Two butterfly species have been bred in the lab to make a third distinct species, the journal Nature reports.
H. heurippa is found in the eastern Andes of Colombia (Image: Christian Salcedo)
In a species, individuals need to be capable of interbreeding to produce fertile offspring.
The study demonstrates that two animal species can evolve to form one, instead of the more common scenario where one species diverges to form two.
The process has been likened to building a new bike from a pair of second-hand ones.
The Heliconius heurippa butterfly appears to be the product of a process called hybrid speciation.
Most species are thought to form when groups of organisms gradually diverge from one another over successive generations.
But these distinctive red and yellow butterflies seem to be the product of two existing varieties.
Hybrid speciation is thought to be rare or absent in animals where, it has been argued, hybrid offspring would be less likely to survive and breed than the parent species.
This is because genes from different species are sometimes "incompatible".
A well known example is the mule - a sterile hybrid between the donkey and the horse. It is useful for carrying heavy loads but is a reproductive dead-end.
Heliconius melpomene (top) and Heliconius cydno (bottom) may have hybridised (Image: Juan Guillermo Montanes/Mauricio Linares)
A team of researchers from Panama, Colombia and the UK managed to recreate Heliconius heurippa in the laboratory by crossing two other species of butterfly; Heliconius cydno and Heliconius melpomene.
"The fact we've recreated this species in the lab provides a pretty convincing route by which the natural species came about," co-author Chris Jiggins, of the University of Edinburgh, told BBC News.
Jesus Mavarez, another author from the Smithsonian Tropical Research Institute in Panama, explained: "We found that a wing pattern almost identical to that of the hybrid can be obtained in months - just three generations of lab crosses between H. cydno and H. melpomene.
"Moreover, natural hybrids from San Cristobal, Venezuela, show wing patterns very similar to H. heurippa, further supporting the idea of a hybrid origin for this species."
In addition, there is growing circumstantial evidence for hybrid speciation in Ragoletis fruit flies, swordtail fish and African cichlid fish.
Some also suspect the American red wolf could be the product of hybridisation between coyotes and wolves.
Colour patterns on the wings of the butterflies may be crucial in forming new species, because they serve as mating cues. These butterflies are extremely choosey about finding mates with their own, species-specific wing pattern.
The wing patterns of H. heurippa individuals make them undesirable as mates for members of their parent species, but attractive to each other - reinforcing patterns of mating that lead to a new species.
These species-specific patterns are also crucial in deterring predators. The butterflies produce toxins when eaten and predators learn to recognise and avoid a specific wing pattern.
This is so finely tuned that butterflies with even slight deviations in colour pattern suffer from higher predation.