Optimising a weighing protocol for sheep

A series of experiments designed to investigate live weight error and analyse the effect fasting and multiple weighing has on live weight measurements were carried out. In the first experiment 24 mixed aged ewes were fasted during a 24 hour period with live weight measurements taken at 0, 2, 4, 6, 8, 10, 12 and 24 hours fasted, with three measurements taken at each time point. At the conclusion of weighing all animals were slaughtered and their gutfill was weighed to enable calculation of their true (digesta-free) weight. In the second experiment 100 Coopworth ewes were weighed as described above on two separate occasions in May and July. On all occasion animals were individually identifiable with either visual ear tags or electronic identification tags (EID). In experiment one live weight of ewes at the start of fasting ranged from 34.4 – 79.6 kg with a mean live weight of 61.3 kg. The mean live weight loss displayed a curvilinear response peaking at 4.1 ± 0.23 kg or 6.7% ± 0.348 at 24h fasting, with a significant effect of time. The regression equations y = -0.0073x² + 0.3445x for absolute live weight loss and y=-0.012x²+0.5636x for proportional live weight loss both explained 98% of the variation. The proportion of true live weight relative to measured live weight over 24 hours fasted increased from 0.85 ± 0.006 to 0.91± 0.004. Despite being fasted for 24h the animals were only able to reach 91% of their true live weight as final digesta ranged from 2.5-7.3 kg with an average of 5.06 ± 0.27 kg. Therefore it was determined that attaining a reliable estimation of an animal’s true body weight is unachievable when fasting them for 24 hours. In experiment two, the live weight of 100 ewes was measured during a 24 hours fasting period with measurements taken at 0, 2, 4, 6, 8, 10, 12 and 24 hours in May which was repeated again in July. The live weight loss peaked at 7.87% in May which was considerably greater than the maximum live weight loss of 3.09% seen in the July experiment. The difference was likely due to a difference in weather conditions, with snowy conditions in July influencing the amount of feed ingested by the animals. The relationship between live weight loss (kg) and (%) ranking at 24 hours fasted and the live weight loss (kg) and (%) at fasting times 2, 4, 6, 8, 10, 12 for experiment 2 (May and July) was analysed. The strongest relationships were seen at 12 hours fasted: y=0.9878x (R² =0.8976); for May, y=0.9567x (R² =0.6623) for July when expressed in kg. As a proportion of weight loss the greatest relationship was seen at 10 hours y=0.716x (R² =0.8578) and y=0.8895x (R² =0.8895) for May and July respectively. Probit analysis to fit 95% of their 24hr weight showed no significant difference was seen between the liveweight range from 0, 2 and 4 hours fasted; 4, 6, 8 hours fasted and 6, 8, 10, 12 hours fasted. In order to analyse the effect of multiple weighing a probit analysis to determine the range in weight which 95% of average the of three weights would fall indicated the only significant difference was at 10hrs fasted between one weight and the average of two weights. In summary, live weight is a fundamental measurement which is used to monitor the production in the sheep industry. However live weight is only an estimate of an animal’s true live weight which is unable to be reliably determined even following 24h fasting due to the variation in digesta weights. Relative to their 24h fasted live weights, fasting improved the accuracy and precision of the live weight estimate with little advantage observed when animals were fasting for periods greater than 8h. The method of multiple weighing had minimal effect on improving accuracy regardless of fasting time