Following a request from the Commission, the EFSA Animal Health and Welfare Panel (AHAW) was asked to deliver a scientific opinion on a multifactorial approach on the use of both animal and non-animal-based measures to assess the welfare of pigs especially those welfare parameters regulated in Council Directive 2008/120/EC4, Article 3(5) and Annex I Chapter I numbers 4 and 8 regulating the provision of manipulable material and avoidance of tail-docking.

In order to address the Terms of Reference (ToRs), the Opinion considers separately the issues of avoidance of tail-docking and provision of functional manipulable materials which meet the needs of the animals, although there is significant interaction between these two issues which is incorporated in the response.

To promote the use of functional manipulable materials, consideration focuses on the identification of when such materials meet the behavioural needs of the animals at different stages in life, the suckling piglet, growing pig from weaning to slaughter and breeding sow and boar, and how this might be assessed in a farm situation.

Since tail-docking of neonatal piglets is carried out to reduce the risk of tail-biting in later life, between weaning and slaughter, consideration focuses on the other risk factors for tail-biting in rearing and finishing pigs, and how these might be better managed and controlled in a farm situation to reduce the need for docking.

Three Terms of Reference

The first term of reference (ToR) is to identify the multiple interactions between risk factors, welfare consequences and animal-based and non-animal-based measures.

A summary of available information in the scientific literature was made on two subjects:

• the absence of functional manipulable materials, for all life stages, and
• tail-biting, for weaner and rearing pigs only.

The second ToR is to identify the strength and predictive capacity of the above identified interactions. To address this ToR, an attempt is made to quantify the relationships described in ToR 1 by carrying out a statistical analysis of information from available databases recording

• multiple welfare outcome indicators in pigs and
• studies on the risk factors for tail-biting.

Two datasets were used for this purpose:

• a set of data from 242 farms in five countries [Spain, France, Finland, Netherlands, and Sweden] collected according to a common Welfare Quality® protocol, and
• a dataset from 1574 farms in Finland, collected by veterinarians during regular herd health visits during 2011 and 2012 providing information regarding the use of 8 different manipulable materials (straw, hay, peat, sawdust, paper, woodchips, wood, toy) together with the presence of tail-biting during the time of the visit to the farm.

The third ToR is to propose a model to evaluate how likely certain welfare consequences may happen given specific risk factors and which animal and/or non-animal-based measures would better fit for the assessment of those consequences. To address this ToR, consideration was given to the processes necessary to construct a ‘diagnostic’ tool-box of animal and non-animal-based measures which can be used to assess the level of risk from the contributing factors in the case of:

• lack of functional manipulable material, or
• tail-biting, and measures which can be used to describe the current extent of the welfare consequences.

Principal Conclusions

Pigs have a need for manipulable materials to satisfy a range of behavioural needs, which can be different in different classes of pig. When these needs are not met, a range of adverse welfare consequences result, one of these being an increased risk for tail-biting in weaners and rearing pigs.

Some manipulable materials, although good at meeting the behavioural needs of pigs, can also have adverse effects on other aspects of pig welfare. These adverse effects have not been adequately studied to ensure safe provision in all cases.

The ability to control the risk of tail-biting through correct identification and alleviation of the predisposing animal, environmental and management factors on that farm is essential when aiming to avoid tail-docking. The presence of these risks can be indicated by a range of resource/management and animal-based measures.

Analyses of an international dataset collected using the Welfare Quality® protocol did not show animal-based measures of behaviour which clearly distinguished between farms providing different types of manipulable material. Category of manipulable material was reflected in severe skin lesions, but not in bitten tails. This may reflect the fact that many farms had pigs with docked tails and there was a confounding between type of manipulable material and tail docking in this dataset.

Analyses of an international dataset using the Welfare Quality® protocol suggested a number of animal and resource-based factors to be important risk factors for tail-biting, but a high degree of uncertainty in the model precludes strong conclusions. The dataset used was not designed to evaluate risk factors for tail-biting and therefore, it had limitations in fitness for this analysis.

