Scottish Animal Welfare Commission: Good Food Nation animal welfare indicators

5. Suggested indicators

5.1 Level 1 indicators

High level indicators relating to system

a) numbers of animals kept in cage-free laying hen systems (free-range, barn/aviary/multi-tier, veranda systems)

b) numbers of animals kept in free farrowing systems (outdoor farrowing, indoor pen systems, temporary crate systems)

c) numbers of dairy calves housed in group pens (individual housed for some period, pair housed, group housed, reared with the mother)

d) numbers of broiler chickens kept at stocking densities of 30 kg/m² or less.

Why these indicators?

These indicators can provide the numbers of animals in Scottish production systems which are kept in methods that should provide higher welfare (note that these indicators should be expressed as numbers of animals and not numbers of farms, since some higher welfare farms typically manage smaller numbers of animals, and welfare is experienced at the level of the individual). Although these indicators are at the method of production level, they are supported by extensive research studies which demonstrate the value to the animals of having important resources to allow highly motivated behaviours to be expressed^[18-22]. There is an underlying assumption that welfare will be improved for most animals in these systems, with some opportunities for positive welfare, and that the numbers of animals kept in the most confining systems used in agricultural production in Scotland can be identified.

A recent Defra-funded project (led by University of Reading, but with extensive inputs from Scottish scientists) has developed a methodology to allow Expert Panel assessment of different farming systems^[23] or practices. In this study, free range laying hen systems were rated as providing higher welfare (score 51.3, on a scale from 0 to 100) compared to barn production (43.8) or colony cage production (32.2). Similarly, free farrowing pig production systems were rated more highly (46.6) compared to limited use of crates (34.3) and current use of farrowing crates (26.8). Housing of broiler chickens at stocking densities of 30 kg/m² were rated more highly (46.6) than conventional stocking densities of 39 kg/m² (37.9). This project did not consider dairy calf housing but similar arguments could be made based on research into the impacts of restrictive housing for this species^[24,25].

Advantages:

• Should be relatively easy to assess
• Could be seen to encompass a broad view of meeting Five Domains and allowing opportunities for some positive welfare
• Would allow baseline data to be determined - movement away from caged and confined systems would indicate a positive shift in welfare
• Data collection of numbers of animals in different systems would be a valuable resource for research and policy

Disadvantages:

• Assesses risks and opportunity for welfare not animal’s actual experience
• Only focused on 3 farmed species/life stages (poultry, farrowing sows, dairy calves) – however these are the most likely animals to be kept in confinement so does allow this important welfare metric to be scored

Indicators associated with membership of an accreditation scheme

(e.g. Red Tractor, RSPCA Assured, Soil Association) or high welfare retailer scheme (e.g. Waitrose, Marks & Spencer):

a) Number of farms accredited by different bodies

b) Inspection reports and/or APHA data

Why these indicators?

Although most accreditation bodies focus on resource-based measures, depending on the accreditation body, they specify animal welfare conditions that can be higher than legislative requirements (for example stocking density, provision of enrichment items etc), and in some cases have animal-based measures that provide information on individual animal experience (e.g. rates of lameness). These measures may also have value in providing information (e.g. numbers of farms accredited) on animal welfare for animal-sourced foods sourced from outside Scotland and in some cases outside the UK. Where higher welfare standards are specified by accreditation bodies, these are usually supported by research which has demonstrated that animals have improved welfare when these features are included, hence these indicators have some ability to assess higher welfare, including positive welfare (e.g. enrichment specifications).

Inspection reports, for example, those collected by APHA, will provide some individual animal data. These tend to focus on health conditions primarily, and on negative welfare which we would consider not to be sufficient on their own to fully assess welfare but could serve to provide additional supplementary information, and which is already collected, to complement some of the other suggested measures.

