Ecologists have studied the wolves and moose on Isle Royale, a remote island in Lake Superior, for more than 50 years. As we report on page 140, after decades of isolation and inbreeding, the wolf population may be on the verge of dying out.

The US National Park Service, which manages the island, is moving slowly in deciding how to proceed. It has three options: total non-­intervention; reintroduction of wolves only after the current population has hit zero; or pre-emptive genetic rescue by bringing in wolves from the mainland to diversify the gene pool. Arguments for non-­intervention tend to rely on the perceived need to let nature take its course. This is nonsense. The whole system is highly artificial: wolves and moose have been on the island for less than 100 years, and human activity has been key to the wolves’ decline. A previous wolf-population crash in the 1980s was caused by a disease transmitted by a domestic dog. Anthropogenic climate change is almost certainly reducing how often ice bridges form to the mainland, which makes it hard for new wolves to come to the island. Some even think that humans put moose on Isle Royale in the first place.

Arguments are more convincing for reintroducing wolves only if the current population dies out: waiting and watching may yield some useful insights into how highly inbred populations function. But the ecologists who run the island’s predator–prey observation study warn that, as the wolves die out, the moose will gorge unchecked on their key food plant, balsam fir, preventing the plant from regenerating. The researchers think that by the time the old wolf population has died out and a new one is established, the ecosystem may have become dominated by pine or spruce, without enough firs to support a moose population that can in turn feed a viable wolf population. If the wolves die out, they could become nearly impossible to reintroduce.

And that might be fine, except that tourists and locals love the wolves of Isle Royale, and the National Park Service was founded with an obligation to protect “the enjoyment of future generations”. Furthermore, the predator–prey study — the world’s longest — would have to end. That would be a shame: it would be difficult to find another place where none of the predators, herbivores or trees are routinely exploited by humans.

The study’s lead ecologists are in favour of genetic rescue. This fairly cheap intervention would allow the project to continue, and would stabilize an ecosystem with which many people feel a deep connection.Some researchers have suggested that any data on re­introduced wolves would have to be treated with caution. Certainly, the influence of the reintroduction would be acknowledged and studied. But the introduced population would not be any more artificial than the population that survived disease, or that which could suffer the effects of climate change.

Isle Royale data help ecology to approach one of its grandest questions. As study leader John Vucetich puts it: “Are eco­systems like other physical systems, governed by law-like patterns and processes, or are they more like human history, where we see one contingency after the next?” The early years of the study seemed to support predictions that in a closed system, predator and prey populations would follow law-like mirror-image cycles, driven by predation pressure. But the data never fitted the theoretical curves that well. And since then, factors from disease to fir abundance, weather, moose ticks and wolf inbreeding have taken turns as the key driver in shaping the populations.

The driver that will shape the future of Isle Royale is now the decision on whether to stage a rescue. Thus of the story of all Earth’s systems is writ small on a wooded isle in a frozen lake: the course of human history is no longer merely analogous to the course of ecology. Ecology depends on human history.