Minimizing the environmental damage that new roads cause is generally regarded as a good thing. But to do that, it helps to understand just how new roads cause the damage of which they are accused.

Recently, a group of researchers led by Dr. Gonzalez conducted an experiment which shows what ecologists have long suspected, but never been able to prove: that immigration is good for the health of animal populations.

A road destroys only a small part of the habitat it traverses, and thus annihilates just a few local populations of creatures. So the argument that road-building itself is bad for biodiversity is not self-evidently correct. Those who nevertheless hold this view usually point to a piece of ecological theory called “meta-population dynamics”. This says that apparently separate local populations of animals are, in fact, parts of much larger populations connected via migration.

According to this theory, when a local population flounders—because of an epidemic, for example—individuals from neighboring communities can fill the gaps. So the more such communities there are, the better the chance of a given local population remaining healthy.

The implications of the theory for conservation are straightforward. Cut local populations off from each other and each is more likely to disappear. And roads are good at doing just that. Testing the theory with experimental roads, however, would be expensive. Dr. Gonzalez’s brainwave was to do the whole thing on a much smaller scale.

Instead of studying, say, a forest, the team looked at moss-covered rocks. These support diverse population of tiny arthropods (insects, mites and so on). On some rocks the researchers left the moss untouched; on others they scraped “roadways” across to leave “isolated” parts. After waiting six months, they found that in the disturbed habitats nearly all the bug populations had declined compared with the undisturbed moss, and 40% of the species had become extinct.

The real test of the recta-population hypothesis came in the second part of the experiment. In this, the researchers scraped away moss much as before, but they left narrow moss paths to bridge the no-bug’s-land between islands. These connected patches were still not as healthy after six months as the unsullied moss, but they did far better than isolated islands—a result that supports the notion that population exchange is necessary to keep an ecosystem healthy.

Whether these results can be translated to large-scale ecosystems remains uncertain. But if they can, they would cause more, not less, concern about the ecological effects of road-building. On the other hand, they also suggest a way out.

In Britain, tunnels are often built under roads for animals of regular habits, such as badgers, to be able to travel their traditional routes without having to tangle with the traffic. Extending that principle, perhaps with special bridges that can support local vegetation and thus allow animals the illusion of an uninterrupted habits, might be a cheap way of letting man and nature rub along a bit better.