Analysis of Cattle Social Transitional Behaviour: Attraction and Repulsion

Abstract

Understanding social interactions in livestock groups could improve management practices, but this can be difficult and time-consuming using traditional methods of live observations and video recordings. Sensor technologies and machine learning techniques could provide insight not previously possible. In this study, based on the animals’ location information acquired by a new cooperative wireless localisation system, unsupervised machine learning approaches were performed to identify the social structure of a small group of cattle yearlings (n=10) and the social behaviour of an individual. The paper first defined the affinity between an animal pair based on the ranks of their distance. Unsupervised clustering algorithms were then performed, including K-means clustering and agglomerative hierarchical clustering. In particular, K-means clustering was applied based on logical and physical distance. By comparing the clustering result based on logical distance and physical distance, the leader animals and the influence of an individual in a herd of cattle were identified, which provides valuable information for studying the behaviour of animal herds. Improvements in device robustness and replication of this work would confirm the practical application of this technology and analysis methodologies.

A Case Study on the Use of Virtual Fencing to Intensively Graze Angus Heifers Using Moving Front and Back-Fences

“…This consistency in how individual heifers interact with the virtual fence across contexts could point to an underlying temperamental predisposition. Cattle vary in their willingness to spend time away from the herd (Hirata et al, 2013;Xu et al, 2020), while ecological studies show that bolder, asocial and exploratory individuals are more likely to disperse from their resident social group (reviewed by Cote et al, 2014). Verdon et al (2020) found bold heifers were more likely to ignore the audio and electrical stimuli delivered in a feed attractant trial.…”

“…Frequency of cues delivered to heifer 9 may thus be higher in allocation 3 than reported. movement patterns than others (i.e., herd "leaders"- Reinhardt and Reinhardt, 1981;Dumont et al, 2005;Ramseyer et al, 2009;Keshavarzi et al, 2020;Xu et al, 2020). Leadership is not synonymous with dominance, but group leaders are likely to be animals that are less sociable, more bold and more explorative (Stricklin and Kautz-Scanavy, 1984;Ramseyer et al, 2009;Della-Rossa et al, 2013;Neave et al, 2018).…”

Validation of Real-Time Kinematic (RTK) Devices on Sheep to Detect Grazing Movement Leaders and Social Networks in Merino Ewes

“…With the development of on-animal sensors and technologies such as GPS devices [ 14 , 15 ], proximity loggers [ 9 , 16 ], or ultra-wideband positional loggers [ 17 , 18 ] there is the potential to deploy these devices on livestock individuals within groups to enable more accurate monitoring of position and/or social relationships. The data from these sensors can provide new insights into how individuals in a group interact and/or influence each other, including affiliative and/or agonistic relationships between group members [ 19 , 20 ], network structure [ 19 ], and resource-use patterns [ 21 ]. Commercially, sensor technologies that allow quantification of social interactions may, for example, enable understanding of how animals learn new technologies such as virtual fencing [ 22 ], detect male-female interactions to determine oestrus [ 23 ] and mating [ 24 ], allow maternal pedigree detection through ewe-lamb contact [ 25 ], or could monitor grazing behaviour [ 26 ].…”

“…Therefore, more frequent and accurate data collection to capture all possible social interactions will result in more precise sub-groupings and allow detection of the social network structure within the group. In addition, more precise data may enable researchers to quantify interactions in situations where animals are all in close physical contact [ 20 ]. While GPS tracking has been used to quantify social relationships between livestock animals [ 30 , 31 , 32 ], GPS typically has a high spatial precision error; one study showed an overestimation of 15.2% or 1.5 km for daily cattle travels without any data filtering [ 33 ], another study showed a contact distance error of 9.5 m with prototype proximity-logging GPS collars on bighorn sheep [ 34 ].…”

Livestock Informatics Toolkit: A Case Study in Visually Characterizing Complex Behavioral Patterns across Multiple Sensor Platforms, Using Novel Unsupervised Machine Learning and Information Theoretic Approaches

“…Clustering algorithms, by virtue of their incredible flexibility, have successfully been applied to a range of PLF data streams [ 7 , 13 , 14 , 15 , 16 , 17 , 18 ]. In our own previous work, we have highlighted the utility of hierarchical clustering-based approaches in leveraging the behavioral co-dependencies of cows housed socially in large groups, in a production environment, in order to recover complex temporal patterns in behavior [ 7 ].…”