The assessment of poultry welfare on a commercial scale is necessary for both research and on a regular basis for auditing, said Professor Marian Stamp Dawkins of the University of Oxford, an internationally renowned expert in the field of animal welfare. She was delivering the Gordon Memorial lecture at the annual meeting of the UK Branch of the World's Poultry Science Association at the University of Nottingham in April.

A small-scale assessment would offer more complete control over the conditions, she said, but what is true for a flock of 50 birds may not hold true for a commercial flock of 50,000.

She identified what she described as the 'package problem', explaining that advice differs depending on the source so that that from animal welfare groups is different from the 'green' movement, for example. Add to this the vital consideration of food safety, and Professor Dawkins said that welfare is more likely to be taken seriously by all stakeholders if it is included in a package of concerns.

Animal welfare is not the public's only concern, she said, adding that the FAO report 'Livestock's Long Shadow' does not mention animal welfare at all and neither does the UK government's report on future food security, 'Food 2030'.

Consumers make their purchasing decisions based on factors other than animal welfare. In a survey carried out last year, animal welfare was the number 4 priority and although 40 per cent of consumers bought at least some organic food on the grounds if its perceived health benefits, animal welfare was a lower priority for them.

Also important, Professor Dawkins highlighted, is the so-called 'sustainability package', and whether animal welfare impacts it at all, positively or negatively. She gave the example of a study of commercial free-range broilers, assessed in terms of animal welfare, economics, environmental impact and taste in the Poultry in the Natural Environment (PINE) project. The hens were kept in movable 600-bird houses. Campylobacter were isolated from this birds and study of the clonal complexes revealed that the strains present in the chickens were the same as those in humans, while those in wild birds were different.

Commercial-scale assessments address both the scaling and package problems as well as providing information on costings and practicalities, Professor Dawkins said.

* "Welfare is more likely to be taken seriously by all stakeholders if it is included in a package of concerns"

Different Types of Commercial Trial

Professor Dawkins explained that there are four main types of commercial trials: natural experiments (epidemiology); research (manipulation); trials conducted by companies and the analysis of commercial data already collected.

Natural experiments

Taking feather-pecking in layers as an example, she said that the behaviour is a serious welfare problem and it is also an economic issue. It is important to prevent feather-pecking as it is difficult to stop later on and control measures (beak trimming) may also have welfare implications.

For this epidemiological study, factors were investigated that correlate with feather damage. There was a great deal of natural variation in terms of the amount of feather damage and the age at which the critical threshold was reached. Factors identified as important included feed intake, carbon dioxide (air quality) and sound level, said Professor Dawkins. Birds that remained on the same farm for rearing and laying had better quality feathers for longer, although the change of location would have been confounded with other changes of feeder type, personnel etc.

Natural experiments have the potential disadvantages of missing 'treatments', absence of a control and correlation rather than causation although this type of trial is relatively easy to manage, according to Professor Dawkins.

Research trials

As an example of a research trial using manipulation, Professor Dawkins gave the example of an experiment investigating the effects of stocking density on broiler welfare, which was carried out by 11 major companies in Denmark and the UK. Each company used each stocking density in two separate trials.

From the results, it was clear that each company had very different average mortality levels. There were large effects of company in terms of overall mortality and for each cause but stocking density turned out not to be among the most important factors. Variations in air and litter quality accounted for much of the difference between the individual companies.

The disadvantages of research trials, said Professor Dawkins, are that they are more difficult to carry out, involve a certain degree of interference and may involve ethical issues. On the other hand, they give better control over the variables and cross-company comparisons can give a fuller picture.

Company research

Yet more difficulties can arise from commercial trials, said Professor Dawkins. Firstly, she mentioned possible conflicts between the commercial and research aims, and second, auditing is required to verify company data. Third, there may be problems with the experimental design and finally, there may be problems getting a Home Office licence, which is needed for almost all welfare research in the UK.

Professor Dawkins cited the example of a trial of this type investigating the welfare needs of commercial ducks and how to provide a water source for ducks without compromising hygiene. Four treatments were used, including nipple drinkers and showers, and these had to be licensed under the 1986 Act because they are not normally provided for ducks. The ducks showed a preference for the showers.

This highlights a need to clarify the law, said Professor Dawkins, because the Act allows the animals to be observed but a licence is required for anything else. This presents challenges for company-run trials and especially in some particularly sensitive areas such as feed restriction of broiler breeders.

Some of the challenges of this type of trial can be overcome, she said, by building up a good co-operation between the parties, respecting both company confidentiality and the need for academic publications, a good trial design and statistics and finally, clarification of the legal position.

Commercial level assessment

A data bank for poultry welfare could be achieved by pooling production and welfare data, much of which is already collected in the UK and the EU, for example, said Professor Dawkins. New technology would be needed, for example, the automated assessment of welfare information already collected and/or done on a smaller scale, in order to build up a full picture of the situation.

Currently, outcomes measured for broiler welfare include mortality, hock condition and foot pad dermatitis (after slaughter). There is also some data on gait score in growing broilers although this work tends to be labour-intensive, somewhat subjective, presents a biosecurity risk and presents only a snap-shot of the situation at the time of assessment.

New Technology Shows Promise

New technology for welfare research needs to be developed, which is inexpensive, easy to use and fully automated, said Professor Dawkins, who then described a camera system using optical flow analysis to compare patterns of intensity in successive images. It does not track individual birds, she explained, but it does reflect the speed and uniformity of the individuals within the flock.

A broiler flock with a high proportion of lame birds will appear to have poor uniformity of movement and so a graph of the flows will be skewed and/or have a long 'tail' when plotted on a distribution graph.

Professor Dawkins explained that this measure of optical flow is highly correlated with gait score: a higher percentage of lame birds will be shown as low mean flow, together with high level of skewing. Furthermore, the wireless system used is very tough, inexpensive and gives continuous readings throughout the flock's life.

In summary, Professor Dawkins stressed the need for commercial level assessment of poultry welfare as part of a package of information on poultry production for producers, retailers and consumers. The new technology of camera and image analysis has much to contribute towards this assessment, she said.

Further Reading

-

For more information on the new technology mentioned in this article, click here.

June 2011