Prediction of faecal output and hay intake by cattle from NIRS estimates of faecal concentrations of orally-dosed polyethyleneglycol

Polyethyleneglycol 6000 (PEG) was selected as a potential indigestible estimator of faecal output in cattle. NIRS-calibration equations of faecal PEG content were obtained with faeces from hay-fed cows added with 0.00–0.10 g PEG content/g faeces. A digestibility trial was conducted with 8 dry cows in digestibility cages for two 3-week periods, during which half the cows received either a Low (9 kg hay, as fed) or High (12 kg hay) feeding level (FL), and were dosed daily during the last 10 days either 175 (Low) or 235 (High FL) g PEG/d. In the last 4 days total faeces were collected, and on the last day rectal samples were grabbed every 4 h during 24 h. Samples were NIRS-scanned to determine faecal PEG content. Using data from total faecal collection, PEG recovery was not complete (0.957 g/g, s.e. 0.010), therefore faecal output was slightly overestimated with PEG data from total faeces (+0.052 g/g, s.e. 0.089), and so was calculated feed intake, although only significantly in the High FL (+0.060 g/g, s.e. 0.021), while estimates were accurate in the Low FL. Recovery was also incomplete in grab samples (0.928 g/g, s.e. 0.039), and faecal PEG content had a large per cow within-day variability (0.40), and the bias of estimation of faecal output was high at any sampling time. The use of a daily composite of all grab samples per-cow reduced the variability of faecal PEG contents, and bias of estimations was below 0.10. It is concluded that dosing PEG and analysing PEG faecal contents with NIRS provides reasonably accurate estimates of faecal output and forage intake in cattle, provided faecal samples are representative of total faeces produced. This implies frequent faecal grab sampling, including nocturnal collection, which may limit the practical use of this technique under field conditions.