Assessment of compensatory predation and re-colonisation using long-term duck nest predator removal data

Removal of predators has been often found ineffective to increase duck nest success. Most often failures were explained by compensatory predation by other predator species and/or rapid recolonisation of the target area by new individuals by the same predator species. We used 13-year data of removing marsh harriers Circus aeruginosus, corvids Corvidae and American minks Mustela vision to test whether (i) removal of an individual predator species increased duck nest success; (ii) removal of an individual predator species decreased subsequent duck nest depredation rates by the same predator species; (iii) removal of one predator species increased subsequent proportion of duck nests depredated by other predator species. We removed 1 590 predators and followed the fates of 3 019 duck nests. Predator removal was measured using a concept-free days, expressed as the number of days of active duck nests during exposure to the removed predator's search if the removal would not happen. Predators were removed from the main duck breeding area and from its surroundings which altogether formed the entire predator removal area. Harrier removal was positively correlated with the apparent duck nest success (P is less than 0.05). However, this was true considering removed from the entire predator removal area, but not when they were removed only from the duck breeding area, thus suggesting that arrival of new harriers from the surroundings was an important factor in determining nest success in the much smaller duck nesting area. Removal of corvids and American mink were not correlated with duck nest success nor the subsequent predation rates of the same species. Mink removal was positively correlated with the proportion of nests depredated by harriers (P is less than 0.05) suggesting that harriers were compensatory predators after mink removal. Re-colonisation and compensatory predation after removing certain predator species may occur in many predator communities thus causing waste of management efforts. We suggest ways of how to evaluate past and ongoing management programmes and to plan future programmes with the aim of providing early diagnostics of a predator problem.