TY - ABST
T1 - Validation of eating duration assessment from automatic feeder systems
AU - Ternman, Emma Matilda
AU - Terré, M.
AU - Bouchon, Matthieu
AU - Meunier, B.
AU - Munksgaard, Lene
AU - Veissier, I
PY - 2021/8
Y1 - 2021/8
N2 - The aim of the study was to test if automatic recordings of when a cow enters and exits a feed bin were comparable with visual observations of the same, and how well total visit duration corresponds to eating duration. Three feed binsystems were tested: Insentec Roughage Intake Control (RIC); BioControl Control and Record Feed Intake (CRFI); MooSystem Intake Control Feeding System (MF). All cows were used to the systems before the study started. The cows’ behaviour was recorded using direct observations by a person standing in front of the bins (RIC, MF) or from video recordings (BC). Eating duration/visit was the summary of head angled down, taking a bite and chewing. Total visit duration included timestamps of the cow entering the bin until exiting. Observations with visit durations <10 s (biologically irrelevant visits) or >2,333 s (1% longest durations) were considered outliers and removed. We divided visits into three classes according to their observed duration: short <150 s; medium 150-599 s; long ≥600 s. To analyse how accurate the system can predict total visit duration, and whether the total visit duration can be used to predict eating duration, the observed duration of visits and duration of eating were tested in mixed models, with class of visit duration, the type of feed bin system, and the interaction between the two as fixed effects. Cow was included as random effect. The analysis was repeated for short, medium and long visits. Duration of a visit to the feeder can be estimated with very high accuracy in all three systems. The mean difference between observed and system duration was -0.21±0.14 s for RIC, -2.63±0.29 s for BC and 0.40±0.19 s for MF and differed between systems (P<0.01). Duration of eating can also be estimated with high accuracy, slope was 1.00±0.035 for short, 0.98±0.016 for medium and 1.00±0.011 for long visits. Cows with RIC spent more time not eating when visiting the feeders than cows in BC or MF, which could be due to the feeder design.
AB - The aim of the study was to test if automatic recordings of when a cow enters and exits a feed bin were comparable with visual observations of the same, and how well total visit duration corresponds to eating duration. Three feed binsystems were tested: Insentec Roughage Intake Control (RIC); BioControl Control and Record Feed Intake (CRFI); MooSystem Intake Control Feeding System (MF). All cows were used to the systems before the study started. The cows’ behaviour was recorded using direct observations by a person standing in front of the bins (RIC, MF) or from video recordings (BC). Eating duration/visit was the summary of head angled down, taking a bite and chewing. Total visit duration included timestamps of the cow entering the bin until exiting. Observations with visit durations <10 s (biologically irrelevant visits) or >2,333 s (1% longest durations) were considered outliers and removed. We divided visits into three classes according to their observed duration: short <150 s; medium 150-599 s; long ≥600 s. To analyse how accurate the system can predict total visit duration, and whether the total visit duration can be used to predict eating duration, the observed duration of visits and duration of eating were tested in mixed models, with class of visit duration, the type of feed bin system, and the interaction between the two as fixed effects. Cow was included as random effect. The analysis was repeated for short, medium and long visits. Duration of a visit to the feeder can be estimated with very high accuracy in all three systems. The mean difference between observed and system duration was -0.21±0.14 s for RIC, -2.63±0.29 s for BC and 0.40±0.19 s for MF and differed between systems (P<0.01). Duration of eating can also be estimated with high accuracy, slope was 1.00±0.035 for short, 0.98±0.016 for medium and 1.00±0.011 for long visits. Cows with RIC spent more time not eating when visiting the feeders than cows in BC or MF, which could be due to the feeder design.
U2 - 10.3920/978-90-8686-918-3
DO - 10.3920/978-90-8686-918-3
M3 - Conference abstract in proceedings
SN - 978-90-8686-366-2
T3 - EAAP Book of Abstracts
SP - 613
EP - 613
BT - Book of Abstracts of the 72nd Annual Meeting of the European Federation of Animal Science
A2 - Strandberg, E.
A2 - Pinotti, L.
A2 - Messori, S.
A2 - Kenny, D.
A2 - Lee, M.
A2 - Hocquette, J.F.
A2 - Cadavez, V.A.P.
A2 - Millet, S.
A2 - Evans, R.
A2 - Veldkamp, T.
A2 - Pastell, M.
A2 - Pollott, G.
PB - Wageningen Academic Publishers
T2 - 72nd Annual Meeting of the European Federation of Animal Science
Y2 - 29 August 2021 through 3 September 2021
ER -