Advantages:

• Data are already collected so making use of what already exists
• Would allow baseline data to be collected and thus for impact of GFN to be tracked in the future

Disadvantages:

• May be data ownership issues and access could be challenging
• May focus primarily on health data
• Likely to need some data integration to be readily used

Indicators that could be collected at a slaughterhouse to provide overview of disease and painful management actions:

a) Castration and tail docking for pigs and sheep

b) Incidence of ecto- and endoparasitism for all species (including fish)

c) Enzootic pneumonia in pigs

d) Liver condemnations (pigs, sheep, cattle)

e) Total number of carcase rejections for welfare related reasons (e.g. bruising)

f) Pododermatitis and bumble foot in poultry

g) Injury and integument alteration (e.g. scratches, lesions, feather condition, hock swellings etc) as relevant for all species

Why these indicators?

These measures can be relatively simple to collect and some of the indicators will already be collected via FSS. They can be assessed at the individual animal level, although may be integrated at flock or herd prevalence level. They do tend to focus on health-related traits and negative welfare indicators, although the ability to, for example, rear pigs successfully with full and clean, unbitten tails could also be indirectly indicative of provision of a high-quality environment. Indicators of carcase bruising reflect negative handling practices (e.g. forcing animals through pen systems such that they collide with pen structures; dragging, hitting or pulling animals by body parts) or impact of transport on slips and falls and collisions with other animals^[26,27,28].

Castration and tail docking are well-known to be painful management activities for which, currently, there is no requirement to provide anaesthetic or analgesia at certain young ages. Reducing the incidence of these practices, as could be assessed by slaughter measures, would reduce negative welfare states experienced on farm, especially if coupled with other indicators that demonstrate an ability to rear these animals without further welfare compromise (such as lesions, bitten tails, pregnant females arriving at slaughter, increased ectoparasite evidence).

Disease conditions such as parasitism (internal and external), respiratory disease, foot and leg problems etc, are typically associated with pain, discomfort, irritation and malaise in livestock. Abattoir measures can provide an overview of the prevalence of these conditions at an individual animal or flock/herd level with can assess the impact of management changes or interventions to reduce incidence. Some of these conditions also provide evidence of poor-quality housing, such as litter or bedding quality, air quality, or stocking density which can be more general welfare indicator.

Advantages:

• Could be integrated into processing of carcases as routine
• Some data already collected which would allow baselines to assessed as well as progress towards improvements (e.g. FSS data)
• Could potentially be at least partially automated in the future

Disadvantages:

• May be data ownership issues and access could be challenging
• May focus primarily on health data
• Likely to need some data integration to be readily used
• Focuses on negative issues only (although successful rearing of pigs with a full tail could be seen as a positive)

High level indicators of humaneness of slaughter processes:

a) Distance travelled to abattoir (e.g. data that Centre of Expertise on Animal Disease Outbreaks (EPIC) collected for SAWC^[29] – not currently routinely collected but it possible)

b) Dead on arrival (DOA) at abattoir (poultry might be most relevant)

c) Non-stun slaughter and stunning methods, mis-stuns – numbers of animals

d) Evidence of catching injuries in poultry

e) Infringements of transport legislation (e.g. transport of pregnant animals or others deemed not fit for transport)

Why these indicators?

Many surveys of consumer welfare concerns have demonstrated that it is this aspect of animal production that people are most concerned about. It is also a highly regulated area of animal welfare, and one where there are already several indicators that are collected e.g. by FSS, and by some high welfare retailers. Transport is well studied as a significant source of stress for animals, and there is good evidence that journey length is an important contributor to poor welfare^[30-33]. Thus, measures that assess the duration of a journey can act as ‘iceberg’ indicators for the welfare of animals when transported to slaughter.

A recent SAWC report^[29] into the welfare issues associated with the provision of slaughterhouses and abattoirs in Scotland demonstrated that journey time could be derived from existing data and highlighted the length and complexity of transport for some animals, including the finding that many animals were not slaughtered at their nearest abattoir. Although these indicators focus on more negative aspects of injury, transport and experience at killing, these are likely to be available metrics and align with areas of consumer concern. Further, when combined with some of the other indicators we have suggested that consider the life of the animal on farm and could include some measures of positive welfare, these indicators allow the assessment of the animal’s quality of life through to the point of slaughter.

