Fodder quality and intake by dairy cows. 1. Preference for oaten hays

Abstract

This work was undertaken to determine whether measurement of preference 30 min after alternative hays are offered to lactating dairy cows adequately predicts preference over a 3-day period and to determine the influence of the concentrations of water-soluble carbon (WSC) and neutral detergent fibre (NDF) of hays on preference. Eleven test hays were chosen to provide an NDF concentration of between 45% and 60% and a WSC concentration of between 10% and 30%. The test hays were each offered to lactating dairy cows as an alternative choice, with two control hays with NDF concentrations of 49% and 54%. A replicated Latin-square design was used to determine the preference of cows for the test hays, using three cows for each test hay versus control-hay comparison. Preference measured 30 min after hays were offered to cows was highly correlated with measures of preference made after 24 h and again for the same measures on the third day of the hays being offered. However, preference measures with the two different control hays were not well related, thus indicating that preference values are control-hay specific. Total hay intake was unchanged with the low-NDF control hay but declined with increasing NDF concentration of the test hays with the higher-NDF control hay. WSC did not influence preference at any one level of NDF of the test hays when the control hay had a low NDF concentration, but there was a small rise in preference at each NDF level when the control hay had a higher NDF concentration. This may have been a result of a contribution of WSC to digestibility as much as to taste response. With the low-NDF control hay, drivers of preference were the NDF concentration and the digestibility of the hay, but with the higher-NDF control hay, the drivers were the NDF concentration and fibre characteristics (shear and potential fibre digestibility). Calculation of the NDF concentration of the total hay consumed, test plus control hays, suggested that a change in drivers of selection is likely at the point where the NDF concentration of the test hay equals that of the control hay. Unless models identifying selection drivers handle non-linear data, misleading results may be obtained.