Reliability of the qualitative behavior assessment as included in the Welfare Quality Assessment protocol for growing pigs

How Far Are We From Data-Driven and Animal-Based Welfare Assessment? A Critical Analysis of European Quality Schemes

Within the European Union, there is no harmonization of farm animal welfare quality schemes for meat and dairy products. Instead, there are several industry-driven initiatives and voluntary schemes that seek to provide information on animal welfare for attentive consumers. This study had two aims. First, we quantified how selected industry-wide quality schemes cover the welfare of pigs and dairy cattle on farms by comparing the evaluation criteria selected by schemes with the animal-, resource- and management-based measures defined in the Welfare Quality protocol (WQ®). Second, we identified how these quality schemes use the data generated along the value chain (sensors, breeding, production, and health recordings) for animal welfare assessments. A total of 12 quality schemes, paying attention to animal welfare but not necessarily limited to welfare, were selected for the analysis. The schemes originated from eight European countries: Finland, Sweden, Denmark, Ireland, the Netherlands, Germany, Austria, and Spain. Among the studied quality schemes, we have identified 19 standards for certification: nine for dairy and 10 for pig production. Most of the analyzed standards were comprehensive in welfare assessment. In total, 15 out of 19 standards corresponded to WQ® in more than 70%. However, this high correspondence was obtained when allowing for different information sources (environment instead of animal) than defined in WQ®. Compared to WQ®, the investigated schemes were lagging in terms of the number of measures evaluated based on the animals, with only five standards, out of 19, using predominantly animal-based measures. The quality schemes mostly applied resource-based instead of animal-based measures while assessing good health and appropriate behavior. The utilization of data generated along the value chain by the quality schemes remains insignificant as only one quality scheme allowed the direct application of sensor technologies for providing information on animal welfare. Nevertheless, several schemes used data from farm recording systems, mostly on animal health. The quality schemes rely mostly on resource-based indicators taken during inspection visits, which reduce the relevance of the welfare assessment. Our results suggest that the quality schemes could be enhanced in terms of data collection by the broader utilization of data generated along the value chain.

The Use of the General Animal-Based Measures Codified Terms in the Scientific Literature on Farm Animal Welfare

Background: The approach to farm animal welfare evaluation has changed and animal-based measures (ABM), defined as the responses of an animal or effects on an animal, were introduced to assess animal welfare. Animal-based measures can be taken directly on the animal or indirectly and include the use of animal records. They can result from a specific event or be the cumulative outcome of many days, weeks, or months. The objective of the current study was to analyze the use of general ABM codified terms in the scientific literature, the presence of their definitions, and the gap mapping of their use across animal species, categories, years of publication, and geographical areas of the corresponding author's institution. The ultimate aim was to propose a common standard terminology to improve communication among stakeholders. In this study, data models were populated by collecting information coming from scientific papers extracted through a transparent and reproducible protocol using Web of Science^TM and filtering for the general ABM codified terms (or synonyms/equivalents). A total of 199 papers were retained, and their full texts were assessed. The frequency of general codified ABM terms was analyzed according to the classification factors listed in the objectives. These papers were prevalently European (159 documents), and the most represented species was cattle. Fifty percent of the papers did not provide a definition of the general ABM terms, and 54% cited other sources as reference for their definition. The results of the study showed a very low penetration of the general codified ABM term in the literature on farm animal welfare, with only 1.5% of the papers including the term ABM. This does not mean that specific ABM are not studied, but rather that these specific ABM are not defined as such under a common umbrella, and there is no consensus on the use of terminology, not even among scientists. Thus, we cannot expect the stakeholders to use a common language and a standardized terminology. The recognition and the inclusion of ABM in the lists of commonly accepted abbreviations of scientific journals could be a first step to harmonize the terminology in the scientific literature.

Reliability of observer ratings: Qualitative behaviour assessments of shelter dogs using a fixed list of descriptors

Qualitative Behaviour Assessment (QBA) is a whole-animal approach used to quantify the expressive style of animals’ behaviour. The aim of this study was to evaluate the inter-observer reliability of QBA of shelter dogs using a fixed list of descriptors. The fixed list of 20 terms was generated using a group of experts and literature reviews. In the pilot study, seven veterinary students scored 12 two-minute video clips, and in the main study, 22 final year veterinary nurse students and third-year veterinary students scored the same videos. The two datasets were analysed using Principal Component Analysis (PCA), and the level of agreement for the main components and individual terms was assessed using Kendall's coefficient of concordance (W). In the pilot study, the observer agreement was 0.89 for PC1 and 0.78 for PC2, indicating high inter-observer agreement. The reliability was similarly high for both components in the main study (0.88 and 0.79, respectively). Results also demonstrated high or moderate agreement for most of the terms included in the fixed list. We propose that this approach can be a useful learning tool for students. Our results support further exploration of this method for the assessment of shelter dog welfare by direct observation.

