Parallelism – or rather lack thereof – in Danish sticklebacks

Sticklebacks2 Lake_Torup

Parallel evolution and the extent to which genes are reused across different populations continues to represent important research questions. In this paper by Anne-Laure Ferchaud and I, we analyzed lateral plate morphology and genomic footprints of selection in three-spine stickleback populations using RAD sequencing (a total of 28,888 SNPs). There are by now several studies that have shown distinct patterns of parallel evolution between marine and freshwater populations, at morphological traits (lateral plate morphs) and at the genomic level, using both SNP chips, RAD sequencing and whole genome sequencing. We wanted to assess if this pattern also applies across freshwater environments that differ significantly from each other, ranging from small lakes with no fish predators and sticklebacks being the dominant species, to large lakes with many fish predators and stickleback being just one among several fish species.

Long story short: we saw surprisingly little parallel evolution across freshwater lakes, and definitely not to the extent that have been reported in other studies. A complicating factor was the finding of signatures of bottlenecks in all freshwater populations (based on analysis of allele frequency spectra), particularly in the two smallest populations. We nevertheless argue that the very low parallelism observed is a genuine finding and not induced by bottlenecks. Among others, the two smallest and most bottlenecked populations were at the same time those that showed the highest degree of parallelism and conformed the most to patterns observed in other studies.

In total, our results suggest that patterns of parallelism depend on heterogeneity across environments. In itself perhaps not such a big surprise, but it stresses the need for considering a wide range of environments before making strong conclusions, for instance about the generality of patterns of parallelism between freshwater-marine stickleback population pairs. It will be interesting to make a direct comparison of these results to RAD data obtained from Greenland sticklebacks. They represent the completely opposite type of environments; simple and species-poor ecosystems that are highly similar across different lakes. So, the prediction there should certainly be lots of parallelism at phenotypic and genomic levels.

Ferchaud AL, Hansen MM (2016) The impact of selection, gene flow and demographic history on heterogeneous genomic divergence: threespine sticklebacks in divergent environments. Molecular Ecology 25, in press. DOI:

This study was funded by the Villum Foundation (grant no. VKR022523 to MMH).

Eels, eels and eels

Eels are tough and long-lived creatures and the same seems to apply to our eel population genomics project. It somehow refuses to die, even though on paper it ought to be finished by now. We published a number of papers in 2014-2015, most of them concerning speciation between European and American eel and the consequences of panmixia in European eel – and fortunately Kristian Sjøgren from the Danish science portal has now written an article about our research. So, in case you want to know what speciation in eels has to do with the Isthmus of Panama and are wondering about the sexual preferences of eels in Iceland, then this is the place to check out. So, here is the link to his story in Danish and in English.

I am sure we will back with more on eels in the future. These creatures are simply so fascinating that we cannot let go of them.

Neighbor-joining tree showing relationships between
American and European eel mitogenomes. From
Jacobsen et al. Heredity 2004.

2014-2015 papers:
Jacobsen MW, Pujolar JM, Bernatchez L, et al. (2014) Genomic footprints of speciation in Atlantic eels (Anguilla anguilla and A. rostrata). Molecular Ecology 23, 4785–4798.
Jacobsen MW, Pujolar JM, Gilbert MTP, et al. (2014) Speciation and demographic history of Atlantic eels (Anguilla anguilla and A. rostrata) revealed by mitogenome sequencing. Heredity 112, 432-442.
Jacobsen MW, Pujolar JM, Hansen MM (2015) Relationship between amino acid changes in mitochondrial ATP6 and life-history variation in anguillid eels. Biology Letters 11, 20150014.
Pujolar JM, Jacobsen MW, Als TD, et al. (2014) Assessing patterns of hybridization between North Atlantic eels using diagnostic single-nucleotide polymorphisms. Heredity 112, 627-637.
Pujolar JM, Jacobsen MW, Als TD, et al. (2014) Genome-wide single-generation signatures of local selection in the panmictic European eel. Molecular Ecology 23, 2514-2528.
Pujolar JM, Jacobsen MW, Bekkevold D, et al. (2015) Signatures of natural selection between life cycle stages separated by metamorphosis in European eel. BMC Genomics 16, 600.
Ulrik MG, Pujolar JM, Ferchaud AL, et al. (2014) Do North Atlantic eels show parallel patterns of spatially varying selection? BMC Evolutionary Biology 14, 138.

A low density SNP resource for sticklebacks


This paper describes development of a SNP resource for three-spine sticklebacks to be used on a Fluidigm platform. In essence, we RAD sequenced ca. 20 individuals from each of one marine and two freshwater populations. This led to identification of > 30,000 SNPs. Based on patterns of genome-wide differentiation, we chose 70 SNPs that are likely to be associated with differential selection in marine and freshwater environments and a further 26 SNPs that we assume mark genomic regions that are not under selection. We are currently using this set of SNPs in studies where we analyze potential hybrid zone dynamics between freshwater and marine sticklebacks and assess selection in recently founded freshwater populations.

R Graphics Output

Anne-Laure Ferchaud had the lead on the paper, NGS was conducted by Beijing Genomics Institute (BGI) and SNPs were genotyped on the Fluidigm machine in Dorte Bekkevold’s lab at DTU Aqua, Silkeborg.

A.-L. Ferchaud, S.H. Pedersen, D. Bekkevold, J. Jian, Y. Niu & M.M. Hansen (2014) A low-density SNP array for analyzing differential selection in freshwater and marine populations of threespine stickleback (Gasterosteus aculeatus). BMC Genomics, 15, 867.

