Indirect estimation of feed intake in dairy cows on pasture based on urine volume or water consumption

Sammanfattning: There is a need to be able to estimate the feed intake of dairy cows on pasture. Since feed intake is correlated to factors such as the drinking water intake and urine volume, the possibility to estimate the pasture intake based on those parameters were investigated. It was also investigated whether the pasture intake could be predicted from the estimated K intake based on the estimated urine volume when the K concentration in the feed was known. High yielding dairy cows of the breeds Swedish Holstein (SH) and Swedish Red (SR) were given access to fresh pasture or an exercise area during daytime and kept indoors at night. The cows in the experiment were divided into two treatment groups; one experimental group with an unknown pasture intake offered new production pasture daily and one control group that only had access to a small exercise area. Both groups were offered concentrate supplementation at the same level according to the lactation curve, but the group with exercise pasture has access to full indoor feeding ad libitum both day and night while the group with production pasture had a restricted silage ration (6 kg DM) given only at night. The feed intake of silage and concentrates indoors was registered individually in both treatment groups. Intake of feed was therefore known in the exercise group while the intake in the pasture group had an unknown element: pasture intake during daytime. The daily water intake was registered and measured by water meters installed in the water bowls in the feeding area. The daily urine volume was predicted in each cow by collecting urine spot samples that were analysed for the content of urea and creatinine. The urine volume was thus estimated based on the cow’s weight and creatinine concentration in the urine. Grass samples from pasture as well as silage and concentrate samples were analysed to determine the content of dry matter (DM), crude protein (CP), ash, energy, potassium (K) and other minerals. Based on dry matter intake (DMI), mineral intake, water intake and urine volume in the control group, simple linear regression and mixed linear regression were made to predict the total feed intake in the experimental group and hence pasture intake by subtracting the known amount of silage and concentrate intake. Drinking water intake (L/day) and urine volume (L/day) in the control group was plotted against the DMI (kg/day) by simple linear regression and gave the equations (y = 0.167 x + 8.66; R² = 0.448) based on the water intake and (y = 0.459 x + 11.83; R² = 0.307) based on the urine volume. When plotting the K intake (g/day) against the estimated urine volume (L/day) by simple linear regression the following equation could be derived; (y = 9.321x + 255; R² = 0.334) and by using mixed linear regression the equation resulted in (y = 5.036x + 366.9; R² = 0.468). Based on those equations the pasture DMI was estimated in the experimental group and gave reasonable intake volumes with the simple linear regression based on drinking water intake and urine volume and also by the simple and the mixed linear model based on the estimated intake of K. Which method that is preferred depends on the conditions since collection of urine spot samples may be more tediously compared to automatic registrations of drinking water intake from water bowls.