The use of sensors applied to precision livestock farming is widespread in many farm species, especially dairy cattle, but there is a dearth of sensors validated for sheep farming. This study aims to validate a dairy cattle sensor collar to assess sheep ingestion, rumination, and other behaviors in two housing conditions: indoor housed and pasture. Twenty crossbred ewes were continuously monitored for 24 h per system, with video recordings analyzed by trained observers to quantify ingestion, rumination, and other behaviors. Precision (r, R2, Bland–Altman) and accuracy (CCC, regression slope) analyses were undertaken to assess sensor performance. The intra-rater reliability of behavior scoring was good (Kappa = 0.84, p < 0.01). In the indoor experiment, ingestion and rumination behaviors showed high precision (r = 0.92 and 0.79, respectively), while only ingestion time was considered accurate (CCC = 0.91). In the outdoor system, ingestion time showed moderate precision (r = 0.83) and accuracy (CCC = 0.80), whereas rumination and other behaviors presented low agreement with visual observations. The findings suggest that, while current sensors can be used to monitor sheep feeding behavior in confined environments, further refinement in algorithm and collar design is needed for effective application in grazing conditions.
Silva et al. (2026) studied this question.