Advantages:

• Data relatively accessible although some integration likely to be needed
• Aligns with consumer concerns for animal welfare
• Can be less biased than some measures on farm (as can be recorded in a similar way across different farms)
• Recorded at the animal level

Disadvantages:

• Only focused on negative aspects of welfare
• May be most relevant to pigs and poultry (although distance to slaughter is relevant to all species)

5.2 Level 2 indicators

These are more specific indicators for some species (and includes some positive welfare aspects), which may cover a wider range of the Five Domains, but whether the data on these could be obtained is not currently known. Some suggestions for means by which these data may be collected and obtained are given.

a) Provision of enrichment e.g. number of farms providing brushes in dairy cows/other species; perches for broiler chickens; dark brooders; others – might be low numbers for some but maybe aspirational if people then become aware that these options exist. It is possible this information could be collected via a farm survey, or sales of particular equipment or via Red Tractor/QMS/RSPCA Assured/Other inspections

b) Turnout for dairy cows (number of days) – in some cases this may be part of farm assurance data

c) Bedding provision for growing/finishing pigs – in some cases this may be part of farm assurance data

d) Use of local anaesthetic for procedures and/or use of analgesia on farm (perhaps via medicine records on farm, or Health Planning data)

e) Lameness incidence (all species) – validated gait scoring systems are available for all the major farmed species although may be more frequently collected for some species. Hoof trimmer data for cattle may also be an option. Lameness is common to all farm species and considered a significant welfare issue for most so does focus on a relevant outcome measure. The Agriculture and Horticulture Development Board (AHDB) might hold some of these data and QMS/Scottish pigs

f) Body condition scoring records in cattle, sheep and pigs, avoiding excessive losses of fatness at key points in the production cycle

g) Incidence of keel bone fracture in laying hens

h) Incidence of notifiable and signal diseases in each species

i) On-farm mortality data (including fish) in numbers of animals (not always collected but this might encourage this level of data collection which is extremely important for welfare management and could be a way to ensure there is more oversight); ‘black loss’ is a vague term used in some species but such vagueness does not support active management to reduce losses

j) SSPCA data on welfare complaints (if upheld, or other data relating to these)

Why these indicators?

Environmental enrichment typically improves environmental complexity and is an important solution to providing positive animal welfare opportunities in intensive farming systems^[34,35]. In pigs, for example, there is an increase in exploratory behaviour with enrichment^[36] compared to housing in barren environments. Evidence for improvements in immune function, resilience and brain health measures have been shown for pigs reared in enriched compared to conventional systems^[37-40]. Similarly, there is evidence that use of brushes by dairy cows can lead to improved calving ease^[41], and that use of dark brooders in laying hens reduces incidence of feather pecking^[42,43]. Anecdotal evidence suggests that brushes are widely used in farmed ruminants and appear to be a resource that animals consider to be valuable. Some legislation makes mention of provision of environmental enrichment, although collection of the amount or type of enrichment is not usually required in a standardised way. Some farm assurance bodies do have options to record this (e.g. Red Tractor, RSPCA Assured, Soil Association for pigs), although these data may not be widely disseminated beyond individual assessors or internally. There is some discussion about whether information on provision of enrichment aligns with actual use, which may require further research or other metrics that can assess this animal-based measure of the value of enrichment.

A review of dairy cow welfare^[44], suggested that cows on pasture-based systems had lower levels of lameness, hoof pathologies, hock lesions, mastitis and mortality, and improved indicators of positive welfare (grazing, lying/resting and affiliative social behaviour) compared to continuously housed cows^[45]. A wider range of behaviour is also seen in cows at pasture than when housed. These data may already be captured in farm assurance measures or requirements and held by industry bodies even if not integrated or shared more widely.

