Petra Tuunainen a, Jarmo Valaja a, Eija Valkonen a & Helena Hepola b a MTT Agrifood Research Finland, Jokioinen,

Transcription

1 This article was downloaded by: [Dr Kenneth Shapiro] On: 09 June 2015, At: 10:19 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Journal of Applied Animal Welfare Science Publication details, including instructions for authors and subscription information: Using Conjoint Analysis to Weight Broiler Welfare Variables Based on Slaughterhouse Data Petra Tuunainen a, Jarmo Valaja a, Eija Valkonen a & Helena Hepola b a MTT Agrifood Research Finland, Jokioinen, Finland b Finnish Food Safety Authority, Helsinki, Finland Published online: 10 Jan To cite this article: Petra Tuunainen, Jarmo Valaja, Eija Valkonen & Helena Hepola (2012) Using Conjoint Analysis to Weight Broiler Welfare Variables Based on Slaughterhouse Data, Journal of Applied Animal Welfare Science, 15:1, 70-79, DOI: / To link to this article: PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content ) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any

2 losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at

3 JOURNAL OF APPLIED ANIMAL WELFARE SCIENCE, 15:70 79, 2012 Copyright Taylor & Francis Group, LLC ISSN: print/ online DOI: / RESEARCH REPORTS Using Conjoint Analysis to Weight Broiler Welfare Variables Based on Slaughterhouse Data Petra Tuunainen, 1 Jarmo Valaja, 1 Eija Valkonen, 1 and Helena Hepola 2 1 MTT Agrifood Research Finland, Jokioinen, Finland 2 Finnish Food Safety Authority, Helsinki, Finland The Council of the European Union adopted the welfare directive for broiler chickens in 2007 (Commission of the European Communities, 2007). The directive defines minimum demands and objectives for broiler production ranging from stocking density to welfare. This study found that the level of nonhuman animal welfare can be determined, for example, with welfare indexes that have different weight values for welfare variables. A total of 20 poultry experts from slaughterhouses, the feed industry, administration, and research received simulated data on 24 flocks. This included data for 5 welfare variables based on postmortem meat inspection information on commercial flocks and based on literature: mortality during rearing, foot-pad score, and the incidence of ascites, dermatitis, and cachexia in routine meat inspection. The experts ranked these flocks from the worst to the best according to animal welfare. Conjoint analysis weighted the relative importance of the welfare variables for broiler welfare. The study based relative importance for broiler welfare on 20 expert opinions: 24.6% for mortality, 35.9% for foot-pad score, 19.7% for the incidence of ascites, 15.7% for the incidence of dermatitis, and 4.1% for the incidence of cachexia. In the future, an index for broiler welfare evaluation could be developed based on the results of this study. Correspondence should be sent to Petra Tuunainen, MTT Agrifood Research Finland, FI Jokioinen, Finland. petra.tuunainen@mtt.fi 70

4 CONJOINT ANALYSIS AND BROILER WELFARE 71 The most abundant welfare problems in broiler chickens (Gallus gallus) are connected with rapid growth rate and poor rearing conditions. Mortality rate is one of the most elementary welfare indicators. After the adoption of the Council of the European Union s welfare directive on chickens kept for meat production (Commission of the European Communities, 2007), the authorities have monitored mortality as cumulative daily mortality, which should be below 1% C 0.06% multiplied by the slaughter age of the flock in at least seven consecutive flocks. Mortality includes the number of chickens who have died and those whom the farmer has culled. If mortality rises above the limits, the authorities lower the permitted stocking density, which reduces the producer s income. Ascites is one of the most common causes of death in chickens kept in meat production (Druyan, Ben-David, & Cahaner, 2007; Julian, 1998). Ascites arises from the rapid growth of birds (breeding influence), but the producer can attenuate the rate of ascites by changing the rearing conditions and reducing stress. In addition, excessively high carbon dioxide and ammonia concentrations strain a bird s lungs and heart (Julian, 1998). Litter quality is also important. A slightly damp litter absorbs ammonia from the air, thus improving air quality (Aftab & Khan, 2005). The rearing temperature should also be high enough to reduce oxygen needs for heat regulation and lessen the burden on the chicken heart (Wideman & Tackett, 2000). Lighting regimens can also be used to alter rearing conditions (Aftab & Khan, 2005). During the dark periods, the birds rest and eat less, which slows growth and reduces the risk of ascites. Rearing conditions also affect the incidence of skin diseases and the health of foot-pads (Bessei, 2006). Too wet litter and high ammonia concentrations cause skin lesions, especially on the chest, and have a great effect on foot-pad scores (Berg, 1998; Bessei, 2006; Meluzzi & Sirri, 2009). In addition, nervous birds can cause skin lesions in each other. Skin damages such as cuts and bruises expose birds to bacteria that can cause cellulitis and dermatitis (Feddes, Emmanuel, Zuidhof, & Korver, 2003). Stocking density is linked to skin lesions, with high stocking density resulting in decreased litter quality and increased likelihood of birds injuring each other (Bessei, 2006). In addition to ascites and skin dermatitis, cachexia (underweight birds) is one of the most prevalent reasons for carcass condemnation in Finland. Cachectic birds are small, and their growth is slower than that of other birds. Small birds may have difficulties in reaching feed and water, and they may suffer from dehydration and undernourishment. Normally, farmers cull these birds, but some of these underweight birds occasionally end up in the slaughterhouse. For example, the level of nonhuman animal welfare can be determined with welfare indexes that have different weight values for the different welfare variables (Hörning, 2001). Such welfare indexes attempt to estimate animal welfare comprehensively on farms and strive to provide a measurable comparison of welfare on different farms (Wemelsfelder, Hunter, Mendl, & Lawrence, 2001).

