Search

Die sichere Erkennung von Eutererkrankungen muss sowohl beim konventionellen als auch beim automatischen Melken gewährleistet sein. Störungen in der Eutergesundheit spiegeln sich zuerst in geänderten Milchinhaltsstoffen, die für das menschliche Auge nicht zu erkennen sind (subklinische Mastitis), und später auch in einer visuell erkennbaren Veränderung der Milchzusammensetzung (klinische Mastitis) wider. Die Aussagekraft der Parameter Elektrische Leitfähigkeit, Spektrale Reflexion sowie Na+ und Cl- Gehalt wurde einzeln und in Kombination hinsichtlich der Erkennung von Eutererkrankungen mittels Diagnostischer Tests bewertet, wobei der Schwerpunkt der Untersuchungen auf der Analyse der Spektralphotometrie(Farbmessung) als zusätzliches Werkzeug einer Mastitisfrüherkennung lag. Zusätzlich wurde überprüft, welche Gemelksfraktion die besten Messwerte bezüglich einer veränderten Eutergesundheit liefern konnte. Die Ergebnisse zeigten, dass durch die Erfassung von elektrischer Leitfähigkeit in Kombination mit der spektralen Reflexion in den ersten 300-500 ml eines jeden Viertelgemelkes eine weitaus bessere Erkennung von Eutergesundheitsstörungen als mit der bloßen visuellen Prüfung des Vorgemelkes erzielt werden konnte. Die Erkennungssicherheit sollte aber mit Blick auf eine automatische Trennung von verkehrsfähiger und nicht verkehrsfähiger Milch noch weiter verbessert werden. In the present exposition different chemical-physical parameters were evaluated for their sensor-controlled use to detect udder diseases in dairy cows having a special regard on spectral photometry. The appraisal of all results was done based upon the laboratory determined quarter specific somatic cell count (SCC) and the corresponding classification. First of all the expressiveness of the simple methods to describe the composition of foremilk (visual inspection by law and CMT) was evaluated under excellent artificial light conditions. For that altogether 1,510 milkings were sampled and evaluated in two herds over a period of 12 days each. These foremilkings were analysed considering the milk consistence, milk colour and the CMT-value. It was found that the sole use of visual inspection (milk consistence and -colour), stipulated by law, allows no accurate statement about the udder health. In merely 15.1 % (29.7 %) of all foremilkings with more than 500,000 cells/ml a change in the milk-colour (-consistence) could be detected. Only the additional use of the CMT made a detection rate of 88.8 % of all foremilkings with more than 500,000 cells/ml possible. By changing the threshold up to 1 Mio. cells/ml the sensitivity went up to 94.7 %. In order to investigate the possibility to detect udder diseases by using the spectral photometry in milk, 1,033 quarter first milk samples (13 cows, 280 milkings) were evaluated. This examination resulted an increased correlation between milk ingredients like fat and protein and spectral reflectance (SR) in wave lengths with more than 530 nm. The highest correlation between SR and SCC and also to the lactose content was found in the band of 400 - 520 nm. Moreover there was a high significant difference between the average value of SR in the band of 430 - 510 nm of quarter foremilk with raised SCC (>100,000 cells/ml) and samples with less than 100,000 cells/ml. Therefore that range was defined as mastitis-band and was solely used for all further analyses. Due to the possibility of using an AMS (automatic milking system) to record relevant parameters during the milking, the optimal point in time for sampling is important to get the most expressive values of the recorded parameters. Consequently the quarter specific milk fractions from 64 milkings were collected. They were analysed on electrical conductivity (EC), spectral reflectance in the mastitis band(SR) and the Na+ and Cl- content (NA, CL). In addition the significance of the SRmastitis of the quarter first milk and the quarter composite milk was compared. The results revealed that the best values for all tested chemical-physical parameters regarding the udder health status were found in the quarter first milk. There was almost no information about the udder health in the quarter composite milk, even though no significant difference between SCC in the quarter foremilk and the quarter composite milk could be detected. The main part of this study was to verify an improved detection of udder diseases by combining the chemical-physical parameters (EC, SR, NA, CL). For this research the quarter first milk samples of 280 milkings were collected from 16 cows, which were automatically milked. The informational value of the parameters was calculated in diagnostic tests (sensitivity and specificity). The sensitivities and the associated cut-off points for the individual and combined parameters were determined at a specificity of at least 95 %. By measuring the values of SRmastitis resp. EC and the calculated values to the reference quarters (inter quarter comparison), 55.6 % resp. 61.4 % of all quarter first milk samples with more than 500,000 cells/ml were detected (specificity=95 %). With the combination of both parameters it was possible to increase the sensitivity up to 73.3 % at the same specificity and threshold. This parameter combination enabled an identification of 85.2 % of all quarter milkings with both, more than 500,000 cells/ml and a positive bacteriological finding. In this case a marginal increased detection rate up to 89.0 % could be reached using an additional combination with direct ion measurement (NA, CL). Altogether it was not possible to detect udder diseases, i.e. subclinical mastitis, prematurely by solely checking the visual appearance of foremilk without any sensor system. In case of using a sensor system, based on chemical-physical parameters, the most meaningful values concerning the udder health status can be gained in the quarter first milk ideally before milk ejection starts. The measurement of milk colour based on spectral reflectance gives some useful information and improves the detection rate of udder diseases in combination with the measurement of EC.