Origin of Lake Constance stickleback

Today, my newest paper was published in Nature Communications. In this collaboration with Kay Lucek, Vitor Sousa, Laurent Excoffier and Ole Seehausen, we have reconstructed the evolutionary history of threespine stickleback found in Lake Constance and surrounding rivers.

Threespine stickleback lake (left) and stream (right) ecotypes of Lake Constance, Central Europe. Females are shown on top and males in nuptial coloration on the bottom.

With extensive demographic modelling and the analysis of both genomic and phenotypic data from the European biogeographic context, we found out where stickleback lake and stream ecotypes come from and whether genetic differences between them arose from new mutations, standing genetic variation or the admixture of old lineages.

Spoiler alert! The title gives it away:


The evolution of lake and stream ecotypes in Lake Constance is thus a prime example for how combinatorial processes can facilitate rapid speciation.

Enjoy the read of our paper and of this blog post on the Nature Research Ecology & Evolution Community forum outlining the story behind the paper.

A combinatorial view on speciation

A review by Joana Meier, Ole Seehausen and me was just published in Trends in Ecology & Evolution. In this manuscript, we highlight a genetic mechanism that has often be found to facilitate rapid speciation and adaptive radiations in speciation genomic research from the recent years (see a few examples in the figure below). The mechanism we highlight is the reassembly of old genetic variants into new combinations or ‘combinatorial mechanism’.


Adopting such a combinatorial view makes some counter-intuitive predictions, for example that speciation rates might be inversely related to the rate of complete reproductive isolation between speciating taxa, and might throw new light on old issues such as the likelihood of speciation in sympatry. I hope you enjoy the read and feel free to let me know your thoughts of our ideas!

Convergent evolution: same challenge, same solution

Today, the first paper from my postdoctoral research with Tom Reimchen has been published in PLOS Biology, sweet! And here’s what it is about:

Blackwater, water with the color of black tea, is found naturally in bog lakes or in the Amazon, but with only red light penetrating the water column, fish have a hard time to detect predators, prey or mates (see photo below). At least they had, until evolution has “repeated itself” in different fish species to improve their color vision in blackwater. We show in our publication that threespine stickleback used the same amino acid changes to invade many blackwater lakes on Haida Gwaii, Canada, and that distant relatives of the sticklebacks living in the Caribbean and in Australia also use the same, but independently evolved changes to do so.


That’s how life in blackwater looks for threespine stickleback: pretty red and no great visibility, even though the water is not turbid. This is a giant stickleback male guarding its nest at Drizzle Lake, Haida Gwaii. Copyright: Thomas E. Reimchen.

In the 1960ies, Tom Reimchen and Eric Moodie discovered the unprecedented morphological diversity of threespine stickleback on the Haida Gwaii archipelago in the North-Eastern Pacific off British Columbia, Canada. In the hundreds of lakes and rivers he studied, he found stickleback populations ranging from full “tank”-like body armor made from bony plates and long spines to complete loss of spines and armor, from 2 cm dwarfs to 12 cm giants, from living between one and 12 years, and from dying after one breeding season or reproducing over multiple years. In his decades of research on the system, Tom demonstrated that predation by trout, birds or insects, the water clarity and the size of a lake are the main factors driving this diversity of stickleback phenotype, life history and behavior. In 2010, Tom teamed up with David Kingsley and Felicity Jones at Stanford University and Federica di Palma at the Broad Institute who sequenced whole genomes of 58 stickleback from Haida Gwaii, to tackle the genetic mechanisms of adaptations to vastly different environments.

Once I started working with this dataset, we discovered that the SWS2 opsin protein, which is used to perceive blue light in the stickleback’s retina, has been mutated at seven amino acids and two of these changes made the protein more sensitive to red-shifted light. This red-shifted SWS2 version allowed sticklebacks to repeatedly colonize deeply stained blackwater lakes on Haida Gwaii after the ice sheets retreated around 10,000 years ago. In addition to these natural populations, we analyzed the data from a selection experiment in which Tom transplanted 100 giant stickleback from a blackwater lake into a tiny, clear water pond in 1993. The genomic data from this experiment now showed that the proportion of the red-shifted protein version rapidly decreased in only 13 generations, as expected from the change in water clarity. Natural selection on color vision thus worked very rapidly on a short timescale.


The evolutionary history of the SWS2 opsin: threespine (Gasterosteus aculeatus) and blackspotted stickleback (G. wheatlandi) have only one SWS2 gene copy, of which Haida Gwaii stickleback have a blue-shifted and a red-shifted variant used to adapt to clear- and blackwater. A gene duplication 200 million years ago in the ancestor of spiny-rayed fish led to two SWS2 copies in most other fish, a blue- and red-shifted copy, which are divergently expressed in clear- and blackwater in bluefin killifish (L. goodei).

When I compared the protein sequences from the Haida Gwaii stickleback to previously published sequences of other fish species, I realized that at least two fish species, bluefin killifish in the Caribbean and black bream in Australia, have adapted to blackwater using the same red-shifted SWS2 and even the same amino acid changes (bluefin killifish). Convergent evolution has thus allowed both killifish, bream and Haida Gwaii stickleback to colonize blackwater habitats with the same molecular mechanism. And evolution was also convergent (or: parallel) within threespine stickleback, as the same allele was used to adapt to blackwater habitat in multiple, replicate lakes on Haida Gwaii within different watersheds. Evolution might thus be much more repeatable and predictable down to the genetic level than what we used to think, even on very different time scales.

Our new publication is freely accessible here: