Abstract :
Based on a repeated sampling the influence of an intramammary infection on the
somatic cell count (SCC), the content of lactoferrin (Lf) and the activity of N-acetyl- -
d-glucosaminidase (NAGase) in goats’ milk was investigated. 58 dairy goats (German
Improved Fawn) were sampled weekly over 3 consecutive weeks and the udder halves
were classified according to the results of the bacteriological analysis of the foremilk
samples and the results of their parallel half into three groups: uninfected halves with
an uninfected parallel half (NoInf/NoInf), uninfected halves with an infected parallel half
(NoInf/Inf) and infected halves with an uninfected parallel half (Inf/NoInf). None of the
goats had two infected halves, thus, this group was omitted. 15 out of the 58 goats
were infected on one udder half. The bacteria detected were Staphylococcus aureus (n = 4),
coagulase-negative staphylococci (n = 7), corynebacteria (n = 3) and esculin-positive streptococci
(n = 1). Log10 SCC, log10 Lf and log10 NAGase were strongly correlated to each other
and changed over the sampling period but not uniformly, revealing a significant effect
of the sampling day on the variables (F6,98 = 29.13, p < .001). This could not be explained
by an underlying effect due to the stage of lactation or the estrus season the animals
were in, and therefore needs further investigation. The infection status had a significant
effect on log10 SCC (F2,103 = 20.22, p < .001), log10 Lf (F2,103 = 11.18, p < .001) and log10 NAGase
(F2,103 = 12.06, p < .001). Inf/NoInf differed significantly from NoInf/NoInf as well as NoInf/Inf
for log10 SCC (p < .01) and log10 Lf (p < .001) whereas the NoInf/NoInf did not differ from
NoInf/Inf indicating that the infected halves did not influence their uninfected parallel half.
For log10 NAGase this was different: infected halves differed significantly from NoInf/NoInf
(p < .01) but not from NoInf/Inf which might be caused by a dependency of the udder halves.
Results support the approach to monitor mastitis in goats by means of Lf or NAGase
instead of SCC. Further studies should explore the effect of other independent variables,
such as estrus, on these indicators and aim for thresholds indicating an intramammary
infection
