Time-Budget of Horses Reared for Meat Production: Influence of Stocking Density on Behavioural Activities and Subsequent Welfare

Abstract

Horses reared for meat production can be kept in intensive breeding farms where they are housed in group pens at high stocking densities. The present study aimed to evaluate whether the expressed behaviours correlated with stocking density, and to compare their time-budget with that of wild-living horses. An ethogram of 13 mutually exclusive behavioural activities was developed. Behavioural observations were performed over a 72 h period on group pens selected on the basis of stocking density and the homogeneity of breed, age, height at the withers, and time since arriving at the farm. Scan sampling (n = 96 scans/horse/day) was used on 22 horses. The mean frequency (%) ± standard deviation (±SD) for each behavioural activity was calculated to obtain the time-budget. The associations between time-budget and stocking density were evaluated using a bivariate analysis. The relationships were analysed by Pearson’s correlation coefficient (r). Our results show that locomotion, playing, and self-grooming positively correlated with a reduction in stocking density, indicating the potential to use these behaviours as positive welfare indicators for young horses kept in group pens. The data also revealed an unusual time-budget, where the main behavioural activity expressed was standing (30.56% ± 6.56%), followed by feeding (30.55% ± 3.59%), lying (27.33% ± 2.05%), and locomotion (4.07% ± 1.06%).

Activity Time Budgets—A Potential Tool to Monitor Equine Welfare?

“…After removing duplicates, reviews, commentaries, single case reports, books and non-English or German articles, 646 papers remained. Following the exclusion of papers that did not focus on equine activity tracking or time budgets, but on different species or other behavioral observations or used an observation period of less than 24 h per day, 12 articles remained included in the qualitative synthesis [ 18 , 28 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ]. A total of 6 (50% of the total) articles were classified as observational studies [ 18 , 28 , 37 , 38 , 39 , 45 ], 5 articles (41.6% of the total) as prospective, non-blinded, non-randomized case series [ 41 , 42 , 43 , 44 , 46 ], 1 article (8%) as prospective, non-blinded, non-randomized controlled trial [ 40 ] and none as randomized controlled trials ( Table 1 ).…”

“…Risk-of-bias assessment ( Table 1 ) revealed the lack of a control or comparison group (only 8.3% of the articles fulfilled this criterion), random assignment of participants to intervention (8.3% of the articles fulfilled this criterion), random selection of participants for assessment (none of the articles fulfilled this criterion), as the most critical concerns. Further limitations of the papers included in this review are the small sample sizes (4–22 horses) and the variable observation methods which were restricted to manual observation in the field in 33% [ 18 , 37 , 38 , 39 ] and manual behavior scoring from video in 50% [ 40 , 41 , 42 , 43 , 45 , 46 ]. Only 25% used biotelemetry devices (25%) [ 28 , 44 , 45 ].…”

“…Only 25% used biotelemetry devices (25%) [ 28 , 44 , 45 ]. Manual quantification of observed behavior was carried out in 83.3% (10/12 studies) of the studies using scan sampling in 33% (4/12 studies) [ 18 , 37 , 38 , 46 ], focal sampling [ 39 , 43 ], and instantaneous sampling in 16.6% (2/12 studies) [ 40 , 41 ] each, and ad libitum sampling [ 45 ] and unspecified methodology [ 42 ] in one paper. In addition, the ethogram, based on which behaviors were categorized, was not provided in 25% [ 39 , 41 , 42 ], and the video analysis software was not detailed in 41.6 % [ 40 , 42 , 44 , 45 , 46 ] of the articles.…”

“…Manual quantification of observed behavior was carried out in 83.3% (10/12 studies) of the studies using scan sampling in 33% (4/12 studies) [ 18 , 37 , 38 , 46 ], focal sampling [ 39 , 43 ], and instantaneous sampling in 16.6% (2/12 studies) [ 40 , 41 ] each, and ad libitum sampling [ 45 ] and unspecified methodology [ 42 ] in one paper. In addition, the ethogram, based on which behaviors were categorized, was not provided in 25% [ 39 , 41 , 42 ], and the video analysis software was not detailed in 41.6 % [ 40 , 42 , 44 , 45 , 46 ] of the articles. Data analysis was limited to descriptive statistics in the observational studies [ 18 , 28 , 37 , 38 , 39 , 45 ] and the other 50% of the included papers detailed further statistics [ 40 , 41 , 42 , 43 , 44 , 46 ].…”

“…In addition, the ethogram, based on which behaviors were categorized, was not provided in 25% [ 39 , 41 , 42 ], and the video analysis software was not detailed in 41.6 % [ 40 , 42 , 44 , 45 , 46 ] of the articles. Data analysis was limited to descriptive statistics in the observational studies [ 18 , 28 , 37 , 38 , 39 , 45 ] and the other 50% of the included papers detailed further statistics [ 40 , 41 , 42 , 43 , 44 , 46 ]. The large variance in study methodology and data presentation did not allow combining the data for meta-analysis.…”