Behaviour of house mice artificially selected for high levels of voluntarywheel running

We have developed a novel model to study the correlated evolution of behavi oural and morphophysiological traits in response to selection for increased locomotor activity. We used selective breeding to increase levels of volun tary wheel running in four replicate lines of laboratory house mice, Mus do mesticus, with four random-bred lines maintained as controls. The experimen t presented here tested for correlated behavioural responses in the wheel-c age complex, with wheels either free to rotate or locked (environmental fac tor). After 13 generations, mice from selected lines ran 2.2 times as many revolutions/day as controls on days 5 and 6 of initial exposure to wheels ( 10 826 versus 4890 revolutions/day, corresponding to 12.1 and 5.5 km/day, r espectively). This increase was caused primarily by mice from selected line s running faster, not more minutes per day. Focal-animal observations confi rmed that the increase in revolutions/day involved more actual running (or climbing in locked wheels), not an increase in coasting (or hanging). Not s urprisingly, access to free versus locked wheels had several effects on beh aviour, including total time spent in wheels, sniffing and biting. However, few behaviours showed statistically significant differences between the se lected and control lines. Selection did not increase the total time spent i n wheels (either free or locked), the frequency of non locomotor activities performed in the wheels, nor the amount of locomotor activity in cages att ached to the wheels; as well, selection did not decrease the amount of time spent sleeping. Thus, wheel running is, at the genetic level, a largely in dependent axis of behaviour. Moreover, the genetic architecture of overall wheel running and its components seem conducive to increasing total distanc e moved without unduly increasing energy or time-related costs. The selecti on experiment also offers a new approach to study the proximate mechanisms of wheel-running behaviour itself. For example, frequencies of sniffing and wire biting were reduced in selected females but not males. This result su ggests that motivation or function of wheel running may differ between the sexes. (C) 1999 The Association for the Study of Animal Behaviour.