Lying on concrete or slatted floors causes discomfort for all species, and can lead to an increase in hock, elbow and joint lesions^[36,46,47]. Provision of mats, straw or other substrates can reduce the potential damage caused by these surfaces. Further, research suggests that animals will actively select particular types of bedding, suggesting they have preferred lying substrates and prefer softer to hard surfaces to allow good quality lying and resting behaviours^[48-50]. Bedding and flooring type and materials may also be recorded as part of farm assurance (e.g. Red Tractor provide advice on suitable bedding materials for cattle and sheep)

Management procedures such as castration, tail docking and disbudding are associated with pain. In most cases there is extensive scientific evidence that the impact of these procedures can be wholly or partially mitigated by provision of anaesthesia and analgesia. In the Defra project^[23] assessing welfare, providing pain relief for castration of lambs led to an improved lifetime welfare score, although no castration also led to further improvements in score. In addition, there are a number of health conditions that can be painful, such as mastitis and lameness, and treatment to reduce pain, as part of other treatments to address the cause of the disease, can increase comfort and welfare. On-farm usage via medicine records is a somewhat blunt instrument to assess this, since incidence will also affect the use of analgesics, but combined with other measures of disease incidence or use of management procedures could be valuable.

Lameness occurs in all terrestrial species, sometimes to a high level (e.g. AHDB figures suggest 30% lameness in dairy cattle), and is associated with pain. Action to reduce lameness and improve foot and leg health across all species would significantly improve welfare. Considerable research into the causes of lameness in each species has been conducted, and methods to mitigate lameness also exist^[36,46]. Improvement in lameness in dairy and beef cattle was included in the Defra project^[23] and resulted in some of the largest changes in rating for welfare (improvements of 13 points for dairy and 5.7 in beef cattle). This was because, to achieve lower rates of lameness (5% in this scenario), several husbandry changes would be required which would provide wider welfare benefits. Thus, assessment of lameness rates can serve as a global indicator of welfare, at least in dairy cows. Industry data (e.g. AHDB) exist for lameness rates for at least some species and could provide baseline and progress data.

Body condition scoring (BCS) is an assessment of the amount of fat and muscle covering the spine and is used as an indicator of absence of prolonged hunger in welfare assessment schemes for cattle and sheep (e.g. Welfare Quality, Animal Welfare Indicators (AWIN)). It is also used as a farm management tool on some farms to ensure provision of feed to animals and to allow more precise management of food resources. Research evidence links low body condition score in sheep with indicators of hunger (willingness to work to access food resources)^[51] and maintaining a good body condition to positive outcomes for lamb and calf survival^[52,53]. AHDB provides advice on carrying out this procedure and scoring templates to collect data. Data on body condition scores may be available from farm assurance records (e.g. RSPCA Assured), APHA records (although often only triggered by complaints which might indicate welfare issues such as low body condition score), and/or AHDB.

Keel bone fracture in laying hens causes pain^[54] and is a significant welfare issue for laying hens who have poor skeletal health through loss of bone calcium associated with high egg production. Keel bone damage also impacts on hen behaviour and can reduce their ability to have positive experiences such as using perches or moving about the house to access food, water and range^[55]. Keel bone damage can occur when hens collide with pen structures and may reflect issues with house design and lack of access to 3-dimensional spaces during rearing in hens^[56]. Keel bone damage can be felt with palpation in live birds or detected at slaughter at end of lay^[57]. Data on keel bone damage may be collected as part of farm assurance for damage in live birds and FSA data for hens at slaughter (end of lay hens are currently not slaughtered in Scotland as there is no abattoir provision for these birds)^[29].

Assessment of the prevalence of notifiable diseases for livestock species (e.g. Bovine TB, sheep scab, bluetongue virus, avian influenza, etc), can provide some insights into the health and welfare of livestock as some of these diseases can cause very severe animal welfare impacts. However, some diseases have never been reported in the UK (e.g. lumpy skin disease in cattle, goat plague) and others have not been reported for more than 25 years (e.g. Aujeszky’s disease)^[58] suggesting these do not provide much indication of current welfare status of animals in Scotland. However, there are a number of non-notifiable disease and endemic diseases in the UK which cause severe animal welfare impacts and may be valuable in understanding welfare impacts. Examples include Bovine viral diarrhoea (BVD), Johne’s disease, Porcine Reproductive and Respiratory Syndrome Virus (PRRSV) in pigs etc. These data may be collected on farm via animal health schemes, APHA inspections and AHDB records.

