Islands can be home to rare and unusual species, which have often evolved in isolation for extremely long periods of time. On many - particularly oceanic islands - there may be no native land mammals, except, perhaps for bats. So when mammalian predators do make it to these islands the effects can be devastating. (Incidentally, it's wrong to say, as many do, that island species evolved in the absence of predators. There were predators all right - look at Harpagornis, New Zealand's extinct giant eagle. Or frogs, & tuataras. Just no mammalian predators.) This is particularly the case where the island's ecosystem is reliant on seabirds - or, more precisely, on seabird poo, which is an ongoing source of nutrients for the soil. Knock out, or just greatly reduce, the seabirds & the whole ecosystem suffers.
Anyway, most people would agree that the 'fix' to this problem is to remove the key introduced predators, & all will be well. But - they'd be wrong, & it's not. This is nature at work, & things are far more complicated than they first appear. As Bergstrom et al. (2009) say, taking out those top predators is only half the story.
For example, they point out that when cats were removed from Little Barrier Island, breeding success of Cook's petrels decreased. Things improved only when rats were also extirpated. And there's another group of introduced mammals that we ignore at our peril - the herbivores. In their recent paper Bergstrom & colleagues present evidence that removal of cats from Macquarie Island, rather than leading to an increase in bird populations, was followed instead by a boom in bunny breeding, & this in turn had hugely damaging effects on the island's ecosystems.
Macquarie Island is a World Heritage site that lies in sub-Antarctic waters. Sealers introduced rabbits to the island in 1878 (presumably as a food source for shipwrecked sailors - I imagine they'd be more palatable than seals!). The rabbits bred like, well, rabbits & reached plague proportions, doing significant damage to the native grasses and also providing a major source of food for the island's cats, themselves introduced in 1818.
Bergstrom et al. explain that the rabbit population had reached 130,000 by 1978, when the myxoma virus was introduced as a control measure. This proved fairly successful (although the myxoma virus** never became permanently established on Macquarie, meaning it had to be released every year.), and rabbit numbers dropped to around 20,000: this was matched by a marked improvement in the state of the island's plants. Unfortunately, when the rabbit supply began to dry up the cats simply switched to eating seabirds, so a cat eradication program was begun in 1985, with the last cat killed in 2000.
You can probably see where this was heading. With the cats gone, the rabbit population exploded, rapidly returning to the highs of the 1970s. The outcome was obvious environmental changes: ecosystems dominated by a mixture of tall plant species were in places replaced by short 'lawns', and an invasive grass species (Poa annua) became more widespread. Based on comparisons of satellite images, the team estimated that 36% of the island's land area was affected to some degree by rabbit activites (feeding & burrowing), with 17.5% 'having moderate to severe change'. And that's not counting the impact of rats & mice, which would also have been preyed on by the cats & whose populations also expanded rapidly when the cats were gone. What's more, you might think that the seabirds would be better off with the removal of cats, but in fact the researchers documented damage to petrel burrows by tunnelling rabbits, exposing the petrels to skua predation.
What does this mean for management of pest populations? The team comments that unanticipated management outcomes not only can lead to new substantial problems, but can also render past effots less valuable than they otherwise might have been. This has certainly been the case on Macquarie, and it's made more likely where pest control (& other management techniques) are funded on a year-by-year, step-by-step basis. Integrated pest management would probably be more expensive - but the ecological pay-off would probably reward such an approach.
D.M. Bergstrom, A. Lucieer, K. Kiefer, J. Wasley, L. Belbin, T.K. Petersen, & S.L Chown (2009) Indirect effects of invasive species removal devastate World Heritage island. Journal of Applied Ecology 46: 73-81 doi:10.1111/j.1365-2664.2008.01601.x
(** Looking around for information on myxomatosis, I found a fascinating blog on the potential significance of the myxoma virus in, of all things, research into cancer treatments.)