Signatures of selection in a recent invasion reveals adaptive divergence in a highly vagile invasive species
Data files
Aug 12, 2020 version files 305.32 MB
-
Sample_Metadata_GBS.csv
204.71 KB
-
snps_05_3_5_neutral_bayescan.vcf
151.69 MB
-
snps_05_3_5_outlier_bayescan.vcf
869.82 KB
-
snps_05_3_5.vcf
152.55 MB
Abstract
A detailed understanding of population genetics in invasive populations helps us to identify drivers of successful alien introductions. Here, we investigate putative signals of selection in Australian populations of invasive common starlings, Sturnus vulgaris, and seek to understand how these have been influenced by introduction history. We used reduced representation sequencing to determine population structure, and identify Single Nucleotide Polymorphisms (SNPs) that are putatively under selection. We found that since their introduction into Australia, starling populations have become genetically differentiated despite the potential for high levels of dispersal, and that starlings have responded to selective pressures imposed by a wide range of environmental conditions across their geographic range. Isolation-by-distance appears to have played a strong role in determining genetic substructure across the starling’s Australian range. Analyses of candidate SNPs that are putatively under selection indicated that aridity, precipitation, and temperature may be important factors driving adaptive variation across the starling’s invasive range in Australia. However, we also noted that the historic introduction regime may leave footprints on sites flagged as being under adaptive selection, and encourage critical interpretation of selection analyses in non-native populations.
We collected a total of 568 starling samples from 24 localities across the invasive range in Australia. We recorded Global Position System (GPS) coordinates upon collection, which was used to obtain climate variables for the sample metadata.
The Cornell University Institute of Biotechnology Genomics Facility conducted library construction and Genotyping-by-Sequencing (GBS), using the restriction enzyme PstI. Each individual received a unique barcode before multiplexing 96 individuals per lane (Illumina HiSeq 2000; 6 lanes, 100 bp single-end reads). We processed raw sequence data using TASSEL 3.0 and called SNPs for all individuals using the UNEAK pipeline.
Full details in the associated manuscript.