Quantifying defence cascade responses as indicators of pig affect and welfare using computer vision methods

Affective states are key determinants of animal welfare. Assessing such states under field conditions is thus an important goal in animal welfare science. The rapid Defence Cascade (DC) response (startle, freeze) to sudden unexpected stimuli is a potential indicator of animal affect; humans and rodents in negative affective states often show potentiated startle magnitude and freeze duration. To be a practical field welfare indicator, quick and easy measurement is necessary. Here we evaluate whether DC responses can be quantified in pigs using computer vision. 280 video clips of induced DC responses made by 12 pigs were analysed by eye to provide 'ground truth' measures of startle magnitude and freeze duration which were also estimated by (i) sparse feature tracking computer vision image analysis of 200 Hz video, (ii) load platform, (iii) Kinect depth camera, and (iv) Kinematic data. Image analysis data strongly predicted ground truth measures and were strongly positively correlated with these and all other estimates of DC responses. Characteristics of the DC-inducing stimulus, pig orientation relative to it, and 'relaxed-tense' pig behaviour prior to it moderated DC responses. Computer vision image analysis thus offers a practical approach to measuring pig DC responses, and potentially pig affect and welfare, under field conditions.

Development of a fixed list of terms for the Qualitative Behavioural Assessment of shelter dogs

The shelter environment may have a severe impact on the dogs’ quality of life, and there is thus a need to develop valid tools to assess their welfare. These tools should be sensitive not only to the animals’ physical health but also to their mental health, including the assessment of positive and negative emotions. Qualitative Behaviour Assessment (QBA) is a ‘whole animal’ measure that captures the expressive quality of an animal’s demeanour, using descriptive terms such as ‘relaxed’, ‘anxious’, and ‘playful’. In this study, for the first time, we developed and tested a fixed-list of qualitative QBA terms for application to kennelled dogs. A list of 20 QBA terms was developed based on literature search and an expert opinion survey. Inter-observer reliability was investigated by asking 11 observers to use these terms to score 13 video clips of kennelled dogs. Principal Component Analysis (PCA) was used to extract four main dimensions explaining 70.9% of the total variation between clips. PC1 characterised curious/playful/excitable/sociable demeanour, PC2 ranged from comfortable/relaxed to anxious/nervous/stressed expression, PC3 described fearful demeanour, and PC4 characterised bored/depressed demeanour. Observers’ agreement on the ranking of video clips on these four expressive dimensions was good (Kendall’s W: 0.60–0.80). ANOVA showed a significant effect of observer on mean clip score on all PCs (p<0.05), due to few observers scoring differently from the rest of the group. Results indicate the potential of the proposed list of QBA terms for sheltered dogs to serve, in alignment with other measures, as a non-invasive assessment tool. However, the observer effect on mean PC scores points towards the need for adequate observer training, particularly in live scoring conditions. The QBA scoring tool can be integrated with existing welfare assessment protocols for shelter dogs and strengthen the power of those protocols to evaluate the animals’ experience in shelters.

Test-Retest Reliability of the 'Welfare Quality® Animal Welfare Assessment Protocol for Sows and Piglets'. Part 1. Assessment of the Welfare Principle of 'Appropriate Behavior'

Simple Summary

The EU-funded Welfare Quality^® project generated species-specific assessment protocols to evaluate the welfare of pigs, poultry, and cattle. With the implementation to be used for certification purposes, it is important that the protocols show consistency over time, which describes the extent to which equal results are achieved if the assessment is performed repetitively. The protocols should not be sensitive to slight changes in the on-farm situation but mirror the long-term welfare situation on-farm. The present study aimed at testing this consistency over time of the indicators included in the ‘Welfare Quality^® animal welfare assessment protocol for sows and piglets’. Thereby, the study focused on the indicators to assess the welfare principle ‘appropriate behavior’. As a result, the indicators applied to assess the animals’ ‘appropriate behavior’ did not represent consistency over time. Thus, further investigation is needed before implementation on-farm. Conclusively, the present study contributes to the development of generally accepted and objective assessment protocols for animal welfare and thereby to the improvement of farm animals’ welfare overall.

Abstract

The present study’s aim was to assess the test−retest reliability (TRR) of the ‘Welfare Quality^® animal welfare assessment protocol for sows and piglets’ focusing on the welfare principle ‘appropriate behavior’. TRR was calculated using Spearman’s rank correlation coefficient (RS), intraclass correlation coefficient (ICC), smallest detectable change (SDC), and limits of agreement (LoA). Principal component analysis (PCA) was used for deeper analysis of the Qualitative Behavior Assessment (QBA). The study was conducted on thirteen farms in Northern Germany, which were visited five times by the same observer. Farm visits 1 (F1; day 0) were compared to farm visits 2 to 5 (F2–F5). The QBA indicated no TRR when applying the statistical parameters introduced above (e.g., ‘playful‘ (F1–F4) RS 0.08 ICC 0.00 SDC 0.50 LoA [−0.62, 0.38]). The PCA detected contradictory TRR. Acceptable TRR could be found for parts of the instantaneous scan sampling (e.g., negative social behavior (F1–F3) RS 0.45 ICC 0.37 SDC 0.02 LoA [−0.03, 0.02]). The human−animal relationship test solely achieved poor TRR, whereas scans for stereotypies showed sufficient TRR (e.g., floor licking (F1–F4) RS 0.63 ICC 0.52 SDC 0.05 LoA [−0.08, 0.04]). Concluding, the principle ‘appropriate behavior’ does not represent TRR and further investigation is needed before implementation on-farm.