Context . Feral cats (Felis catus) threaten biodiversity in many parts of the world, including Australia. Low-level culling is often used to reduce their impact, but in open cat populations the effectiveness of culling is uncertain. This is partly because options for assessing this management action have been restricted to estimating cat activity rather than abundance.

Aims . We measured the response, including relative abundance, of feral cats to a 13-month pulse of low-level culling in two open sites in southern Tasmania.

Methods . To do this we used remote cameras and our analysis included identification of individual feral cats. We compared estimates of relative abundance obtained via capture–mark–recapture and minimum numbers known to be alive, and estimates of activity obtained using probability of detection and general index methods, pre- and post-culling. We also compared trends in cat activity and abundance over the same time period at two further sites where culling was not conducted.

Key results . Contrary to expectation, the relative abundance and activity of feral cats increased in the cull-sites, even though the numbers of cats captured per unit effort during the culling period declined. Increases in minimum numbers of cats known to be alive ranged from 75% to 211% during the culling period, compared with pre- and post-cull estimates, and probably occurred due to influxes of new individuals after dominant resident cats were removed.

Conclusions . Our results showed that low-level ad hoc culling of feral cats can have unwanted and unexpected outcomes, and confirmed the importance of monitoring if such management actions are implemented.

Implications . If culling is used to reduce cat impacts in open populations, it should be as part of a multi-faceted approach and may need to be strategic, systematic and ongoing if it is to be effective.