When I started my PhD at the University of Glasgow with Kathryn Elmer in 2013, I didn’t think that a year later I would be back at the place I had just left. I had just spent the previous three years studying for my Bachelor’s degree in Constance, Germany. But there I was, back in Constance, on a cold and foggy November day (Fig. 1), with my office mate and PhD colleague Madeleine Carruthers, ready to start our new project on the whitefish of Lake Constance.

However, the story behind this paper actually began approximately 45 years earlier (way before some of the collaborators on this paper were even born), during a time when the future of the beautiful Lake Constance, and its inhabitants, looked bleak. This was due to the influx of unfiltered nutrients from household waste and agriculture into Lake Constance, and other Alpine lakes, which led to severe eutrophication events – outbreaks of algal blooms and subsequently the depletion of oxygen within the affected lakes - and had an ecosystem-wide impact. The eutrophication of Lake Constance had catastrophic consequences for some of the inhabiting species, particularly the European whitefish.

Prior to the eutrophication of Lake Constance, four whitefish subspecies were present in the Upper basin of the lake, which were separated in diet, spawning time and location. Two benthic species, the shallow-water dwelling sandfelchen (Coregonus lavaretus arenicolus) and deep-water kilch (C. l. gutturosus), and two pelagic species, the blaufelchen (C. l. wartmanni) and gangfisch (C. l. macrophthalmus) (Fig. 2). However, eutrophication resulted in the extinction of the deep-water benthic kilch, the collapse of sandfelchen stocks, and the breakdown of reproductive barriers between the subspecies, causing strong hybridisation and ‘speciation reversal’ (Vonlanthen et al. 2012), with similar scenarios reported for other Alpine lakes that had underwent eutrophication during this time (Vonlanthen et al. 2012). At the end of the eutrophic phase, only the two pelagic subspecies (blaufelchen and gangfisch) remained as viable populations within the Upper basin of Lake Constance, with another benthic subspecies, the weissfelchen, remaining within the Lower basin of the lake, which is connected to the Upper basin by a fast flowing stretch of river. Overall, eutrophication caused a drastic reduction in whitefish diversity in Lake Constance, best documented by the reduction in the range of gill rakers by around 28% across subspecies. For the non-fish geeks here, gill rakers are feeding apparatus in the head of the fish which are used to filter prey items, with long and numerous gill rakers adapted to feeding on fine plankton, and short and few gill rakers adapted to feeding on zoobenthos. 

But this is not the end of the story! In the 1980s, remediation efforts were made to the lower nutrient inputs, and Lake Constance slowly started to return to its pristine oligotrophic state (Fig. 3). At first, no obvious changes were observed in remaining subspecies of European whitefish. However, one day in the late 2000’s, Alfred Sulger, a local fisherman and fisherman for the Limnological institute at the University of Konstanz, told Jasminca Behrmann-Godel over a cup of coffee that he, and some other fishermen and anglers, had recently started catching gangfisch at much deeper depths during spawning than usual. This is interesting for two reasons. First, it tells us that gangfisch have re-expanded their spawning depth range, from a narrower shallow range (up to 10m) during the eutrophication period, to a much broader and deeper range (up to 50m) since the recovery of the lake. Second, European whitefish trophic specialists are known to separate by spawning depth, suggesting that gangfisch within Lake Constance could be exhibiting early signs of diversification. Thus, Jasminca and her colleagues from the Limnological Institute at the University of Konstanz, and Philipp Hirsch from the University of Basel, went to a known gangfisch spawning site, close to the institute (Egg, which was also one of the three spawning sites investigated in our paper), to catch gangfisch at different spawning depths and look for differences in their morphology, trophic niche and genetic composition. And there was the first evidence - gangfisch were diverging in Lake Constance! Jasminca and colleagues uncovered considerable morphological, trophic, and even some genetic divergence in gangfisch along the depth gradient (reported in Hirsch et al. 2013).

