Cutting edge: Ben Shelton

Exposing the infidelity of fly catchers has allowed scientists to catch natural selection in the species juggling act.

In the face of the cold accounting of natural selection, few things would seem less adaptive than breeding with an individual of another species. The hybrids it produces usually perform poorly. Often, one sex is sterile or the hybrid has a reduced chance of survival. This is not surprising. Combining genes that work well in one species with those of another should produce an organism that falls between two stools. Hybrid unfitness is an important stage on the road to speciation - when one species diverges into two.

During 20 years studying thousands of pairs of flycatchers on the Swedish island of Gotland, we occasionally recorded mixed pairs of two flycatcher species (collared and pied). These were not common - usually only a few per cent of pairings in any given year. Not so long ago, if I had been asked why such mixed pairings occurred, I would have answered rather vaguely and without hard evidence in terms of species-recognition errors, constraints on mate choice, or individuals having learnt the wrong-mate preference.

Two years ago I stumbled across an odd pattern in our long-term databases. When ringed nestlings produced by mixed pairs (usually a male pied flycatcher with a female collared flycatcher) were recaptured a year or so later as adults, it seemed we had made mistakes in identification. Rather than identifying them as hybrids, almost 60 per cent were recorded as pure collared flycatchers. We knew from DNA tests that our field identifications were usually reliable, so an alternative explanation was needed. Something was clearly happening in mixed pairs.

Soon after DNA fingerprinting was invented in 1985, it was applied to several species of wild bird. It was a revelation: frequently, male birds were found to be rearing nestlings unrelated to them. Observation showed females in these species cuckolded their mates by making clandestine visits to copulate with neighbouring males. Our long-term flycatcher data suggested precisely the same thing might be happening in the mixed-species pairs, but at an unusually high rate. We needed genetic typing to be sure. A year collecting blood samples from as many mixed families as possible, followed by months in the lab, produced startling confirmation of our suspicions. On average, 57 per cent of the supposedly hybrid offspring were actually pure-bred. Instead of producing genetically unfit hybrid offspring, the females were, in fact, copulating with a male of their own species thus producing pure offspring. This suggested that females in mixed pairs were doing much better, in terms of reproductive success than we would have guessed.

Additionally, later in the breeding season when success is generally low, females in mixed pairs did better than those in pure pairs. This compensated for the few unfit hybrid offspring that they produced. Combining these findings allowed us to predict the date in the breeding season when a female flycatcher would gain more from a mixed-species pairing than a pure pairing. To our amazement, this prediction matched observations.

The results stood my ideas on their head. The combination of a long-term study and a collaborative effort to collect and process DNA samples from families of flycatchers showed that what appeared maladaptive may actually be just the right thing to do under certain circumstances. Whether these processes accelerate the rate at which two species diverge remains to be seen, but they do offer a solution to the puzzle of why female birds mate with other males within species.

Ben Sheldon is a Royal Society university research fellow in the department of zoology, Oxford University.