The study of speciation—the formation of new species—has had a long history in evolutionary biology, but the past few decades have seen leaps in how we think about the process that creates biodiversity. We now know that natural selection is almost always heavily involved in the process, and that new species can form even when there is some ongoing hybridization between the evolving lineages.
One way to approach a recently evolved pair of species is to ask (1) why did they split in the first place and (2) how do they keep separate.
Question 1 often involves determining whether the new species are adapted to different environments, while question 2 looks at whether the organisms in question mate preferentially with their own species, whether their hybrids are inviable or sterile, etc. The two questions are more closely linked than might appear at first. For example, hybrids might not be able to compete with either parent by virtue of being halfway between two species that are adapted to use very specific resources. Another example is when adaptations to different environments also prevent the two species from recognizing each other as mates. In fact, when the answers to both questions are linked, as in the two examples I just mentioned, speciation becomes easier.
Most speciation studies look at existing species, or highly diverged varieties of a single species (which are assumed to be on their way to speciation). This study, however, used a single well-studied species, a beetle called Callosobruchus maculatus that eats mung beans and related plants, and tried to start speciation experimentally.
The beetles had been raised in the laboratory on black-eyed beans for many generations. Then the researchers created 18 new strains: nine used black beetles that were still raised on black-eyed beans (BE treatment), the other nine consisted of brown beetles raised on mung beans (M treatment). Each generation, 20 beetles were allowed to mate and lay their eggs on their treatment bean. After ten generations, the researchers asked whether (1) females preferred to lay eggs on the type of beans on which they had been raised and (2) females chose mates that had been raised on the same type of beans as them.
To test their first question, they simply gave females a half M/half BE mix and counted the eggs on each. And indeed, M females laid more eggs on M beans while BE females laid more eggs on BE beans.
To test their second question, they placed 20 males and 20 females from an M strain together with 20 males and 20 females from a BE strain in an environment with half M and half BE beans. Remember that the M strain beetles were all brown and the BE beetles were all black. Colour in these beetles is controlled by a single gene, and a cross between a brown and a black beetle gives “intermediate”-coloured offspring. Additionally, in the lab population that founded these experimental strains, beetles don’t choose their mates based on colour – in fact, they mate in the dark! Now, to determine whether mating is assortative—whether M females mate mostly with M males and BE females with BE males—the authors simply needed to compare the proportion of intermediate offspring to the proportion they’d expect if the beetles mated at random.
The authors didn’t simply check for assortative mating once—in fact, they continued the test for eight generations. This is important, because although assortative mating might have evolved, it doesn’t necessarily have to be permanent, especially if there is no penalty to mating with the “wrong” strain—if the hybrids, for example, are perfectly healthy.
In the first generation of the assortative mating test, there were fewer intermediate offspring than expected in 13 of the 18 lines—meaning M females tended to mate disproportionately with M males and BE females with BE males. However, with each subsequent generation, assortative mating decreased, and after four generations of contact with the other lines it had disappeared altogether. So although the beetles could evolve a preference for a new type of bean and a preference for mating with their own strain, it wasn’t permanent—meaning speciation got started, but was curtailed.
This result drives home a few messages about speciation that other studies have suggested. First, speciation is hard. Even if only a few beetles mate with the “wrong” type, eventually the hybrids will swamp the “purebred” types and the population will collapse back to random mating. Second, a single barrier to reproduction is often not enough to cause speciation. By this I mean that assortative mating alone didn’t stem the flow of hybrids, but if, say, hybrids couldn’t eat mung beans as well as M beetles and black-eyed beans as well as BE beetles and as a result often didn’t survive to adulthood, there would be fewer hybrids to contribute to the swamping process. Then there might even be selection for the parents to avoid mating with the other type, but that’s a matter for further experiments! Third, when we see assortative mating in nature, we shouldn’t assume that it’s permanent. Other forces may be required to maintain it, and these forces may depend upon changeable environmental factors.
Finally, this result fits with findings from other studies that suggest that premating isolation—barriers to finding and mating with the other neo-species—evolves before postmating isolation—factors that affect the performance of hybrids—during speciation. However, this could simply be a byproduct of the particular species they used, e.g. it might be “easier” to evolve mate preference than to lose the ability to eat one kind of bean in these beetles. But then again, that might tell us something about why speciation couldn’t happen—perhaps it’s simply easier in organisms and environments where postmating isolation evolves readily.
There is good evidence in this paper that the differences in mate and bean choice at the end of the selection phase were genetic in nature (the authors tested whether a single generation on a new type of bean created a preference for that bean, and it didn’t). It would be interesting to investigate the genes underlying mate and egg-laying site choice, because this might shed more light on why assortative mating eventually collapsed. In particular, if genes for mate choice and bean choice are unlinked (if they’re on different chromosomes, or for away from each other on the same chromosome), it’s quite difficult to maintain assortative mating without additional natural or artificial selection. If there were related species in which these genes were actually linked, speciation in the lab might be possible…but I begin to fantasize; this would require a ton of work!
This paper clearly shows that mate choice based on habitat can evolve, and that it can persist for a few generations after contact with the other strain. But it also illustrates the difficulty of speciation and suggests that changes in geography and environmental factors might lead to transient fluctuations in assortative mating levels, only sometimes leading to speciation.
Rova, E., & Björklund, M. (2011). Can Preference for Oviposition Sites Initiate Reproductive Isolation in Callosobruchus maculatus? PLoS ONE, 6 (1) DOI: 10.1371/journal.pone.0014628