiences where the animal is neither experiencing positive or negative states.
Good welfare is not just about minimising negative states. It is also about promoting positive states.
Isn’t this all a bit anthropomorphic?
Criticism towards the use of emotional language and measures within welfare assessment is inevitable and anthropomorphism has historically been seen as a dirty word in animal behaviour studies and animal care. However, biologist Gordon Burghardt, defined ‘critical anthropomorphism’ — using our own human intuition, feelings and understanding as a starting point for comprehending animal cognition; that using the sentience of the observer to generate hypotheses about the state of that animal’s welfare in light of scientific knowledge of the species and the environment it occupies – is valid. That in identifying with that animal’s own perception of its environment, one can make informed decisions about its care and welfare.
An animal’s emotional state can be considered as the immediate response to a particular stimuli. For example, excitement.
An animal’s mood can be be considered to be the more long-term state (hours, days or even weeks) that an animal feels. For example, contentment.
The Discrete Emotional Theory believes there are a set of basic emotions that are distinguishable by expression and biological processes. These basic emotions are anger, disgust, fear, happiness, sadness and surprise.
The Dimensional Emotional Theory believes action of a small number of underlying systems can generate or construct many mental categories of feelings, including discrete emotions. The Core Affect is a neurophysiological state that underlies simply feeling good or bad, drowsy or energised and can account for the differing mental categories in the same way as the dimensional theory.
If measuring affective states is hard in humans, is it possible to do in animals? While measuring an animal’s mood or affective state is not easy, there are increasingly more studies demonstrating that it is possible. Firstly we can relate to our studies on human emotional measures. Assuming we are looking for discrete emotions for example, we can study facial expressions and neurophysiological profiles in animals to help determine animal emotions and moods. Or we can use cognitive bias, telomere attrition or hippocampus neurogenesis measures to determine possible emotions within the Core Affect model. But there is no simple one to one relationship between a state and measures and the human model may only be translatable to closely related species.
We can also use Intuitive Inference – certain behaviours are intuitively likely to reflect certain affective states. Particular situations are intuitively likely to generate particular affective states. However again, non-evidence-based inference can be wrong and there is potentially limited use in distantly related taxa.
Finally, we can take a theoretical approach and assume certain behaviours reflect certain affective states. Emotions are elicited by rewarding or harmful stimuli. A rewarding stimuli is anything that an animal will work for and a harmful one is anything that it will work to avoid. Working to avoid means they are in a negative state. These operational definitions are most likely to avoid anthropomorphism criticism.
Animal welfare encompasses both the affective (emotional) and physical state of an animal. However, measuring an emotional state is hard enough in humans, let alone in animals where we cannot ask them how they feel! As a result, prescribed (numerical) measures such as the size of an enclosure, the depth of a pool, the nutritional content of food and the healthcare provided, are often used to assess welfare.
But this does not directly measure the animal’s affective state, but simply the care that is being provided. While this can and does have an important influence on how the animal feels, assessing numerical measures alone means we still cannot be certain when an animal is experiencing positive welfare.
An outcome-based measure is more easily linked to the emotional/affective aspect of animal welfare. Aligning to current welfare frameworks, measures can look at identifying positive goal-directed behaviours from the animals. For example, can a tiger carry out natural and normal (positive) behaviours such as stalking, marking, bathing etc, within its enclosure at all times?
We encourage zoos to move away from using only prescribing numerical or minimum-based standards to measure against, but instead look at outcome-based measures. These outcome-based criteria are measurable in the sense that they are a product of the care provided. These measures also include behavioural studies such as cognitive bias and judgement bias research.
A good assessment will provide objective, unbiased information and seek opportunities for improvement. It will aim to identify conformance with the agreed standard and effectiveness of the facility.
A Collection or Enclosure Assessment is a Systematic, Independent and Documented process for obtaining information about the whole of an animal collection (a collection is considered to be all animals held at a facility) or a specific enclosure, in regards to an animal’s welfare. An assessment provides evidence and objective evaluation to determine the extent to which assessment standard criteria are fulfilled.
Assessing a facility is a valuable management tool which can reveal major opportunities for animal welfare and operational improvement, as well as cost reduction. An assessment can be used:
The majority of animal management practices that impinge on an animal’s welfare can be improved so they no longer have a negative impact. From changing diets and encouraging positive feeding behaviours to providing appropriate social groupings and species-appropriate enrichment, often simple changes can significantly improve an animal’s welfare.
Some management practices however, cannot simply be improved and we consider them unacceptable or non-negotiable practices. If we observe these practices during our welfare assessment process, we recommend that they must cease for animal welfare to improve.