Self-regulation within outbreak populations of feral house mice: a test of alternative models.

1. Outbreaks of feral house mice, Mus domesticus, in Australia represent a fundamental failure of the behavioural control mechanisms of population density, as proposed in the hypothesis of self-regulation. 2. Mice have the potential to keep numbers in check via a suite of spacing behaviours; however, the self-regulation hypothesis implies that some social change occurs that permits the population to erupt. It also suggests that at different phases of an outbreak, distinct patterns of social activity are evident. 3. We compare predictions from two models encapsulating the self-regulation hypothesis as applied to feral house mice in south-eastern Australia. Each model may be distinguished by the timing of aggressiveness between mice that leads to a closed social system. We compare individual turnover, residency and territoriality in each sex and age cohort during the increase, peak and low phases of a population outbreak that peaked in 2001. 4. The activity of 438 mice was monitored via intensive mark-recapture trapping and an automated event recording system that detected the activity of 300 marked individuals at burrow entrances. 5. Our findings support the second model, which suggests that mice switch from an almost asocial structure at low densities to a territorial system as abundance increases. Adult females appear more likely than males or juveniles to make the significant social shift. The trigger for this change remains unclear and several alternative mechanisms are proposed.