The relationship between common data-based indicators and the welfare of Swiss dairy herds

Exploring uses for an algorithmically generated Animal Welfare Indicator for welfare assessment of dairy herds

On-farm welfare assessment is time-consuming and costly. Assessing welfare using routinely collected herd data has been proposed as a more economical alternative. The online Animal Welfare Indicator (AWI), developed by a Norwegian dairy cooperative, applies an algorithm to routinely collected health, production, and management data to "indicate" aspects of animal welfare at herd level. The overall AWI score is based on 10 AWI subindicator scores, representative of elements of animal welfare such as claw health, udder health, and mortality. Our cross-sectional study explored 2 ways in which the AWI may enable more efficient welfare assessment of Norwegian dairy herds. First, we investigated using the AWI to reduce the duration of on-farm assessments by replacing on-farm measures. Second, we examined reducing the number of on-farm welfare assessments by using the AWI to predict which herds have poorer welfare with respect to specific on-farm measures. Using Spearman rank analyses, we investigated if the AWI scores for 157 herds were associated with 24 on-farm welfare variables measured contemporaneously by Welfare Quality assessment. The mortality AWI subindicator score and the percentage mortality in the previous 12 mo were moderately correlated, as were the udder health AWI subindicator score and the percentage high somatic cell count (SCC) in the previous 3 recordings. Only negligible or weak correlations were found between the other AWI scores and the on-farm assessment variables. We built Generalized Linear Models using AWI scores as independent variables to predict herds with poorer welfare. Herds were classified as having poorer welfare based on their results in specific on-farm welfare measures. We evaluated the models' predictive ability and accuracy. Moderately accurate models were built for predicting poorer herds with respect to high SCC, mortality, and moderate or severe lameness. The other models were less accurate. The AWI scores were generally unsuitable as replacements of on-farm welfare measures. The AWI subindicators for udder health and mortality could replace the on-farm welfare measures related to those 2 topics, but there was some overlap in the data used to calculate them. Despite a lack of independence, the use of those 2 AWI subindicators may marginally reduce the duration of on-farm assessments. A prediction model based on AWI scores showed potential for identifying herds with poorer welfare in terms of moderate or severe lameness, facilitating more efficient use of resources for on-farm lameness assessment. As a consequence of the data used in the AWI, it was only reflective of health-related welfare outcomes.

Development of a data-driven method for assessing health and welfare in the most common livestock species in Switzerland: The Smart Animal Health project

Improving animal health and welfare in livestock systems depends on reliable proxies for assessment and monitoring. The aim of this project was to develop a novel method that relies on animal-based indicators and data-driven metrics for assessing health and welfare at farm level for the most common livestock species in Switzerland. Method development followed a uniform multi-stage process for each species. Scientific literature was systematically reviewed to identify potential health and welfare indicators for cattle, sheep, goats, pigs and poultry. Suitable indicators were applied in the field and compared with outcomes of the Welfare Quality® scores of a given farm. To identify farms at risk for violations of animal welfare regulations, several agricultural and animal health databases were interconnected and various supervised machine-learning techniques were applied to model the status of farms. Literature reviews identified a variety of indicators, some of which are well established, while others lack reliability or practicability, or still need further validation. Data quality and availability strongly varied among animal species, with most data available for dairy cows and pigs. Data-based indicators were almost exclusively limited to the categories “Animal health” and “Husbandry and feeding”. The assessment of “Appropriate behavior” and “Freedom from pain, suffering, harm and anxiety” depended largely on indicators that had to be assessed and monitored on-farm. The different machine-learning techniques used to identify farms for risk-based animal welfare inspections reached similar classification performances with sensitivities above 80%. Features with the highest predictive weights were: Participation in federal ecological and animal welfare programs, farm demographics and farmers' notification discipline for animal movements. A common method with individual sets of indicators for each species was developed. The results show that, depending on data availability for the individual animal categories, models based on proxy data can achieve high correlations with animal health and welfare assessed on-farm. Nevertheless, for sufficient validity, a combination of data-based indicators and on-farm assessments is currently required. For a broad implementation of the methods, alternatives to extensive manual on-farm assessments are needed, whereby smart farming technologies have great potential to support the assessment if the specific monitoring goals are defined.