On farm mortality rates, particularly of young animals, can provide an unambiguous welfare indicator, although these records may not be kept or communicated. Rates of losses of animals, including fish species, at key life stages (e.g. birth, weaning, finishing, involuntary culling of breeding animals, losses when transferring from one stage to another such as hatch to setting in poultry, transferring to sea cages in fish) can provide information on vulnerable time points as well as allow monitoring of improvements in practice over time. A recent SAWC report on fish sentience^[59] also highlighted the need to consider animals as individuals, and not weight of the species, which is important when considering welfare as the individual quality of life experience of animals. In addition, in some species (fish, sheep), unexplained losses are considered as ‘black loss’ and may reflect some acceptance of mortality within the industry. However, loss of animals reflects some degree of animal suffering and poor welfare, and monitoring of losses can provide an ‘iceberg’ indicator for welfare and a benchmark for improvements.

Advantages:

• Many are much more aligned with the Five Domains and provide assessments of positive as well as negative welfare
• Indicators are assessed primarily at the animal level and not at level of resource provision
• May also encourage data collection of these common welfare concerns and opportunities that may indirectly lead to improvements once evidence is more apparent

Disadvantages:

• Much of the data may not be accessible very readily and may not be retrospective (for baseline evidence against which to assess progress)
• Data ownership and sharing issues
• Some issues may require assessor training to achieve reliability (e.g. BCS, lameness)
• Some species more represented than others

5.3 Level 3 indicators

These indicators are more aspirational measures, many of which seek to provide more insights into the overall animal experience, particularly of positive aspects of animal welfare, are suggested by recent research activity and might be possible for future recording.

a) QBA implementation on farm (although not currently widely used except in research there is some commercial interest in developing this approach further, e.g. Waitrose, Ikea)

b) Automated assessment of e.g. food and water use in housed systems, weight change, use of range in free-range poultry, pig emotions in faces, lameness detection in some systems, enrichment use etc.

c) Integrated platforms that collate data from multiple sources to derive a more multi-dimensional welfare index

d) Food labelling that indicates method of production and some indicator of key positive welfare attributes; consumer purchasing of higher welfare products

e) QR codes for system information that can be scanned on product – if interested, the purchaser could go to specific areas

f) Stockpersonship measures – training, certification

Why these indicators?

Qualitative Behaviour Assessment (QBA) is starting to be used more widely as a method of welfare assessment that provides an holistic overview of the animal’s emotional state and can span both positive and negative aspects of welfare assessment within a single measure^[60,61]. There have been various trials of QBA outside of a research context, and the development and testing of an App to collect data is underway, which should allow this to be used more easily on farm. Waitrose have made some movements towards implementing this within their supply chains (e.g. Waitrose Animal Welfare App | Waitrose & Partners), work is underway with the Hong Kong Jockey Club to develop this for use with racehorses and other industry uses are in discussion.

The use of automated and precision livestock farming approaches to welfare monitoring and management have been increasing in the last 10 years or so, with large pan-EU projects assessing the ability of these methods to provide welfare monitoring (Home - ClearFarm and Project – TechCare Project for cattle, pigs, sheep and goats). In addition there are a range of other automated systems, including assessment of food and water intake in intensive systems, automated weighing systems, walk over force plates for assessing lameness, use of machine learning to assess emotional state in pigs faces, Global Navigation Satellite Systems (GNSS) to monitor animal movement, 3D video monitoring of tail posture to predict tail-biting, assessments of air quality, temperature humidity index and bedding moisture, animal noises and vocalisations associated with positive and negative states and optical flow in broiler houses among others. The increased use of AI in monitoring is also providing some more opportunities although there are also ethical concerns raised in some areas^[62]. Although the vast majority of these monitoring and assessing methods are still in development and require external validation^[63,64], there is considerable scope to automate at least some assessments of animals in the future to provide further metrics related to animal welfare.

