Abstract #M118

# M118
Investigation into the accuracy of a commercially available activity meter for measuring grazing duration.
Emer Kennedy*1, James Moloney1, Donagh P. Berry1, Michelle Liddane1, Frank Buckley1, 1Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland.

Grass intake is influenced by grazing duration. Current methods of measuring grazing behavior are best suited to research environments. Activity meters, used primarily for oestrus detection, are now incorporating behavioral measures. The hypothesis was that an activity meter would provide similar information on grazing duration to that generated by a behavior recorder. One hundred 24-h records were obtained from grazing Holstein-Friesian cows between August 26 and September 30, 2014. Grazing time was recorded using IGER behavior recorders (IGER), which were processed using the ‘Graze’ analysis software. All cows were fitted with MooMonitor+ activity meters (MOO; Dairymaster, Causeway, Co. Kerry). Total grazing time data from the MOO was obtained from the Dairymaster database. Total 24-h grazing duration determined by the IGER was linearly regressed on the MOO grazing duration to determine the association between each method; no intercept was fitted in the model. A fixed effects linear model was used to quantify the association between parity, measurement day and IGER recorder number on the IGER recorded grazing duration. Two-way interactions between the MOO grazing duration and the main fixed effects were considered in the model to determine if the association between the MOO and IGER grazing duration differed by any of the main effects. Mean 24-h grazing duration was 543 (±6.39 SEM) min for IGER and 540 (±5.9 SEM) min for MOO. A correlation of r = 0.72 existed between the IGER and MOO grazing durations. The regression coefficient of IGER grazing duration on MOO grazing duration was 1.003 (±0.008 standard error) which was not different from the expectation of unity should the measured differences in grazing duration between the 2 devices be equivalent. Neither parity nor calendar day was associated with the difference in grazing duration estimated by the 2 devices. The regression coefficient of the IGER on the MOO grazing duration varied from 0.87 to 1.1. The results from this study, albeit from a limited data set, indicate good concordance between 24-h grazing duration estimated from the IGER and MOO recorders.

Key Words: grazing behavior, dairy cow, activity meter