Aim of the study was to investigate the influence of long-term pure-br
ed selection with respect to combining ability of different mouse line
s for endurance fitness tested on a treadmill in two environments. Com
plete diallel was carried out in semi-barrier-system and under convent
ional conditions. Used lines were selected for fertility (line F: 73 g
enerations), growth (line G: 52 generations) and treadmill performance
(line E: 28 generations); a further used line was a control C. Differ
ent models were available for estimation crossbreeding effects. GLS-me
ans: The largest positive effects were found for treadmill performance
at 42. day of age in all crossbreeds with line E, independent if E wa
s the dam or sire line. In semi-barrier-system heterotic effects were
frequent: the best cross was ExC. In conventional environment cross Ex
E was superior. Good results for treadmill performance were also obser
ved if line C was used as parental line. Line F and especially line G
had a worse effect on this performance of their Fl crosses. In all env
ironments the worst crosses of the diallel were GxF, FxG and CxG. Fixe
d effects: It is evident that in all models and both environments the
following ranking for general combining ability is found: E>C>F>G. Fix
ed effects of all models showed that both crosses of CxE and of CxF ha
ve the highest specific combining ability/specific direct heterosis in
treadmill performance.