Assessment of RumiWatch noseband sensors for the quantification of ingestive behaviors of dairy cows at grazing or fed in stalls

The RumiWatch System (Itin+Hoch GmbH, Liestal, Switzerland) is a sensor system that enables quantification of ingestive behaviors of cattle that may be useful, amongst other things, for estimating dry matter intake. There are currently multiple versions of the RumiWatch Converter software used to convert the recorded data into specific animal behaviors through the use of algorithms. The objective of this research was to evaluate the use of RumiWatch System noseband sensors for quantifying the number of prehension bites and mastication chews in grazing and stall-fed dairy cows using RumiWatch Converter software versions 0.7.3.36 (RWC36) and 0.7.4.13 (RWC13). Two experiments using lactating Holstein-Friesian dairy cattle were conducted. In Experiment I, ten cows grazed perennial ryegrass pasture (Lolium perenne L.) in individual plots that measured 35 m2, with a pasture mass of 2500 kg DM/ha and pasture allowance of 28 kg DM/d. In Experiment II, 14 cows were offered 16 kg DM/d of perennial ryegrass pasture silage in individual indoor feed stalls. In both experiments, ingestive behaviors were quantified with the jaw movement halters and through visual observation using video recordings. The performance of the RumiWatch Converter software was evaluated through Lin’s concordance correlation coefficients with video observation. For grazing cows, there was greater concordance between the RumiWatch System and video observation for the number of prehension bites (0.94 for RWC36 and 0.75 for RWC13) than for the number of mastication chews (0.20 for RWC36 and 0.14 for RWC13). In cows fed in stalls, there was also greater concordance for the number of prehension bites (0.54 for RWC36 and 0.64 for RWC13) than for the number of mastication chews (0.46 for RWC36 and 0.58 for RWC13). However, the RWC36 was more accurate for quantifying feeding behaviors of grazing cows than the RWC13, while for stall fed cows the reverse was true. Overall, both RumiWatch Converters were better at quantifying prehension bites in grazing cows than in stall fed cows, but better at quantifying mastication chews in stall fed cows than grazing cows. These data enable a new understanding of the how these RumiWatch sensors should best be used for measuring cows' feeding behaviors in different feeding systems.