5 72 TUUNAINEN, VALAJA, VALKONEN, HEPOLA Researchers from different parts of Europe have developed, studied, and applied a variety of welfare indexes (Johnsen, Johanssenson, & Sandoe, 2001; Wemelsfelder et al., 2001). Some of these welfare evaluation methods, such as an environment-based welfare index, focus on the assessment of animal husbandry conditions (Bartussek, 2001; Johnsen et al., 2001); others concentrate on measurements made on the animal (animal-based index), such as animal health, yield, and abnormal behavior (Johnsen et al., 2001). Environment-based indexes measure only environmental factors and try to predict their impact on animal welfare. Thus, an environment-based index actually measures not welfare itself but the requirements for good welfare. Animal-based indexes, on the other hand, can be troublesome because they may require laborious measurements of factors such as behavior. Environmental variables have the advantage of generally remaining relatively unchanged; moreover, they can be measured objectively, whereas measurements of animal-based variables are often subjective. A welfare index can also combine the two methods. So that it can give the best possible description of animal welfare, an ideal welfare index combines several different environmental and animal-based variables (Algers & Berg, 2001). After the variables for a welfare index are chosen, their weighting factors must be decided on (Bracke, Metz, & Spruijt, 2001). The difficulty is to find a suitable factor for each variable. For example, is the social contact more important than the space allowance? If so, how much more important is it? Weighting factors can all be equal, or each variable can have a different weighting factor. Equally weighted variables are easier to use; however, differences between variables are lost (Bracke et al., 2001; Scott, Nolan, & Fitzpatrick, 2001). Because all variables do not affect welfare with the same intensity, the differences between flocks become more visible when each variable has a different weight. The determination of weighting factors can be based on intuition, literature, measured animal preference, or expert opinions (Bracke et al., 2001; Den Ouden, Nijsing, Dijkhuizen, & Huirne, 1997; Haslam & Kestin, 2003; Scott et al., 2001). The European Welfare Quality (2009) project is a common European project that has developed standardized ways of measuring welfare. Welfare Quality focuses on animalbased variables; it is therefore laborious and requires well-trained assessors. The purpose of this study was to make use of routinely collected data on broiler production to find suitable variables for measuring broiler welfare and to weight the variables based on expert opinions with the help of a conjoint analysis. The results of this study can be used to develop a welfare index readily applicable to broiler production. MATERIALS AND METHODS We collected mortality data and postmortem meat inspection information on

