The Commentary article by Josh Donlan and colleagues (“Re-wilding North America” Nature 436, 913–914; 2005) argues for the introduction of old-world mammals to North America, on the grounds that these are proxies for megafauna that lived there at the end of the Pleistocene. This perspective overlooks environmental changes that have occurred during the intervening millennia, and that have produced qualitatively different communities in new ecological equilibria.

For example, global climate change since the Pleistocene extinctions makes the restoration of vanished ecosystems through large-mammal introduction quite unlikely. Such environmental change also increases the risk that introduced species might respond in unexpected ways. Indeed, one of the reintroductions proposed by Donlan and colleagues — the camel — was previously attempted during the nineteenth century (T. L. Connelly, Southwestern Historical Quarterly 69, 442–462; 1966). At that time, camels seemed an ideal beast of burden for use in the North American arid regions, but American environments ultimately proved inhospitable. By 1900, both public and private introduction programmes had failed, and the remaining feral camels had died.

The inherent unpredictability associated with disturbing ecosystems means that, while some introductions might follow the camel's fate, others might prove all too capable of adapting to contemporary North American environments, potentially at the expense of other species of conservation value. Indeed, introduced species are now a major cause of biodiversity losses worldwide.