Environmental sinners in Medieval monastries

What is a “natural” population and when did humans start to have a significant impact on genetic population structure? These are highly relevant questions in regions that have been densely populated for many centuries, such as Denmark. This is particularly the case for freshwater habitats where establishment of water mills has taken place since ca. 1100 AD as part of a virtual Medieval industrial revolution; the mills were used for manufacturing wool products and for grinding flour. From ca. 1400 AD permanent weirs and dams were established that effectively restricted both upstream and downstream migration of many organisms, including freshwater fishes (see example here). Many of these water mills were established and run by monks and nuns from the monastries.

brown trout Denmark is all lowland, and under natural circumstances nearly all freshwater habitats within river systems should be accessible for migratory fish species, such as anadromous (sea-run) brown trout (Salmo trutta). In some of the larger and deeper lakes, lake-dwelling brown trout can be found that spawn in tributaries and migrate into the lakes instead of the sea. At the same time, however, most of these lakes are in fact isolated due to water mills. This raises the question whether the lake populations are natural, i.e. founded thousands of years ago; or if they represent remnants of  original, coherent anadromous trout populations that have become fragmented due to establishment of the water mills. In order to answer this question we focused on two lake populations (Lake Mossø and Lake Hald) in the Danish Gudenaa River system, both of which became isolated around 1400 AD when local monastries built impassable dams. We also included anadromous brown trout from the lower part of the river system.

We analyzed 44 microsatellites and used two methods for dating divergence time: IMa (Hey & Nielsen 2007) and Approximate Bayesian Computation (implemented in DIYABC; Cornuet et al. 2008). We were quite amazed by the clarity of the results. Both populations had become isolated ca. 6-700 years bp, coinciding very well with the time the dams were built, and essentially ruling out natural establishment of the lake populations after the last Glaciation. The figure below shows the probability density distribution of divergence time between the two lake populations (A: Lake Hald and B: Lake Mossø) and the downstream anadromous trout population, estimated using IMa. There are three replicates for each divergence time estimate.

QuartzWe also estimated effective population size in the two lake populations to be quite low, on the order of 150-250. Although this is not critically low in the short term, we estimated that only approximately half of the initial potential response to selection remained. Hence, the evolutionary potential of the populations will become increasingly compromised over time. So, even though the munks and nuns were probably quite unaware of this, they were nevertheless environmental sinners in their earthly lives.

M.M. Hansen, M.T. Limborg, A.-L. Ferchaud & J.-M. Pujolar (2014). The effects of Medieval dams on genetic divergence and demographic history in brown trout populations. BMC Evolutionary Biology, 14, 122.

Greenland field work 2014


This year we had two sampling trips to Greenland, though I personally only participated in one of them. The purpose was to sample Arctic char and threespine sticklebacks for our project “Evolutionary potential and contemporary evolution in a changing Arctic environment: Arctic charr and three-spine stickleback in Greenland”. On the first trip in August, Magnus and Shenglin were accompanied by experienced Greenland traveller Terkel Broe Christensen, and they successfully sampled a number of rivers and lakes in Kangerlussuaq, Maniitsoq and the “neighborhood”.

King Kong BarIn early September Shenglin, Magnus and I had the pleasure of being accompanied by Louis Bernatchez (Université Laval, Quebec City, Canada) and Dylan Fraser (Concordia University, Montreal, Canada) to the southernmost tip of Greenland, more specifically Qaqortoq. We started out by being stuck in Kangerlussuaq for two days due to bad weather. It would be an exaggeration to call this the Greenland equivalent of Las Vegas, and as we had already sampled the lakes and rivers in this area we were eager to move on.

DSCN0406We arrived to Qaqortoq several days delayed. Now we were really busy and basically had to do our  sampling in half the time that we originally planned. Fortunately, we had a great guide and boat captain, Emil, and by hard work and some luck we managed to sample all the localities that we had planned.  Much of the sampling of Arctic char took place several kilometres from the river outlets in rough terrain and had to be conducted by rod fishing. However, everybody showed good team spirit and volunteered for this laborious and unpleasant task….. The dog on the photo belongs to a shepherd in the area. It was nice to have it around as fresh polar bear scats had been observed recently in this place; we hoped the bear might consider the dog for a starter before turning on us, but in the end we never saw it.

DSCN0424It was close to spawning season for the Arctic char, and their spawning coloration is often amazing. In some places both “orange” and “red” morphs were present. Do they represent polymorphism within a population or do they in fact represent different populations? Time (and RAD sequencing) will tell.DSC_3697

The sticklebacks were generally less appealing, and the vast majority were filled with endoparasites.They seem to follow the classical pattern of low plate morphs in freshwater and full plated morphs in the sea. Overall, the freshwater ecosystems are very simple and species-poor with Arctic char and three-spine stickleback and in some cases American eel being the only fish species present.

sticklebacksIn the end we managed to sample 6 Arctic char populations and 3 stickleback populations. As an added bonus we discovered an amazing Thai restaurant in Qaqortoq, certainly one of the best  I have ever visited, though the ingredients were slightly unorthodox (musk ox, reindeer, redfish…..).

DSCN0487Southern Greenland was the home of the Norse settlers (Eric the Red, Leif Ericsson and their descendants), who lived there from ca. 985 to at least 1408, but vanished afterwards. We had a closer look at Hvalsey Church, an amazing ruin of a church built ca. 1300.

Next year we will sample rivers and lakes in north-western Greenland. I have started counting the days!