Jacobs, A. | bioRxiv Science

bioRxiv2018

Convergence in form and function overcomes non-parallel evolutionary histories in Arctic charr

Understanding the extent to which evolution is predictable under multifarious selection is a longstanding question in evolutionary biology. However, the interplay of stochastic and contingent factors influencing the extent of parallelism in nature is not well understood. To test the predictability of evolution, we studied a natural experiment on different organismal levels across lakes and evolutionary lineages of a freshwater salmonid fish, Arctic charr (Salvelinus alpinus). We identified significant phenotypic parallelism between Arctic charr ecotype pairs within a continuum of parallel evolution and highly parallel adaptive morphological traits. Variability in phenotypic predictability was explained by complex demographic histories, differing genomic backgrounds and genomic responses to selection, variable genetic associations with ecotype, and environmental variation. Remarkably, gene expression was highly similar across ecotype replicates, and explained the observed parallelism continuum. Our findings suggest that parallel evolution by non-parallel evolutionary routes is possible when the regulatory molecular phenotype compensates for divergent histories.

evolutionary biology↗

bioRxiv2017

Genetic fingerprinting of salmon louse (Lepeophtheirus salmonis) populations in the North-East Atlantic using a random forest classification approach

Caligid sea lice represent a significant threat to salmonid aquaculture worldwide. Lepeophtheirus salmonis is the predominant species that occurs in the Northern Hemisphere. Dispersal of sea lice between marine aquaculture sites and geographic regions is thought to occur rapidly via planktonic transport of larvae. Population genetic analyses have consistently shown minimal population genetic structure in North Atlantic L. salmonis, frustrating efforts to track louse populations, improve targeted control measures and understand local adaption to environmental conditions. The aim of this study was to test the power of reduced representation library sequencing (IIb-RAD sequencing) coupled with random forest machine learning algorithms to define markers for fine-scale discrimination of louse populations. We identified 1286 robustly supported SNPs among four L. salmonis populations from Ireland (N=2, 27 individuals), Scotland (N=1, 11 individuals) and North Norway (N=1, 12 individuals). Weak global structure (FSC = 0.018, p<0.0001) and only one significant pairwise FST comparison was observed (Scotland vs Kenmare Bay, (FST = 0.018, p<0.0001)) using all 1286 SNPs. The application of a random forest machine-learning algorithm identified 98 discriminatory SNPs that dramatically improved population assignment (DAPC assignment probability = 1), increased global Fsc = 0.098, (p<0.0001) and resulted in pairwise comparisons that all showed highly significant Fst-values (range = 0.081 - 0.096, p<0.0001). Out of 19 SNPs found to be under directional selection between populations, 12 corresponded to the discriminatory SNPs identified using random forest. Taken together our data suggest that L. salmonis SNP diversity exists with which it is possible to discriminate differences between nearby populations given suitable marker selection approaches, and that such differences might have an adaptive basis. We discuss these data in light of sea lice adaption to anthropogenic and environmental pressures as well as novel approaches to track and predict sea louse dispersal.

ecology↗