At this point, Kathryn and Jasminca decided that it was a good idea to take a closer look at the Lake Constance gangfisch, and investigate in greater detail whether gangfisch were in fact diversifying following the re-oligotrophication of the lake, and if so, how? This was set to be an exciting project, as this would be the first time (to our knowledge) that a collapsed species-complex was observed to have re-diversified following ecosystem recovery. In 2014, Madeleine and I packed our backs to join Jasminca in Konstanz to carry out the field sampling for the planned project, with the invaluable help of local fishermen Hendrik Thiele and Alfred Sulger. Over the course of three weeks, the fishermen caught hundreds of fish and we spent long days in the basement of the Limnological Institute measuring, photographing, and tissue sampling many of them, as well as gutting all of them. All individuals that were not used for our study were returned to the fishermen that helped us. Madeleine also spent days patiently dissecting fish eyes to look for parasites. Kathryn, including family, also joint us to help with the processing of the large number of samples, take us out to the beautiful Konstanz Christmas market for some well-deserved Gluehwein, and to visit friends, as Kathryn had also worked in Konstanz prior to moving to the University of Glasgow. Unfortunately for everyone else working in the Institute, our work made the whole building stink of fish for the entire time. As a compensation, we treated everyone to a nice lunch with freshly smoked whitefish. Some of the pleasures working with a deliciously tasting species.

Nearly three years later, when the genomic, transcriptomic, stable isotope, parasite and phenotypic data were finally all processed, the picture started to become clearer. Gangfisch displayed eco-morphological divergence, mainly in the number of gill rakers and body shape, along the depth gradient, but the extent of this differed across sampling sites within the lake. Furthermore, this phenotypic divergence was coupled to ecological differences across gangfisch, such as differences in parasite abundance and stable isotope ratios, suggesting that gangfisch really do display wide-ranging eco-morphological divergence. However, to show that this diversity was not present at the end of the eutrophication period was not an easy feat. Thankfully, Lake Constance is one of the best studied lakes in Europe and we found enough comparable data to confirm our initial hypothesis that gangfisch had rapidly diversified since the start of the re-oligotrophication of Lake Constance in the late 1980s. This ~25year period corresponds to around just 8 generations in European whitefish! This opened up the question of how gangfisch could diversify so rapidly - was it a purely plastic response or did we actually see an evolutionary change with underlying genetic diversification?

Remarkably, we did find that the observed variation in phenotypic diversity correlated with patterns of neutral genetic divergence in gangfisch, and we were able to identify a range of loci associated with divergent phenotypic traits, such as gill raker number, that were putatively under selection. For us, this suggested that the rapid divergence we observed was not only phenotypic plasticity but did in fact represent rapid evolutionary change. Furthermore, we were able to indirectly show that these putatively adaptive loci under selection most likely introgressed from the other (extant and extinct) whitefish subspecies in Lake Constance into gangfisch during the eutrophication-driven hybridisation and speciation-reversal. This confirmed results from Vonlanthen et al. (2012), showing that microsatellite alleles from the extant profundal-benthic kilch introgressed into gangfisch during the speciation reversal, and suggests that this introgression was adaptive. Finally, using the ecological transcriptomic data, we were able to show that co-expressed gene modules associated with phenotypic divergence were enriched for several biological pathways that that have been previously linked to trophic adaptation and divergence in freshwater fishes.

Overall, our study empirically shows that a species-complex can recover from a drastic collapse in functional diversity when the ecosystem is restored quickly enough, even after decades of disturbance. In their simulation study, Gilman & Behm (2011) suggest that diversity is more likely to re-emerge if this hybridization is temporal and strong, and if a sparse-genetic basis underlies the adaptive phenotypes; since adaptive loci are less likely to get lost under that scenario. Our results fit this hypothesis, but also add that the introgression of adaptive genetic material across (sub)species could potentially support the (partial) re-emergence of functional diversity in a different species, even if the previous ecological specialist went extinct. However, this does not mean that the initial biodiversity can be restored, but only that the lost functional diversity can at least partially re-emerge under the right conditions. It remains to be seen how much of this diversity can re-emerge over time and where the adaptive genetic material in gangfisch actually came from.

Other studies referred to in this post:

Vonlanthen et al. "Eutrophication causes speciation reversal in whitefish adaptive radiations." Nature 482.7385 (2012): 357.

Hirsch et al. "Phenotypic and genetic divergence within a single whitefish form–detecting the potential for future divergence." Evolutionary applications 6.8 (2013): 1119-1132.

Gilman &  Behm "Hybridization, species collapse, and species reemergence after disturbance to premating mechanisms of reproductive isolation." Evolution 65.9 (2011): 2592-2605.