Sensitivity of the integrated Welfare Quality^® scores to changing values of individual dairy cattle welfare measures

Abstract

School, Brewhouse Hill, Wheathampstead, Hertfordshire AL4 8AN, UK www.ufaw.org.uk Animal Welfare 2018, 27: 157-166 ISSN 0962-7286 doi: 10.7120/09627286.27

A New, Practical Animal Welfare Assessment for Dairy Farmers

“…This was confirmed in the study of Heath et al [ 5 ], where 88% of the farms could be classified correctly with “absence of prolonged thirst”, a component of the first principle, only. De Graaf et al [ 8 ] also reported that the absence of prolonged thirst and the Qualitative Behaviour Assessment were the most influential measures. Heath et al [ 5 ] suggested that the protocol could be shortened to just 15 min with the same outcome.…”

“…However, the execution of the extensive WQ protocol [ 3 ] is time-consuming (almost a full day is needed), which has hampered its implementation as a routine welfare check on a dairy farm [ 4 , 5 ]. Furthermore, the relative contribution of certain measures/criteria to the end classification of the WQ is disputed [ 5 , 6 , 7 , 8 ]. For example, in the study of Heath et al [ 5 ], it appeared that they could classify the final outcome of the WQ protocol [ 3 ] correctly in 88% of the farms, with the result only for “Absence of prolonged thirst”.…”

Using Expert Elicitation to Abridge the Welfare Quality® Protocol for Monitoring the Most Adverse Dairy Cattle Welfare Impairments

“…Step 2: Severity Scores Median expert severity scores were highest for severe lameness (92, interquartile range = 90-97) and mortality (90, 69-100) followed by leanness (61, 50-71) and SCC > 400,000 (73, 43-80), and lowest for hairless patches (18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34) and wounds/swellings (40-58).…”

“…The WQ protocols have been praised for being very comprehensive and for the implementation of a hierarchical approach to integrate data on a multitude of predominantly animal-based welfare measures enabling the assignment of farms or herds to one of the four overall welfare categories (not classified, acceptable, enhanced, and excellent). Although issues about consistency over time (6)(7)(8)(9) and about reliance on complete and standardized farm/slaughterhouse records (10)(11)(12) have been raised, the WQ protocols have been criticized mainly with regard to the (i) the feasibility [mainly labor costs per farm, e.g., (11,13)], (ii) the variable quality of the welfare measures included in the protocol (8,10,14), and (iii) the way these measures are aggregated into an overall WI (15)(16)(17)(18)(19)(20)(21). Indeed, uptake of the WQ protocols by the authorities and food industry at large for improving and better marketing of farm animal welfare has been below expectation.…”

“…The dairy cattle protocol, for example, relies on resource-based measures for assessing 3 of the 12 welfare criteria (water availability and cleanliness for the criterion absence of prolonged thirst, tethering for the criterion ease of movement, and pasture access for the criterion expression of other behaviors). It is particularly worrying that sensitivity analyses have revealed that a limited number of (often resourcebased) measures seem to have a disproportionally large effect on the overall welfare categorization [e.g., 88% of the overall dairy cattle welfare categorization is predicted by water availability and cleanliness (17)], whereas some key (often animal-based) measures such as lameness and mortality have a negligible effect (16)(17)(18)21). This appears to be an unwanted side effect of the very complex and hard-to-understand (and hence poorly transparent to most end-users) integration method, which was needed to aggregate so many measures of different scales with different thresholds.…”