6 CONJOINT ANALYSIS AND BROILER WELFARE commercial flocks in Finland. These data were analyzed, and the analysis and literature were used to choose the following five welfare variables for the index: (a) mortality, (b) foot-pad score, (c) the incidence of ascites, (d) the incidence of skin dermatitis, and (e) the incidence of cachexia. To measure the experts opinions, we composed a set of welfare profiles for evaluation. These welfare profiles had to contain different levels of variables that overlap each other as little as possible so that the respondents were easily able to choose which profile they preferred. The selection of different class levels of the variables is based on the relevant literature and the data analysis of the 90 commercial flocks. Mortality includes all dead and culled birds from flock. The foot-pad lesions were defined from 100 birds per flock and classified according to a Swedish foot-pad health surveillance program (Ekstrand, Carpenter, Andersson, & Algers, 1998); however, the actual score was calculated according to newer instructions (Svensk Fågel, 2005). Ascites, skin dermatitis, and cachexia were determined as number of rejected carcasses per flock in meat inspection. Table 1 shows the class levels of each variable (subclasses). Variables and subclasses were turned into welfare profiles, meaning that all significant combinations are present. It is possible to make 4*3*3*2*2 D 144 different welfare profiles from the five variables and their subclasses. This amount would have been impossible to handle, so we used orthogonal series (SAS Version 9.1) to compose 24 welfare profiles (model flocks) from the five variables and their subclasses. Poultry experts from slaughterhouses, the feed industry, public administration, and research received simulated data on 24 flocks. The experts sorted these flocks from the worst (1) to the best (24) according to animal welfare. Conjoint analysis weighted the relative importance of the welfare variables (Table 2). The statistical analysis was made using conjoint analysis (SAS Version 9.1), which finds the best possible combination of the given variables, so that some variables are weighted more than others. The welfare variables were assigned TABLE 1 The Subclasses of Each Welfare Variable Mortality a Foot-pad Score b Ascites c Dermatitis c Cachexia c <3.2% <40 p <0.5% <0.15% <0.2% % p % >0.15% >0.2% % >80 p >1.5% >4.2% a Dead and culled birds per flock. b Foot-pad scores of 100 birds per flock: no lesions D 0 p, mild lesions D 0.5 p, and severe lesions D 2 p (Svensk Fågel, 2005). c Rejected carcasses per flock in meat inspection.

7 74 TUUNAINEN, VALAJA, VALKONEN, HEPOLA TABLE 2 Welfare Profiles (Simulated Model Flocks) Model Foot-pad Flock Mortality a Score b Ascites c Dermatitis c Cachexia c Note. The model flocks were composed using 5 welfare variables and their subclasses so that each model flock is unique and overlaps the other model flocks as little as possible. a Dead and culled birds per flock. b Foot-pad scores of 100 birds per flock: no lesions D 0 p, mild lesions D 0.5 p, and severe lesions D 2 p (Svensk Fågel, 2005). c Rejected carcasses per flock in meat inspection. relative importance values based on the experts opinions. We calculated the averages of the importance values assigned by the respondents to each welfare variable to construct a mathematical expression for a total index score. In addition, we assessed the extent of agreement between the experts using Kendall s Coefficient of Concordance, Kendall s W (SAS Version 9.1, macro %MAGREE). RESULTS We received responses from 21 poultry experts; one of the experts, however, did not rate all the model flocks. Some values would thus have been missing from this evaluation; therefore, we decided to exclude it and use only 20 evaluations

8 CONJOINT ANALYSIS AND BROILER WELFARE 75 TABLE 3 Conjoint Analysis Results: The Relative Importance Values for Selected Welfare Variables Based on 20 Expert Opinions Expert Mortality a Foot-pad Score b Ascites c Dermatitis c Cachexia c a Dead and culled birds per flock. b Foot-pad scores of 100 birds per flock: no lesions D 0 p, mild lesions D 0.5 p, and severe lesions D 2 p (Svensk Fågel, 2005). c Rejected carcasses per flock in meat inspection. in the analysis. The results of the conjoint analysis are shown in Table 3. Based on the experts opinions, the mean relative importance for broiler welfare was 24.6% for mortality, 35.9% for foot-pad score, 19.7% for the incidence of ascites, 15.7% for the incidence of dermatitis, and 4.1% for the incidence of cachexia (Figure 1). The coefficient of concordance (Kendall s W) or the agreement among the experts was 0.517, which was moderately good (p <.0001). Kendall s W ranges from 0 (no agreement) to 1 (complete agreement). DISCUSSION A number of experimental studies have attempted to combine a number of different welfare indicators to produce a single numerical value to indicate animal welfare (Bracke et al., 2001; Hörning, 2001). Conjoint analysis is a popular marketing research technique that finds the best solution from among

