An automatic system to monitor lying, standing and walking behaviour of grazing animals

Sensors incorporating mercury tilt switches were used to measure lying, standing and walking behaviour of grazing cattle and sheep. The lying/standing sensor was suspended below the animal, whilst the walking sensor was fitted to the shin of the animal's foreleg. Both switches gave a digital-low output when vertical or near vertical and a digital-high output when more than 45 degrees from vertical. The lying/standing sensor was operated when the animal lay down and the walking sensor was operated by leg movement, and also by the animal lying down. The sensor outputs were connected to digital recorders mounted on the animals. Sheep fitted with lying/standing recording equipment were observed every 5 min during daylight over 24 h. The index of concordance between the automatic and manual observations of lying/standing was 0.953. Cattle and sheep, fitted with equipment to record walking, were filmed by video camera for at least 10 min each. The number of steps counted by observers of the video recordings were compared with results from the automatic system. Mean square prediction error (MSPE) analysis of the automatic-system results showed they had an error of 7.7% for both cattle and sheep. This error was a percentage of the mean of the actual observed results. Cattle fitted with equipment to record lying/standing and walking were observed every 10 min during daylight over 24 h. The concordance between automatic and manual observations of lying/standing was 0.990 using the lying/standing sensor and 0.996 using the walking sensor. The presence or absence of walking behaviour was also noted and the concordance for the automatic system results with this was 0.96. The automatic system described provides a reliable means of recording lying, standing and walking behaviour of cattle and sheep over 24 h periods. (C) 1997 Elsevier Science B.V.