Effectiveness of energy dispersive X-ray fluorescence (ED-XRF) for the prediction of mineral elements in individual cow urine

Mineral elements, along with proteins, free amino acids, and sugars, represent one of the main indicators of heat stress in dairy cows urine. For these reasons, there is interest in the development of novel techniques aimed to predict urine mineral elements in reasonable time and low costs. The objective of the present research was to assess the effectiveness of energy dispersive X-ray fluorescence (ED-XRF) for the prediction of Ca, K, Mg, Na, P, S and Cl in dairy cows urine. Individual urine samples were collected from 45 Holstein Friesian cows in 9 commercial dairy farms located in north of Italy. One aliquot was used for mineral extraction followed by ICP OES analysis. The second aliquot was used for ED-XRF analysis on untreated urine samples. The training of the ED-XRF method was performed on a subset of 30 individual urine samples, as a regression equation between mineral elements obtained through ICP-OES analysis and normalized impulses profiled by ED-XRF instrument. Instead, the testing of the ED-XRF method was performed excluding samples used in the training dataset and considering only the remaining samples (n = 15) as a regression equation between minerals measured through ICP-OES versus minerals predicted through ED-XRF using the training model. Results highlighted moderate accuracy of the ED-XRF technique for the quantification of Ca, with coefficient of determination in testing (R2) of 0.60 while greater accuracies were obtained for the quantification of the other elements, with R2 ranging from 0.86 (Mg and K) to 0.99 (Cl).