9 76 TUUNAINEN, VALAJA, VALKONEN, HEPOLA FIGURE 1 Summary of importance weighting of welfare variables. Each bar represents the relative importance for broiler welfare, and each error bar represents 95% confidence intervals for the mean. different alternatives. It has also been used to evaluate pig and broiler welfare using expert opinions (Den Ouden et al., 1997; Haslam & Kestin, 2003) and to elicit consumers preferences in egg production (Mesías, Martínez-Carrasco, Martínes, & Gaspar, 2010). In conjoint, the experts considered mortality (24.6%) and the incidence of ascites (19.7%) to be good variables for measuring welfare. The incidence of dermatitis (15.7%) and cachexia (4.1%) were seen as less important factors in animal welfare. Haslam & Kestin (2003) used similar welfare variables in their study: mortality, stocking density, percentage of birds with a gait score of 3 or more, percentage of birds with hock burn lesions, level of enrichment provision, and level of emergency provision. They found that experts considered mortality the most important variable (importance of 26%), followed by gait score (24%) and level of hock burn lesions (16%). These results agree with our findings that foot health and mortality are seen as good variables for measuring broiler welfare. We chose all variables from routinely collected data from farm and slaughterhouses. These variables can practically be collected and are likely to be both accurately recorded and comparable between farms. Our results address only the study of the selected variables, and the results may change if the number of welfare variables is reduced or increased. Similarly, the results could change if, for example, we used the welfare directive s limit values to categorize mortality. However, it must be taken into account that it is difficult to use a large number of variables in arranging model flocks according to their relative welfare. It was

10 CONJOINT ANALYSIS AND BROILER WELFARE 77 also hard to evaluate the welfare of some model flocks because some of the variables subclasses had very similar values. In total, we received 21 responses from poultry experts, but some values were missing from one of these responses. Therefore, we only analyzed 20 responses. The agreement among the experts (Kendall s W) was moderately good (0.517). This fairly low result may reflect the different priorities, experience, education, and knowledge of the experts consulted. Some of the experts concentrated on just one or two variables and sorted the model flocks simply by putting those variables in order, whereas others results distributed weight more evenly between the variables. This is visible in the importance values (Table 2). Haslam and Kestin (2003) had three different expert groups: researchers, veterinary surgeons with a postgraduate qualification in welfare science (DWEL), and veterinary surgeons with a postgraduate qualification in poultry medicine and production (DPMP). The Coefficient of Concordance between all the experts was 0.64; the Coefficients of Concordance between the members in each of the expert groups of researchers, DPMP veterinary surgeons, and DWEL veterinary surgeons were 0.67, 0.7, and 0.77, respectively. Haslam and Kestin s (2003) results are higher than ours. Possibly, our experts came from more varying backgrounds, which made the differences between experts more visible. In the absence of ability to ask the birds, the relative importance of different welfare measures must necessarily be determined using the expert opinions. Number and background of the experts always affect the results. Our index needs to be validated and compared with the other welfare indexes and measurements. The index variables were based on postmortem animal-based measurements. Similar parameters have also been used in a Welfare Quality (2009) model. Postmortem parameters are registered from the whole flock and are not dependent on sampling methods when live birds are evaluated. However, in postmortem evaluation, management improvements could be made only to following flocks. Mortality was highly ranked by experts, but it is a rather conflicting welfare indicator. Mortality consists of many welfare problems that all end in death but does not give any indication for farmers to improve production. If culling were registered with actual mortality, the experts could better differentiate the negative impact of the mortality and the positive impact of the culling on animal welfare. CONCLUSION A welfare index based on routinely collected data is feasible and does not require laborious visits to farms. In the future, the results from this study could be used to create an index for the evaluation of broiler welfare. The importance values

