How many genes does it take to make a new species?

Three-spined sticklebacks are speciation machines. When retreating glaciers exposed lakes and rivers around the coasts of northern North America and Eurasia, these armor-plated little fish colonized the new freshwater habitats from the ocean, and adapted to the threats and resources they found there. But colonists kept coming from the ocean, and sometimes they found not an empty lake or a population of sticklebacks like themselves, but unfamiliar fishes that ate some of the same things they did.

Competition between the new arrivals and their evolutionary cousins gave an advantage to fish that relied less on the resources they both used. And, eventually, in many of those freshwater lakes, there were two types of stickleback: one that made a living in the shallow limnetic zone of ponds, eating free-swimming plankton; and one in the deeper benthic zone, snapping up prey from the bottom sediment or off of rocks and vegetation. Across many different bodies of water where they’ve been found, benthic and limnetic sticklebacks mate mostly within their types, meeting the classical definition of separate species.

The specific genetic differences between freshwater sticklebacks and their oceanic ancestors have been dissected in detail—they mostly boil down to change at a single gene that reduced the bony armor plating in freshwater populations. Now, a study published recently in Nature has dissected the genetic differences between the benthic and limnetic forms, and it shows that the genetic basis of sticklebacks’ repeated ecological speciation has been quite a bit more complicated than their move from the ocean.

Matthew Arnegard and a team from the University of British Columbia, the University of California, Davis, Uppsala University, and Stanford University, started by collecting benthic and limnetic fish from a British Columbia lake that has been a focus of stickleback studies for decades. They artificially fertilized eggs of one form with sperm of the other (and vice-versa) to create a population of hybrid sticklebacks.

At the beginning of the stickleback breeding season, Arnegard and his coauthors introduced the population of hybrid fish—the first-generation offspring of benthic-limnetic matings—into an artificial pond on the campus of the University of British Columbia. The sticklebacks did what fish do: eat, avoid predators, and make lots of new fish. In the fall, the coauthors collected a large sample of the young second-generation hybrid population and started taking measurements.

First, they established that the second-generation hybrid fish were starting to sort into the benthic and limnetic niches occupied by their grandparents. Stable isotope readings of carbon and nitrogen from the sticklebacks’ tissue showed that the fish were arrayed between two isotope profiles, indicative of two different diets, and fish at the two ends of this spectrum were the largest, meaning they had grown the fastest.

In the figure above, those are fish represented by points in the clusters labeled “L” and “B”. A third cluster, labeled “A,” is a group of fish that deviated significantly from the previously known limnetic and benthic diets—they grew more slowly than any other fish in the sample. Inventorying the gut contents of the fish in each cluster confirmed they had been eating different mixtures of aquatic invertebrates, corresponding to their availability in the benthic or limnetic zones. Measuring the fish taken from each group, the authors found that the intermediate fish had mouth cavities and lower jaws like their limnetic grandparents, but upper jaws more like benthics.

Finally, the team conducted a quantitative trait locus analysis to identify regions of the stickleback genome where fish with similar measurements shared genetic markers. They found 76 different genome regions, QTLs, that showed strong associations to traits that separate benthic sticklebacks from limnetic sticklebacks, spread widely across the genome. Fish with more benthic-like alleles at these QTLs also had more benthic-like isotopic dietary profiles—a nice demonstration of the connection between the genetic data, the morphological differences, and the actual ecological differences of the two different stickleback forms.

To test for non-additive interactions between QTLs, Arnegard et al. compared regression models predicting the fishes’ measurements with the QTL alleles, and a model that allowed for both dominance effects and non-additive interactions proved to fit the data best. That model identified four pairwise interactions among QTLs, but including them explained about 5% more variation than a model without interactions. It’s fair to say, as the authors do, that most of the predictive power was from simple additive effects.

Because they used a relatively small set of genetic markers—just 408 across the entire stickleback genome—Arnegard and his coauthors couldn’t define the QTLs they found with much precision. Some of them cover more than half of a chromosome! So it’s not impossible that there could be smaller loci with non-additive interaction effects lurking within some of those QTLs. Still, I think the finding that regions across more than half of the different stickleback chromosomes contribute additively to the differences between benthic and limnetic fish is pretty solid evidence that the multiple parallel stickleback speciation events have been created by differences at many loci of small effect.

That’s interesting both because it contrasts with the single-locus story of differentiation between oceanic and freshwater sticklebacks, and because it hints at a broader range of genetic models underlying speciation. As the authors note, speciation involving many loci of small effect may reflect adaptation from standing variation—in this case, genetic variation that may originate in the freshwater sticklebacks’ oceanic ancestors.

References

Arnegard ME, B Matthews, KB Marchinko, GL Conte, S Kabir, N Bedford, S Bergek, YF Chan, FC Jones & DM Kingsley. 2014. Genetics of ecological divergence during speciation, Nature, 511 (7509) 307-311. DOI: 10.1038/nature13301

Barrett RDH, SM Rogers, & D Schluter. 2008. Natural selection on a major armor gene in threespine stickleback, Science, 322 (5899) 255-257. DOI: 10.1126/science.1159978
McPhail JD. 1992. Ecology and evolution of sympatric sticklebacks (Gasterosteus): evidence for a species-pair in Paxton Lake, Texada Island, British Columbia , Canadian Journal of Zoology, 70 (2) 361-369. DOI: 10.1139/z92-054

Schluter D and JD McPhail. 1992. Ecological character displacement and speciation in sticklebacks, The American Naturalist, 140 (1) 85-108. DOI: 10.1086/285404