Abstract
This study was performed to evaluate the accuracy and precision of an automated water trough system (Intergado, Pro-Eficiencia Solucao para Agronegocios S.A., Betim, Minas Gerais, Brazil) for measuring water intake (WI) of beef cattle. The dataset was obtained from thirteen Angus x Hereford registered crossbred breeding bulls over a 90 d research trial. Animals were consuming beardless wheat (Triticum aestivum) hay near maintenance requirements with free access to water and a trace mineral block. The animals were assigned to 1 of 2 shaded pens provided with a total of eight automated water trough systems. For each voluntary visit to the water trough, the system recorded the animal identification, water consumed, and initial BW. The final BW was determined following each drinking event by accounting for the weight of water disappeared upon drinking (±0.050 kg), scale recorded weights (±0.500 kg), and flow of water consumed (L/s). Data were analyzed based on plotting observed against predicted for the identity line (slope = 0 and intercept = 1) and goodness-of-fit, both were assessed using the Model Evaluation System (Tedeschi 2006). Differences were considered significant at P ≤ 0.10. The intercept was found to be different than zero (P < 0.001) but slope was similar to 1 (P = 0.0658). Differences were attributed to the inherent precision of the scales. The WI values presented a mean bias of 0.140 kg. The predicted WI accounted for 99.9% of the observed WI. Goodness-of-fit indicated very high values for modeling efficiency (0.999), accuracy (0.999), and the overall measurement of precision and accuracy (CCC; 0.998). Additionally, the major part of the prediction error was random (89%), indicating that only 11% of the error was associated with the prediction technique. Therefore, the Intergado automated water trough system is recommended for the monitoring of individual water intake for beef cattle.
Smoothing Secant Line Slope Using Aggregation Fischer Burmeister Function