11 78 TUUNAINEN, VALAJA, VALKONEN, HEPOLA could be used as weighting factors for the variables in calculating a welfareindex score for commercial broiler flocks. Such a welfare-index score could be used as a measure of broiler welfare with which flocks and farms could be compared. ACKNOWLEDGMENTS This research was supported by the Finnish Ministry of Agriculture and Forestry and was performed in collaboration with Finland s Poultry Association, Atria Oy, HK Ruokatalo Oy, and Saarioinen Oy. The authors would like to thank all of those who helped with the planning on this study, data collection, and analysis. REFERENCES Aftab, U., & Khan, A. A. (2005). Strategies to alleviate the incidence of ascites in broilers: A review. Brazilian Journal of Poultry Science, 7, Algers, B., & Berg, C. (2001). Monitoring animal welfare on commercial broiler farms in Sweden. Acta Agriculturae Scandinavica, Section A, Animal Science, 30, Bartussek, H. (2001). An historical account of the development of the Animal Needs Index ANI-35L as part of the attempt to promote and regulate farm animal welfare in Austria: An example of the interaction between animal welfare science and society. Acta Agriculturae Scandinavica, Section A, Animal Science, 30, Berg, C. C. (1998). Foot-pad dermatitis in broilers and turkeys: Prevalence, risk factors and prevention [Doctoral thesis, Swedish University of Agricultural Sciences, Uppsala, Sweden]. Acta Universitatis Agriculturae Sueciae Vetenaria, Series 36. Bessei, W. (2006). Welfare of broilers: A review. Worldt s Poultry Science Journal, 62, Bracke, M. B. M., Metz, J. H. M., & Spruijt, B. M. (2001). Development of a decision support system to assess farm animal welfare. Acta Agriculturae Scandinavica, Section A, Animal Science, 30, Commission of the European Communities. (2007). Council Directive 2007/43/EC of 28 June 2007 laying down minimum rules for the protection of chickens kept for meat production. Official Journal of the European Union, L 182, Den Ouden, M., Nijsing, J. T., Dijkhuizen, A. A., & Huirne, R. B. M. (1997). Economic optimization of pork production-marketing chains: Model input on animal welfare and costs. Livestock Production Science, 48, Druyan, S., Ben-David, A., & Cahaner, A. (2007). Development of ascites-resistant and ascitessusceptible broiler lines. Poultry Science, 86, Ekstrand, C., Carpenter, T. E., Andersson, I., & Algers, B. (1998). Prevalence and control of foot-pad dermatitis in broilers in Sweden. British Poultry Science, 39, Feddes, J. J. R., Emmanuel, E. J., Zuidhof, M. J., & Korver, D. R. (2003). Ventilation rate, air circulation and bird disturbance: Effects on the incidence of cellulitis and broiler performance. Journal of Applied Poultry Research, 12, Haslam S. M., & Kestin, S. C. (2003). Use of conjoint analysis to weight welfare assessment measures for broiler chickens in UK husbandry systems. Animal Welfare, 12,

12 CONJOINT ANALYSIS AND BROILER WELFARE 79 Hörning, B. (2001). The assessment of housing conditions of dairy cows in littered loose housing systems using three scoring methods. Acta Agriculturae Scandinavica, Section A, Animal Science, 30, Johnsen, P. F., Johanssenson, T., & Sandoe, P. (2001). Assessment of farm animal welfare at herd level: Many goals, many methods. Acta Agriculturae Scandinavica, Section A, Animal Science, 30, Julian, R. J. (1998). Rapid growth problems: Ascites and skeletal deformities in broilers. Poultry Science, 77, Meluzzi, A., & Sirri, F. (2009). Welfare of broiler chickens. Italian Journal of Animal Science, 8, Mesías, J., Martínez-Carrasco, F., Martínes, J. M., & Gaspar, P. (2010). Functional and organic eggs as an alternative to conventional production: A conjoint analysis of consumers preferences. Journal of the Science of Food and Agriculture, 91, Scott, E. M., Nolan, A. M., & Fitzpatrick, J. L. (2001). Conceptual and methodological issues related to welfare assessment: A framework for measurement. Acta Agriculturae Scandinavica, Section A, Animal Science, 30, Svensk Fågel. (2005). Branschorganisationen Svensk Fågels Djuromsorgsprogram [The animal provision program of the Swedish Poultry Meat Association]. Retrieved from se/attachments/35/299.pdf Welfare Quality. (2009). Welfare Quality assessment protocol for poultry (broilers, laying hens). Lelystad, The Netherlands: Welfare Quality Consortium. Wemelsfelder, F., Hunter, T. E. A., Mendl, M. T., & Lawrence, A. B. (2001). Assessing the whole: A free choice profiling approach. Animal Behaviour, 62, Wideman, R. F., & Tackett, C. D. (2000). Cardio-pulmonary function in broilers reared at warm or cool temperatures: Effect of acute inhalation of 100% oxygen. Poultry Science, 79,