Forage quality: innovative analytical approaches and practical applications for dairy cattle feeding

This doctoral thesis investigates the potential of innovative analytical approaches for the determination of forage quality and nutritional value as well as their applications in dairy cattle feeding. Three trials were conducted to provide a comprehensive discussion of the different techniques tested and their potential applications from an economic perspective in forage evaluation. In Trial 1, the prediction ability of a hand-held NIR spectroscopy device was tested to determine dry matter (DM), ash, crude protein (CP), ether extract (EE), amylase treated neutral detergent fiber (aNDF), neutral detergent insoluble crude protein (NDICP), acid detergent fiber (ADF), acid detergent insoluble crude protein (ADICP), acid detergent lignin (ADL), non-fiber carbohydrates (NFC) and digestible neutral detergent fiber at different time points (NDFD12, NDFD30, NDFD120, NDFD240), net energy at production levels of intake – with calculated NDFD at 48 hours – (NELpEst), net energy at production levels of intake – with in vitro NDFD determination at 48 hours – (NELpV), relative feed value (RFV) and relative forage quality (RFQ) of ground hay forages (alfalfa and mixed hay). The highest predictive performance was observed for CP determination (R2 = 0.86), while the lowest was observed for EE determination (R2 = 0.27). Good predictive performances were also found for the aNDF and ADF determination with R2 of 0.86 and 0.74, respectively. In Trial 2, the prediction ability of two portable NIR spectroscopy devices was tested to determine DM, CP, ADF, aNDF and ash content of two different unground forage categories (alfalfa and mixed hay). The potential of sample fractionation by sieving through the use of the PSPS to improve the device's prediction ability for alfalfa hay categories was investigated. The predictive performance increased for CP (R2 increased from 0.37 to 0.86), ADF (R2 increased from 0.51 to 0.63), aNDF (R2 increased from 0.40 to 0.58), and ash (R2 increased from 0.11 to 0.38). On the other hand, a lower performance was observed for DM (R2 decreased from 0.87 to 0.38). In Trial 3, the prediction ability of a third hand-held NIR spectroscopy device was tested to discriminate the botanical composition of pasture (legumes vs grass) from fresh pasture mixtures. Two scanning methods (stationary and sliding) were tested to determine the best solution for calibration purposes. The best performance was found when the sliding mode and a PLS calibration model were applied (R2 = 0.68) while the worst performance was found when the stationary method was combined with a SVM calibration model (R2 = 0.16). In the application chapter, a new tool for determining the economic value of forages, the forage value as production factor (FVPF), was developed. Specifically, the FVPF was created starting from the NELp and NEL equations of the NRC 2001 and by adding adjustment coefficients such as milk price (MilkPx), Feed costs (FeedC), Milk production costs (MPC), dietary concentrate proportion (DConP), dietary forage proportion (DFP) and the concentrate price (ConPx). Two different FVPF (FVPFEst and FVPFV) were developed based on the different NDFD method determination used (calculated and in vitro, respectively). After comparing both FVPFEst and FVPFV values with forage market prices across different years (from 2019 to 2022), it was found that FVPF showed strong potential to valorize forages in the dairy farm context by covering a wider range of values, complementing at the same time market prices, which normally are not adjusted based on forage quality and characteristics. The NIR spectroscopy implemented through portable devices represents an innovative analytical method to be applied for forage analysis in dairy cattle feeding. Despite the technical issues to be solved in the calibration phase of NIR devices for specific nutrients and forage components, their performances appear very promising. Given the numerous benefits that this technology and its application can offer, further studies are addressed in the area of sample presentation for calibration improvements. The development of accurate and trustable calibration curves, as well as of NIRS instruments usable for technicians and nutritionists, are essential for the precision feeding implementation that leads, among the others, to an improved sustainability of the dairy enterprises.