The Welfare Quality® dataset indicated the possibility for undocked pigs to be housed and managed in a way which does not imply an increased risk for tail-biting. However, this requires further investigation in more comprehensive datasets where the overall farm prevalence of bitten tails, including animals in hospital pens and euthanized/culled animals, is recorded.

Analyses of a large Finnish dataset with undocked pigs showed that use of straw was associated with reduced tail-biting prevalence relative to the other types of manipulable material (including objects) present on Finnish farms. No other manipulable material gave consistent reduction in tail-biting across both weaner and rearing pigs compared to the population average.

The adequacy of provision of manipulable material could be assessed under farm conditions by reference to a permitted list of materials, but this approach has major practical and biological limitations. A better resource-based approach would be to judge the functionality of the manipulable materials to meet the behavioural need of the pigs by the properties which that material possesses.

Because the human view-point may not correctly interpret the pigs’ perception of material suitability, it would be preferable in a tool-box to use animal-based measures for the assessment. The type of manipulable material supplied has an effect on the prevalence of severe skin lesions. It is also affects prevalence of bitten tails but this measure may be less sensitive if tails are docked. However, the specificity of both lesion measures to assess the functionality of manipulable material is limited. Therefore, a practical tool-box should contain direct behavioural measures.

Animal-based behavioural measures of functionality of the supplied manipulable material need to be simple and robust under farm measurement conditions. The ratio between material-directed exploration and other redirected exploration to pen mates and pen fittings has been suggested for this purpose. However, no comprehensive measure has yet been scientifically validated for this purpose, although studies currently in progress are addressing this question.

A simple tool-box for on-farm use to assess the functionality of the supplied manipulable material is proposed, which includes a combination of the most important resource-based and animal-based measures based on the current state of knowledge.

The presence of known risk factors for tail-biting can be assessed on farm using a tool-box containing both resource/management-based and animal-based measures. These outcome measures may not always be specific for a given risk factor, but the occurrence of a measure suggestive that a risk factor may be present indicates the need for further investigation.

With present knowledge, the relative importance of different risk factors as hazards for tail-biting and the interactions between these risk factors cannot be scientifically quantified. Further studies are needed for this purpose. These should provide the data necessary to weight different risk factors in decision-support tools which can provide customised risk assessment for individual farms.

A simple tool-box for on farm use to assess the presence and strength of risk factors for tail-biting is proposed, which includes a combination of the most important resource-based and animal-based measures based on the current state of knowledge.

Recommendations

The nine recommendations arising from the work are:

1. Any study on manipulable materials should consider possible adverse effects and their alleviation.
2. Further research should be carried out into the causal relationship between the general pig health and tail-biting risk.
3. There is a need for more comprehensive analyses of existing data-sets collected for the purpose of evaluating risk factors for tail-biting in different farm typologies. This could better indicate the relative importance of different risk factors for the occurrence and severity of tail-biting outbreaks, and the way in which these factors interact.
4. In order to assess the true prevalence and importance of the risk factors for tail-biting and their interactions, further harmonised data collection across the range of European farming circumstances is needed. A proposal is made (Appendix J) for a data model which might be used in such a study.
5. There is a need to investigate the farmers’ acceptance level of tail-biting relation to their previous experiences of this problem and perceived ability to limit the level of injury.
6. There is a need for further studies to provide guidance on how to house and manage undocked pigs under different farm circumstances without uncontrollable tail-biting outbreaks.
7. Tail-biting and severe skin lesions should be included in a tool-box to assess the functionality of manipulable material, although it is recognised that these may be caused by many risk factors.
8. Validation of a practical on farm assessment protocol for functionality of manipulable material based on behavioural measures should be carried out, in order to provide a sensitive tool-box measure for use also in docked pigs.
9. The further development and validation, from robust epidemiological data, of decision-support tools for customised assessment of tail-biting risk factors on individual farms is strongly recommended. Such tools could assist farmers to identify, and prioritise correction of, the most important hazards for tail-biting on their own unit.

May 2014