Proximity Interactions in a Permanently Housed Dairy Herd: Network Structure, Consistency, and Individual Differences

Abstract

Understanding the herd structure of housed dairy cows has the potential to reveal preferential interactions, detect changes in behavior indicative of illness, and optimize farm management regimes. This study investigated the structure and consistency of the proximity interaction network of a permanently housed commercial dairy herd throughout October 2014, using data collected from a wireless local positioning system. Herd-level networks were determined from sustained proximity interactions (pairs of cows continuously within three meters for 60 s or longer), and assessed for social differentiation, temporal stability, and the influence of individual-level characteristics such as lameness, parity, and days in milk. We determined the level of inter-individual variation in proximity interactions across the full barn housing, and for specific functional zones within it (feeding, non-feeding). The observed networks were highly connected and temporally varied, with significant preferential assortment, and inter-individual variation in daily interactions in the non-feeding zone. We found no clear social assortment by lameness, parity, or days in milk. Our study demonstrates the potential benefits of automated tracking technology to monitor the proximity interactions of individual animals within large, commercially relevant groups of livestock.

Exploring the Potential of Precision Livestock Farming Technologies to Help Address Farm Animal Welfare

“…In cattle, systems have been developed to monitor agonistic behaviors based on sensors (Foris et al, 2019), while image-based technologies can monitor mounting behaviors (Chung et al, 2015;Guo et al, 2019) and social interactions (Guzhva et al, 2016), and accelerometers can estimate locomotor play in calves (Luu et al, 2013). Proximity interactions of individual dairy cows within large herds can also be monitored using local positioning sensor network (Chopra et al, 2020). Imageand RFID-based technologies in the poultry sector allow to monitor human-animal interactions (HAI) (Lian et al, 2019), nesting (Li et al, 2017), and perching behaviors (Nakarmi et al, 2014;Wang et al, 2019).…”

An Information-Theoretic Approach to Detect the Associations of GPS-Tracked Heifers in Pasture

“…The importance of understanding and decoding the social structures of livestock becomes increasingly important in modern farming [1,2], since the knowledge can help to make species-appropriate management decisions [2][3][4][5]. Social network analysis has been performed to establish the structures of agonistic behavior in pig groups [6][7][8][9], to estimate the spread of diseases in cattle [10][11][12], and to determine the social structures or agonistic behavior of cattle [10,[13][14][15][16][17].…”

“…However, especially for position data such as GNSS, the identification of real interactions is a problem, and associations of animals have to be estimated based on their pairwise distances [1,4]. Often, distance thresholds are determined beyond which two animals are defined to have a contact [1,[13][14][15]17].…”

“…However, the results strongly depend on the chosen null model (also known as the background model), a procedure that generates datasets (i.e., by randomization), which in turn, can be used to compare the original data to [25]. Although for some applications, strategies exist to define a proper null model [4,[25][26][27], for other problems (newer problems such as GNSS data), determining the most suitable null model remains challenging [26]. Especially for GNSS data, null model construction by randomizing the raw coordinate data is difficult due to the fact that the autocorrelation of the individual tracks has to be considered and the walking paths of animals should not be interrupted [25].…”

Where do we find missing data in a commercial real-time location system? Evidence from 2 dairy farms

“…Loss of data in the feeding area is an important consideration if a researcher is interested in analyzing proximity networks and social behavior because a large number of displacements and allogrooming occur among cows in the feeding area, sometimes over short periods (Miller and Wood-Gush, 1991;Val-Laillet et al, 2009). Filters to smooth RTLS data have been applied in earlier studies of social interactions (Chopra et al, 2020;Rocha et al, 2020), but the effect of missing data has not, to the best of our knowledge, been investigated in detail. When using a social network graph to analyze proximity interactions, missing observations of an individual can dramatically change the structure of the network graph, especially for nodes that have high betweenness (Krause et al, 2007).…”