Evolutionary genomics: Insights from the invasive European starlings

Abstract

Two fundamental questions for evolutionary studies are the speed at which evolution occurs, and the way that this evolution may present itself within an organism's genome. Evolutionary studies on invasive populations are poised to tackle some of these pressing questions, including understanding the mechanisms behind rapid adaptation, and how it facilitates population persistence within a novel environment. Investigation of these questions are assisted through recent developments in experimental, sequencing, and analytical protocols; in particular, the growing accessibility of next generation sequencing has enabled a broader range of taxa to be characterised. In this perspective, we discuss recent genetic findings within the invasive European starlings in Australia, and outline some critical next steps within this research system. Further, we use discoveries within this study system to guide discussion of pressing future research directions more generally within the fields of population and evolutionary genetics, including the use of historic specimens, phenotypic data, non-SNP genetic variants (e.g., structural variants), and pan-genomes. In particular, we emphasise the need for exploratory genomics studies across a range of invasive taxa so we can begin understanding broad mechanisms that underpin rapid adaptation in these systems. Understanding how genetic diversity arises and is maintained in a population, and how this contributes to adaptability, requires a deep understanding of how evolution functions at the molecular level, and is of fundamental importance for the future studies and preservation of biodiversity across the globe.

Keywords: Sturnus vulgaris; museum specimens; plasticity; population genetics; rapid adaptation; structural variants.

FIGURE 1

Summary diagram of evolutionary trends for Sturnus vulgaris (S. vulgaris) within the invasive Australian range. Panel (A) depicts an artist’s image of a male S. vulgaris in breeding season. Panel (B) depicts the Australian range of S. vulgaris (approximately based on eBird data retrieved 2018), with approximate genetic sub-structuring indicated in purple (Australia_EAST) and blue (Australia_SOUTH), and with introduction sites indicated (yellow circles) next to the year of first introduction. Panel (C) depicts differences in genetic differentiation between the two Australian subpopulations (Australia_EAST and Australia_SOUTH) and the native range. F_ST values were obtained from comparisons to Newcastle, UK, from Stuart et al. (2022d). Panel (D) depicts a subset of the chi-squared results that assessed the occurrence of putative outliers and non-outlier SNPs across macro-, micro-, and major sex (Z) chromosome from Stuart et al. (2022d). The table visualises the Pearson residuals, where the circle area is proportional to the amount of the cell contribution, positive residuals (indicating a positive correlation) are in blue, and negative residuals (indicating a negative correlation) are in orange. Panel (E) depicts the positive interaction between phenotypic and genetic dispersion, which is positively affected by the level of ground cover vegetation and annual precipitation variation, data pulled from Stuart et al. (2022a). Panel (F) depicts a revigo (Supek et al., 2011) gene ontology term summary plot of all coding regions that were identified as significant across loci identified in studies on the Australian starling population: all statistically-outlier and environmentally-associated loci (Stuart et al., 2021); all divergent and parallel loci (Stuart et al., 2022d); all statistically outlier SNPs and SVs identified by Bayescan v2.1 (Foll and Gaggiotti, 2008) (Stuart et al., 2022c); all morphologically-associated loci that were also under selection (Stuart et al., 2022a). Log size is indicative of the frequency of the GO terms.

FIGURE 2

Visualisation of the possible outcomes and data requirements of cross study synthesis in evolutionary and population genetics. Panel (A) depicts common research outcomes from studies of invasive, conservation, and agricultural population and evolutionary genomics research, and their overlap. Panel (B) depicts the need for highly curated and reproducible datasets to enable confident cross-study comparisons. Best practices will often be specific to the variant types being called and the research setting they are being conducted in (e.g., Olson et al., 2015; Cameron et al., 2019; Koboldt 2020), and will often be limited by funding and resource availability. In general, minimal best practices should be informed by both the technical limitations of the sequence technology used (e.g., quality checks of the right type and stringency), as well as biologically informed decision making (e.g., accounting for population structuring when conducting downstream analysis such as outlier detection).