An assessment of Walk-over-Weighing to estimate short-term individual forage intake in sheep

The main limitation for determining feed efficiency of freely grazing ruminants is measurement of daily individual feed intake. This paper describes an investigation that assessed a method for estimating intake of forage based on changes in BW of ewes. A total of 24 dry and non-pregnant Romane ewes (12 hoggets, HOG; mean±SD 51.8±2.8 kg BW; body condition score (BCS) 2.6±0.2; and 12 adults, ADU; 60.4±8.5 kg BW; BCS 2.7±0.8) were selected for the study and moved from their rangeland system to a confined pen with controlled conditions and equipped with individual automatic feeders. The experiment lasted for 28 days (21 days adaptation and 7 days feed intake measurement). Ewes were fed hay and trained to use the electronic feeders (one feeding station per ewe) in which actual daily intake (H_intake24) was measured. The pens were designed to maximize movement of trained ewes through an automated Walk-over-Weighing device, by using water and mineral salts as attractants. Total individual intake of hay measured in the automatic feeder at each meal (H_intake) was compared with indirect estimates of feed intake determined using differences in the BW of the ewes (∆BW) before and 1 h following morning and afternoon feeding at fixed times. The BW, BCS, H_intake, H_intake24, as well as plasma non-esterified fatty acids (NEFA), glucose and insulin profiles were determined. The BW was higher in ADU v. HOG but BCS was not affected by parity. The H_intake24 was affected by day of experiment as a consequence of reduced availability and intake of water on one day. Plasma glucose, NEFA and insulin were not affected by parity or day of experiment. The H_Intake was and ∆BW tended to be higher in the morning in HOG, whereas H_intake was and ∆BW tended to be higher in ADU at the afternoon meal. Irrespective of parity or feeding time, there was very strong correlation (r²=0.93) between H_intake and ∆BW. This relationship confirms that our indirect method of estimating individual forage intake was reliable within the strictly controlled conditions of the present experiment. The method appears suitable for use in short-term intensive group feeding situations, and has potential to be further developed for longer-term forage intake studies, with a view to developing a method for freely grazing ruminants.