No evidence of selection
We found no evidence for either local adaptation or global selection
that might be linked to the range expansion. For local adaptation we
used FST to compare the expanded regions (Washington
State and the eastern populations) to their nearest native range region
(the Bay Area of California and the Pacific coast of southern
California), but no obvious peaks stood out for either comparison (Fig.
2A–B). Chromosomal mean pFSTs were p = 0.7 ± 0.2 and no
SNPs were significant after Bonferroni or false discovery rate
corrections. Based on the most significant 1% of p-values from pFst, we
found 6,079 SNPs that were shared between the two expanded region
comparisons, only slightly higher than the expected number of high
FST shared SNPs (5,599 SNPs). An alternative to
divergence due to spatially varying selection is that with high gene
flow homogenizing genetic diversity, selection from resources and
environments in the expanded range would affect allele frequencies in
the entire species. However, we also found no evidence of selective
sweeps when analyzing the site frequency spectrum generated from all
samples (Fig. 2C). In fact, the composite likelihood ratio (CLR)
statistics representing the test for selective sweeps were all below
2.0, much lower than standard cutoffs found in many studies .