Acta Univ. Agric. Silvic. Mendelianae Brun. 2016, 64, 1155-1166

https://doi.org/10.11118/actaun201664041155
Published online 2016-08-30

The Evaluation of Real Time Milk Analyse Result Reliability in the Czech Republic

Oto Hanuš1, Luděk Stádník2, Marcela Klimešová1, Martin Tomáška3, Lucie Hasoňová4, Daniel Falta5, Josef Kučera5, Jaroslav Kopecký1, Radoslava Jedelská1

1Dairy Research Institute, Ltd., Prague, Czech Republic
2Czech University of Life Sciences, Faculty of Agrobiology, Food and Natural Resources, Prague, Czech Republic
3Dairy Research Institute, Žilina, Slovakia
4University of South Bohemia, Faculty of Agriculture, České Budějovice, Czech Republic
5Mendel University in Brno, Faculty of AgriSciences, Department of Animal Breeding, Czech Republic

The good result reliability of regular analyzes of milk composition could improve the health monitoring of dairy cows and herd management. The aim of this study was the analysis of measurement of abilities and properties of RT (Real Time) system (AfiLab = AfiMilk (NIR measurement unit (near infrared spectroscopy) and electrical conductivity (C) of milk by conductometry) + AfiFarm (calibration and interpretation software)) for the analysis of individual milk samples (IMSs). There were 2 × 30 IMSs in the experiment. The reference values (RVs) of milk components and properties (fat (F), proteins (P), lactose (L), C and the somatic cell count (SCC)) were determined by conventional (direct and indirect: conductometry (C); infrared spectroscopy 1) with the filter technology and 2) with the Fourier transformations (F, P, L); fluoro-opto-electronic cell counting (SCC) in the film on the rotation disc (1) and by flow cytometry (2)) methods. AfiLab method (alternative) showed less close relationships as compared to the RVs as relationships between reference methods. This was expected. However, these relationships (r) were mostly significant: F from .597 to .738 (P ≤ 0.01 and ≤ 0.001); P from .284 to .787 (P > 0.05 and P ≤ 0.001); C .773 (P ≤ 0.001). Correlations (r) were not significant (P > 0.05): L from −.013 to .194; SCC from −.148 to −.133. Variability of the RVs explained the following percentages of variability in AfiLab results: F to 54.4 %; P to 61.9 %; L only 3.8 %; C to 59.7 %. Explanatory power (reliability) of AfiLab results to the animal is increasing with the regularity of their measurements (principle of real time application). Correlation values r (x minus 1.64 × sd for confidence interval (one-sided) at a level of 95 %) can be used for an alternative method in assessing the calibration quality. These limits are F 0.564, P 0.784 and C 0.715 and can be essential with the further implementation of this advanced technology of dairy herd management.

Funding

This paper was supported by projects NAZV KUS QJ1210301, „S“ grant MŠMT ČR, APVV-0357-12, NAZV QH 81210, IGA FA MENDELU TP 5/2014 and RO 1415.

References

52 live references