Spectroscopy for assessing the nutritional value of pastures and enteric methane emissions from dairy cows in northern Sweden

We aimed to estimate the nutritional value of leys using short-wave infrared (SWIR) spectroscopy. Samples were collected from 12 different field spots at seven dairy farms in Northern Sweden during summer 2021 and 2022. Sub-samples were chemically analysed for nutrient content, gas production and digestibility, and the remaining material was scanned with the Specim SWIR camera. Multivariate regression and machine learning algorithms were tested to estimate organic matter, crude protein, aNDF, ADF, non-fibre carbohydrates, energy, in vitro true organic matter digestibility, and enteric total gas and methane emissions predicted in vivo, where the SWIR spectral reflectance values were used as explanatory variables. The nutrient chemical composition of pastures varied among farms and months, and higher nutritional value was observed in June and August, irrespectively of year. Preliminary results show satisfactory performances for most quality parameters (R2 ranging from 0.83 to 0.99, normalized RMSE ranging from 0.4 to 5.1) and indicate the potential of spectroscopy to assess pasture quality, making it a viable alternative for nutritional monitoring. We plan to increase the size of the dataset to improve the robustness and generalization of the models.