S. Waiblinger et al., The development of an epidemiologically based on-farm welfare assessment system for use with dairy cows, ACT AG SC A, 2001, pp. 73-77
A scientifically based and operational welfare assessment tool for advice a
nd control purposes must involve measures that are reliable and valid, be b
oth easily and swiftly operated, and reveal the causes of impaired welfare.
In our view these criteria can only be fulfilled by applying a mixture of
animal-related parameters, i.e. direct welfare indicators, and indirect hou
sing and human-related parameters, i.e, influencing factors. However, altho
ugh our knowledge of the effects of individual influencing factors on anima
l health and well-being is in part well developed, relatively little is kno
wn about the effects of the interaction of these factors and their relative
importance within the animal's complex environment. We therefore propose t
o undertake epidemiological research and use multivariate analysis in order
to construct assessment models relating to particular aspects of welfare.
The integration of individual evaluations into one overall welfare assessme
nt may entail a loss of information and is, we suggest, in many cases unnec
essary.