Predicting diet digestibility and crude protein content from the faeces of grazing cattle

Summary

Supplementation of grazing cattle is becoming increasingly important in the northern beef industry due to a range of production and marketing imperatives. Cost effective supplementation demands a knowledge of the diet quality from pasture alone because the quality of the base diet determines whether supplement is needed (when to feed), what to feed and how much to feed to meet a desired level of performance, as well as the biological and therefore economic response to feeding supplement. Unfortunately, no commercially applicable technology has been available for determining diet quality in grazing cattle. Recent developments in NIRS (near infra-red reflectance spectroscopy), however, have opened up new opportunities and overseas research has demonstrated the potential for predicting diet quality from faecal analysis using NIRS. Project CS.253 was established to develop NIRS calibration equations for predicting diet digestibility and crude protein content from cattle faeces. The experimental methodology involved the collection and processing of faecal samples from both grazing and pen fed cattle and the acquisition and chemical analysis of forage samples representing the grazed and pen fed diets. Calibration equations were developed by determining the regression relationships between dietary reference values and faecal NIR spectra using the appropriate computer software. Additional work was performed to determine the potential of faecal NIRS for predicting forage intake and botanical composition and to resolve technical problems associated with the estimation of in vivo digestibility using in vitro techniques. Project results demonstrated that faecal NIRS can provide accurate predictions of dietary N and dietary C~C4 composition. The relationship between actual and predicted digestibility coefficients was less precise than for dietary N. Although faecal NIRS predictions of forage intake were poor, faecal NIRS showed good potential for predicting the intake of digestible dry matter. Additional data is needed to expand the size and diversity of calibration sample sets to improve the reliability of predictions before equations can be applied commercially with confidence and there is a need for on-going validation and refinement of calibration equations. Nevertheless, the sample sets and equations developed in CS.253 provide a valuable and sound foundation for future expansion and application. Overall, project results indicate that substantial benefits can be expected from faecal NIRS applications in industry and research. Contrary to generally accepted dogma, in vivo digestibility was not a good indicator of forage quality and intake was not closely correlated with digestibility. Pepsin-cellulase in vitro digestibility was a better indicator of forage quality since forage intake was more closely related to in vitro than to in vivo digestibility. Faecal NIRS was a better predictor of the intake of digestible dry matter than any measure of digestibility. Accurate estimates of in vivo digestibility by pepsin-cellulase in vitro analysis were limited by the poor relationship between in vivo and in vitro digestibility. Project data showed that in vivo digestibility could be predicted more accurately by faecal NIRS than by in vitro analysis of dietary forage. The pepsin-cellulase in vitro digestibility of forage was lower than that of comparable extrusa. The difference was greater for grasses than legumes and the difference increased as digestibility decreased. These differences have important implications for estimating in vivo digestibility from in vitro analysis.