Projects per year
Abstract
Application
Precision livestock farming technologies already in use for management purposes (oestrus/illness detection) may be used to improve understanding of animal welfare in an automated way, inform farmers on herd health/dynamics and provide a verifiable, non-biased indicator of positive welfare.
Introduction
UK consumers have an interest in dairy cattle welfare, with 93% of respondents to a 2009 study (Ellis et al., 2009) confirming a willingness to pay more for good dairy welfare practices. Providing dairy cows pasture access allows increased opportunity for expression of natural behaviours, may improve health, and many consumers consider it a clear indicator of good welfare and “happy cows”. This work looked to compare automatically collected sensor data with manually collected behavioural data to explore positive animal welfare on UK dairy farms.
Materials and methods
Qualitative Behaviour Assessment (QBA) (Andreasen et al., 2013) was carried out twice on four farms, at pasture and during housing, and 20 QBA metrics examined in relation to animal-mounted sensor outputs. On each farm, for each location, 20 animals were randomly scored by one staff member, giving 160 data points (80 housed and 80 pasture). Sensor data from leg-mounted accelerometers were acquired during housing and pasture for lying times and step count. Principal components analysis (PCA) was used to analyse QBA data in R using parallel analysis.
Results
Behavioural assessment showed animals at pasture displayed more positive behaviours. Positive PC1 scores were observed in 67.5% of pasture cows, where PC1 score refers to mood and runs negative to positive. In comparison, PC1 for housed cows showed 61% of animals had a negative behavioural score. Sensor data, particularly step count and lying time, differed with cattle location - housed or pasture – and correlated with QBA data. For example, step count was positively correlated with “happiness” (p = <0.001, r = 0.4), PC1 score (p = <0.001, r = 0.3) and others. There was significantly greater “clustering” of lying times at pasture (p = 0.004), with cows lying for similar times compared to housed cows. Clustering at pasture was attributed to the potential for cows to lie down at the same time, i.e., lying synchrony. As a result of increased lying space at pasture, cows have the ability to exhibit similar behaviours at the same time, which can be limited indoors by housing and management practices. Behavioural synchrony - e.g., lying and feeding synchrony - has been shown to be a positive welfare indicator in cattle (Fregonesi and Leaver, 2002, Napolitano et al., 2009). A smaller spread of QBA scores was noted at pasture, also suggesting higher levels of synchrony, with more animals exhibiting similar levels of behaviours.
Conclusions
This work suggests that it is possible to use automatically collected sensor data to understand positive welfare in dairy cattle. The potential for an automated link between existing farm management sensor data and positive cattle welfare is the subject of continuing work (https://www.digitaldairychain.co.uk).
Precision livestock farming technologies already in use for management purposes (oestrus/illness detection) may be used to improve understanding of animal welfare in an automated way, inform farmers on herd health/dynamics and provide a verifiable, non-biased indicator of positive welfare.
Introduction
UK consumers have an interest in dairy cattle welfare, with 93% of respondents to a 2009 study (Ellis et al., 2009) confirming a willingness to pay more for good dairy welfare practices. Providing dairy cows pasture access allows increased opportunity for expression of natural behaviours, may improve health, and many consumers consider it a clear indicator of good welfare and “happy cows”. This work looked to compare automatically collected sensor data with manually collected behavioural data to explore positive animal welfare on UK dairy farms.
Materials and methods
Qualitative Behaviour Assessment (QBA) (Andreasen et al., 2013) was carried out twice on four farms, at pasture and during housing, and 20 QBA metrics examined in relation to animal-mounted sensor outputs. On each farm, for each location, 20 animals were randomly scored by one staff member, giving 160 data points (80 housed and 80 pasture). Sensor data from leg-mounted accelerometers were acquired during housing and pasture for lying times and step count. Principal components analysis (PCA) was used to analyse QBA data in R using parallel analysis.
Results
Behavioural assessment showed animals at pasture displayed more positive behaviours. Positive PC1 scores were observed in 67.5% of pasture cows, where PC1 score refers to mood and runs negative to positive. In comparison, PC1 for housed cows showed 61% of animals had a negative behavioural score. Sensor data, particularly step count and lying time, differed with cattle location - housed or pasture – and correlated with QBA data. For example, step count was positively correlated with “happiness” (p = <0.001, r = 0.4), PC1 score (p = <0.001, r = 0.3) and others. There was significantly greater “clustering” of lying times at pasture (p = 0.004), with cows lying for similar times compared to housed cows. Clustering at pasture was attributed to the potential for cows to lie down at the same time, i.e., lying synchrony. As a result of increased lying space at pasture, cows have the ability to exhibit similar behaviours at the same time, which can be limited indoors by housing and management practices. Behavioural synchrony - e.g., lying and feeding synchrony - has been shown to be a positive welfare indicator in cattle (Fregonesi and Leaver, 2002, Napolitano et al., 2009). A smaller spread of QBA scores was noted at pasture, also suggesting higher levels of synchrony, with more animals exhibiting similar levels of behaviours.
Conclusions
This work suggests that it is possible to use automatically collected sensor data to understand positive welfare in dairy cattle. The potential for an automated link between existing farm management sensor data and positive cattle welfare is the subject of continuing work (https://www.digitaldairychain.co.uk).
Original language | English |
---|---|
Pages (from-to) | 315-315 |
Number of pages | 1 |
Journal | Animal - science proceedings |
Volume | 14 |
Issue number | 2 |
Early online date | 31 Mar 2023 |
DOIs | |
Publication status | Published - 30 Apr 2023 |
Event | British Society of Animal Science 2023: Animal Science: delivering for all our needs - International Convention Centre, Birmingham, United Kingdom Duration: 28 Mar 2023 → 30 Mar 2023 https://bsas.org.uk/conference-2023 |
Keywords
- dairy cattle
- animal welfare
- livestock farming
Fingerprint
Dive into the research topics of 'Use of existing precision livestock farming tools to identify positive welfare in dairy cattle'. Together they form a unique fingerprint.Projects
- 1 Active
-
Digital Dairy Value-Chain for South-West Scotland and Cumbria (Strength in Places Fund)
Michie, C. (Principal Investigator), Wagner, B. (Academic), Wilson, J. (Academic), Andonovic, I. (Co-investigator), Atkinson, R. (Co-investigator), Bellekens, X. (Co-investigator), Galloway, S. (Co-investigator), Kelly, N. (Co-investigator), Lengden, M. (Co-investigator), Revie, M. (Co-investigator), Stewart, R. (Co-investigator), Tachtatzis, C. (Co-investigator) & Ward, A. (Co-investigator)
1/02/22 → 31/01/27
Project: Research