Various direct selection traits were tested for improved udder health
through breeding on the basis of evidence from a commercial dairy with
a high concentration of animals. The traits differ according to the m
anner of the findings (clinical and bacteriological), in the duration
of the diseases under consideration, and in the duration of the testin
g period, which includes one or more or partial lactations. Clinical s
ecretion findings displayed the higher estimated heritability values o
f h2 = 0,04 to 0, 1 7, compared with bacteriological findings of h2 =
0,01 to 0,06. A trait which includes all diseases in the testing perio
d is hardly advantageous to the breeder when compared with an alternat
ive trait which allows only the difference between healthy and disease
d, An extension of the testing period effects an increase in the herit
ability of the clinical findings. The estimated values of the bacterio
logical findings, however, remain uninfluenced by the duration of the
testing period. Of all variants tested, clinical findings prove to be
the alternative trait in the first lactation which, as most favourable
selection trait, is recommended for future testing.