Animal welfare is multidimensional and experienced at the level of the individual. No single measure will be able to provide a complete overview of welfare state of each animal. Our suggestions for numbers of animals in each production system is an attempt to do this at a fairly crude level based on assumptions that more animals in systems that have been demonstrated to provide higher welfare will result in higher overall welfare, although there may be some individual animals that do still have poor welfare even in higher welfare systems. Integration of a range of different types of measures, through a platform that can weight the different levels of evidence, may provide the most complete view on animal welfare. For instance, the Welfare Footprint Framework^[65] weights different intensities of ‘pleasure’ and ‘pain’ across an animal’s lifetime. This approach requires detailed data on animal experience to be available to ensure some measure of validity, but the overall framework concept is likely very helpful. There are some developments in this area in the EU (see below) although this will likely mean collection of more data than is currently collected, and research to develop a platform that would allow these data to be integrated in a meaningful way.

Food labelling with method of production, or even with QR codes that could allow consumers to select different levels of information or see the conditions under which animals are kept, can be one way to allow producers to be rewarded for maintaining higher welfare standards. These also allow consumers of animal-source foods to express their preferences for higher welfare standards. This is also relevant when considering the welfare of animals consumed in Scotland, rather than only those raised and slaughtered in Scotland. Most of our indicators have focused on the welfare of animals in Scotland, as the place where there is most availability and accessibility of data that can be collected. However, if most food is actually sourced from outside Scotland, then these indicators do not capture the whole picture with respect to welfare state of animal-sourced food consumed in Scotland. This can be particularly an issue for processed foods, where the provenance of ingredients may not be clear and often the lowest welfare standard products are incorporated (e.g. eggs from colony cages, whereas purchase of shell eggs is largely free range). Developing a food labelling scheme, which requires method of production to be made clear, can help consumers to understand the welfare impact of all their animal-sourced food purchasing.

Many studies have clearly emphasised the impact of human behaviour and human interactions on welfare^[66], and this is identified as part of the latest version of the Five Domains Model^[1]. These studies have also shown that training in animal-centred handling and positive welfare can lead to improvements in attitudes and approaches to animal management with benefits for both human well-being and animal welfare. These positive benefits have been captured in the One Welfare framework^[67], and increases in training in animal welfare, handling, and positive attitudes to animals could be a method of improving animal welfare across systems. In some countries in the EU such ‘farmer schools’ or training are recognised as an important part of having ethical farming practices and achieving production goals^[68]. Certification schemes for these attributes may improve on-farm morale and pride in skills and working methods. Cow signals and Pig signals are examples of training approaches that help those working with animals understand and interpret the behaviour of these animals which can improve ease of handling, stockperson safety and animal welfare.

Scotland has long been an important centre for animal welfare science research and development, as the birthplace of the first scientific society focusing on the behaviour and welfare of production animals in 1966. Scotland is also home to some of the pioneering research in this field in the 1980s, such as the Edinburgh pig park that demonstrated that the behavioural needs of production animals had not disappeared just because they had been in housed in conditions that prevented their expression^[69]. This research has grown and developed over the intervening 50 years, contributing to a deep understanding of animal behaviour, welfare assessment, positive welfare, and the development of solutions to common welfare problems. The Scottish Government has been at the forefront of supporting these developments, recognising the importance of animal welfare in good food production systems. Encouraging these developments in knowledge acquisition, co-creation of solutions and dissemination to achieve animal welfare impacts should be seen as an integral part of a Good Food Nation.

Advantages:

• These indicators are much more aligned with contemporary understanding of animal welfare
• Could allow identification of, and progress, in welfare improvements
• Could make Scotland a world leader in welfare assessment and understanding (NB, some EU countries are already exploring some of these approaches, especially automation and integration of measures)

Disadvantages:

• Will need considerable development work
• Many indicators are at the research stage and not yet ready for implementation
• Likely to be significant expense in developing these for general use
• May be push back in some sectors/or for some indicators (e.g. stockperson training in animal welfare)