Variability of water intake in dairy cows for the assessment of water efficiency

Feed efficiency and precision feeding strategies are essential to reduce the environmental impact of dairy and beef sector. Water is a precious resource, and because cattle are far from being considered as a waterneutral livestock, identification of superior animals is advisable. Water efficiency can be described by the amount of water ingested per kg of milk or meat produced. In this study, we investigated phenotypic effects affecting daily water intake (WI) and water efficiency index (EFF1) calculated as ratio of WI to milk yield (MY). Similarly, the EFF2 was calculated by using WI predicted by reticular boli manufactured by smaXtec animal care GmbH. Twenty-four Holstein cows with bolus inserted, 7 primiparous and 17 multiparous, monitored for 28 d (Jan.–May 2024) at the University of Bologna experimental farm. Every day individual WI at each bout and MY at each milking were recorded through inhouse developed drinkers. To study the phenotypic variability of the traits we used a linear mixed model where cow was the random effect and parity (1, ≥2), lactation stage (≤80, 81–100, ≥101 d in milk) and their interaction were fixed. Measured WI and predicted WI averaged 183.0 and 176.3 L/d. Primiparous drank less than multiparous (155.4 vs. 194.2 L/d; P < 0.001) and WI increased with days in milk (166.0,175.8, and 182.5 L/d; P < 0.001). As regards EFF1, primiparous were the most efficient (3.4 vs. 3.8 L/kg milk; P < 0.001) as well as cows in ≤80 d in milk (3.2 L/kg milk; P < 0.001). None of the fixed effects were significant for EFF2, although trends for parity and lactation stage were numerically similar to EFF1. The Pearson correlation between EFF1 and EFF2 was 0.49 (P < 0.001) and that between measured and predicted WI was 0.24 (P < 0.001). MY was significantly correlated with both EFF1 (−0.44) and EFF2 (−0.46). Results indicate that WI is variable and, at the same time, suggest that there is potential to select water-efficient cows without compromising MY. Further research may explore genetic aspects and inclusion of WI in the definition of feed efficiency. Finally, sensors for large-scale accurate phenotyping of WI are also needed.